1. GROUNDWATER MONITORING
  2. GUIDANCE MANUAL
      1. DEPARTMENT OF ENVIRONMENTAL PROTECTION
      2. BUREAU OF WATERSHED MANAGEMENT
  3. TABLE OF CONTENTS
      1. Page
      2. Page
      3. Page
      4. Page
      5. APPENDIX:
      6. FIGURES:
      7. TABLES:
      8. INTRODUCTION
  4. CHAPTER 1: OVERVIEW
      1. 1.1 INTRODUCTION
      2. 1.2 AMBIENT MONITORING
      3. 1.3 COMPLIANCE MONITORING
      4. 1.4 ASSESSMENT MONITORING
      5. 1.5 REMEDIATION MONITORING
      6. 1.6 POST-CLOSURE MONITORING
  5. CHAPTER 2: MONITORING WELL TYPES AND CONSTRUCTION
      1. 2.1 OBJECTIVES OF MONITORING WELLS
      2. 2.2 TYPES OF MONITORING SYSTEMS
      3. Figure 1. Recommended construction of an open borehole well.
      4. Figure 2. Recommended construction of a single screened well.
      5. Figure 3. Example of a well cluster and a multiple screened well.
      6. 2.3 CHOICE OF MONITORING SYSTEM
      7. 2.4 MINIMUM CONSTRUCTION STANDARDS
      8. 2.4.1 Materials
      9. 2.4.2 Assembly and Installation
      10. 2.4.3 Well Development
      11. 2.4.4 Record Keeping and Reporting
      12. 2.5 REFERENCES
  6. CHAPTER 3: LOCATIONS AND DEPTHS OF MONITORING WELLS
      1. 3.1 IMPORTANCE
      2. 3.2 APPROACH TO DETERMINING MONITORING LOCATIONS AND DEPTHS
      3. 3.2.1 Ambient Monitoring
      4. 3.2.2 Compliance Monitoring
      5. 3.2.3 Assessment Monitoring
      6. 3.2.4 Remediation and Post-Closure Monitoring
      7. 3.3 FACTORS IN DETERMINING TARGET ZONES FOR MONITORING
      8. Figure 4. Examples of target zones.
      9. 3.3.1 Groundwater Movement
      10. 3.3.1.1 Geological Factors
      11. Figure 5. Monitoring well screens placed too deeply below the
      12. target zone to detect contamination.
      13. 3.3.1.2 Groundwater Boundaries
      14. Figure 6. Effect of fractures on the spread of contamination.
      15. 3.3.1.3 Karst Terrane
      16. Figure 7. Ineffective monitoring wells in a carbonate aquifer.
      17. 3.3.1.4 Deep Mined Areas
      18. 3.3.2 Contaminant Distribution
      19. 3.4 AREAL PLACEMENT OF WELLS
      20. Figure 8. Determination of monitoring locations at a landfill site
      21. based on assessment reconnaissance studies.
      22. 3.5 WELL DEPTHS, SCREEN LENGTHS, AND OPEN INTERVALS
      23. 3.5.1 Ambient and Compliance Monitoring
      24. 3.5.2 Assessment and Remediation Monitoring
      25. 3.6 NUMBER OF WELLS
      26. 3.7 WELL YIELD
      27. 3.7.1 Fractured Rock
      28. 3.7.2 Heterogeneous Unconsolidated Formations
      29. 3.7.3 Areas of Uniformly Low Yield
      30. 3.8 REFERENCES
  7. CHAPTER 4: ANALYTE SELECTION AND MONITORING FREQUENCY
      1. 4.1 IMPORTANCE OF ANALYTE SELECTION
      2. 4.2 DEVELOPMENT OF AN ANALYTE LIST
      3. 4.2.1 Ambient Monitoring
      4. 4.2.2 Compliance Monitoring
      5. 4.2.3 Assessment Monitoring
      6. 4.2.4 Remediation Monitoring
      7. 4.2.5 Post-Closure Monitoring
      8. 4.3 DETECTION LEVELS AND METHODOLOGIES
      9. 4.4 DURATION AND FREQUENCY OF MONITORING PERIOD
      10. 4.4.1 Ambient Monitoring
      11. 4.4.2 Compliance Monitoring
      12. 4.4.3 Assessment Monitoring
      13. 4.4.4 Remediation Monitoring
      14. 4.4.5 Post-Closure Monitoring
      15. 4.4.6 Cessation of Monitoring
      16. 4.5 REFERENCE
      17. CHAPTER 5: STATISTICAL ANALYSIS OF MONITORING DATA
      18. 5.1 PURPOSE
      19. Chapter Topic
      20. 5.2 STATISTICAL APPROACHES
      21. 5.2.1 Parametric Procedure Assumptions
      22. 5.2.2 Nonparametric Procedure Assumptions
      23. 5.3 DATA AND SAMPLING CONSIDERATIONS
      24. 5.3.1 Data Variability
      25. 5.3.1.1 Spatial Variability
      26. 5.3.1.2 Temporal Variability
      27. 5.3.1.3 Seasonality
      28. 5.3.2 Significance Levels
      29. 5.3.3 Data Sufficiency and Limitations
      30. 5.3.3.1 Independence
      31. 5.3.3.2 Transformations and Distribution
      32. 5.3.3.3 Dataset Size
      33. 5.3.3.4 Outliers
      34. 5.3.3.5 Censored Data
      35. 5.3.4 Network Design
      36. Table 1. Procedures for handling nondetects (from the EPA Addendum, 1992).
      37. Percentage of
      38. 5.4 STATISTICAL PROCEDURES
      39. Figure 9. General guidance on selection of statistical procedures.
      40. 5.4.1 Graphical Procedures
      41. 5.4.1.1 Boxplots
      42. 5.4.1.2 Time Series Plots
      43. 5.4.1.3 Control Charts
      44. 5.4.2 Summary Statistics
      45. 5.4.3 Interval Tests
      46. 5.4.3.1 Statistical Intervals
      47. 5.4.3.2 Tolerance Intervals
      48. Prediction Intervals
      49. 5.4.3.4 Confidence Intervals
      50. 5.4.3.5 Two-Phase Retesting Strategies
      51. 5.4.4 Well-to-Well Comparison Tests
      52. 5.4.4.1 Analysis of Variance (ANOVA)
      53. 5.4.4.2 Kruskal-Wallis Test
      54. 5.4.4.3 Wilcoxon Rank Sum
      55. 5.4.4.4 t-test
      56. 5.4.5 Trend Tests
      57. 5.4.5.1 Considerations
      58. 5.4.5.2 Parametric Trend Tests
      59. 5.4.5.3 Nonparametric Trend Tests
      60. 5.5 REFERENCES
  8. CHAPTER 6: GROUNDWATER SAMPLING TECHNIQUES
      1. 6.1 IMPORTANCE OF SAMPLING TECHNIQUE
      2. 6.2 SAMPLE COLLECTION DEVICES
      3. Table 2. Advantages and disadvantages of different sampling devices.
      4. ADVANTAGES DISADVANTAGES
      5. 6.3 SAMPLE COLLECTION PROCEDURES
      6. 6.3.1 Protective Clothing
      7. 6.3.2 Water Levels
      8. 6.3.3 Field Measurements
      9. 6.3.4 Purging
      10. 6.3.4.1 Criteria Based on the Number of Bore Volumes
      11. 6.3.4.2 Criteria Based on Stabilization of Indicator Parameters
      12. 6.3.4.3 Special Problems of Low Yielding Wells
      13. 6.3.4.4 Summary on Purging
      14. 6.3.5 Management of Purge Water
      15. 6.3.6 Private Wells
      16. 6.3.7 Filtering
      17. 6.3.8 Sample Preservation
      18. Table 3. Suggested procedure for the management of purge water
      19. from groundwater sampling.
      20. TYPE OF
      21. GROUNDWATER
      22. ACTION
      23. 6.3.9 Decontamination of Sampling Devices
      24. 6.4 REFERENCES
  9. CHAPTER 7: WELL ABANDONMENT PROCEDURES
      1. 7.1 INTRODUCTION
      2. 7.2 WELL CHARACTERIZATION
      3. 7.3 WELL PREPARATION
      4. 7.4 MATERIALS AND METHODS
      5. 7.4.1 Aggregate
      6. 7.4.2 Sealants
      7. 7.4.3 Bridge Seals
      8. 7.5 RECOMMENDATIONS
  10. Figure 10. Summary of procedures for well abandonment.
      1. 1,2,3,4,5
      2. 1,2,3,4,5
      3. 1,3,5
      4. 1,3,4,5
      5. 1,3,4,5
      6. 7.5.1 Casing Seal
      7. 7.5.2 Wells in Unconfined or Semi-Confined Conditions
      8. 7.5.3 Wells at Contaminated Sites
      9. 7.5.4 Wells in Cavernous Rocks
      10. 7.5.5 Multiple Aquifer Wells
      11. 7.5.6 Flowing Wells
      12. 7.5.7 Wells with Complicating Factors at Contaminated Sites
      13. 7.5.8 Monitoring Wells
      14. 7.6 EXISTING REGULATIONS AND STANDARDS
      15. 7.7 REPORTING
      16. 7.8 REFERENCES
  11. WELL ABANDONMENT FORM
  12. CHAPTER 8: QUALITY ASSURANCE/QUALITY CONTROL REQUIREMENTS
      1. 8.1 PURPOSE
      2. 8.2 DESIGN
      3. 8.3 ELEMENTS
      4. 8.4 REFERENCE
  13. APPENDIX
      1. DECONTAMINATION PROCEDURES
  14. Bureau of Watershed Management
  15. P.O. Box 8555
  16. Harrisburg, PA 17105-8555
  17. 717-787-5259

GROUNDWATER MONITORING

Back to top


GUIDANCE MANUAL
Commonwealth of Pennsylvania
Department of Environmental Protection
December 1, 2001
For more information, visit us through the PA PowerPort at
www.state.pa.us or visit DEP directly at www.dep.state.pa.us

383-3000-001 / December 1, 2001 / Page i
DEPARTMENT OF ENVIRONMENTAL PROTECTION
BUREAU OF WATERSHED MANAGEMENT
DOCUMENT NUMBER:
383-3000-001
EFFECTIVE DATE:
January 1, 1999
Minor changes were made on pages i, ii, iii, 4, 5, 6, 7, 14, 16, 17, 19, 22, 44,
and 69 (Dec. 1, 2001).
TITLE:
Groundwater Monitoring Guidance Manual
AUTHORITY:
Federal Clean Water Act, Pennsylvania Clean Streams Law, Act of 1937
P.L. 1987 No. 394, as amended
POLICY:
This guidance defines the general principles and practices for monitoring
groundwater quality in Pennsylvania.
PURPOSE:
This manual has been prepared as a guide for DEP hydrogeologists,
groundwater consultants, and private industry for implementing a
comprehensive monitoring program consistent with the established
principles and objectives for protection of the Commonwealth’s
groundwater resources.
APPLICABILITY:
This guidance applies to all local, state and federal agencies and
programs with groundwater quality monitoring responsibilities.
DISCLAIMER:
The policies and procedures outlined in this guidance document are
intended to supplement existing requirements. Nothing in the policies or
procedures shall affect regulatory requirements.
The policies and procedures herein are not an adjudication or a
regulation. There is no intent on the part of the Department to give
these rules that weight or deference. This document establishes the
framework within which DEP will exercise its administrative discretion in
the future. DEP reserves the discretion to deviate from this policy
statement if circumstances warrant.
PAGE LENGTH:
84 pages
LOCATION:
Volume 26, Tab 01F
DEFINITIONS:
None

383-3000-001 / December 1, 2001 / Page ii

Back to top


TABLE OF CONTENTS
Page
INTRODUCTION .................................................................................................................................................... 1
CHAPTER 1: OVERVIEW......................................................................................................................................... 2
1.1 INTRODUCTION...........................................................................................................................................2
1.2 AMBIENT MONITORING.............................................................................................................................2
1.3 COMPLIANCE MONITORING ...................................................................................................................2
1.4 ASSESSMENT MONITORING.......................................................................................................................3
1.5 REMEDIATION MONITORING....................................................................................................................3
1.6 POST-CLOSURE MONITORING..................................................................................................................3
CHAPTER 2: MONITORING WELL TYPES AND CONSTRUCTION ......................................................................... 4
2.1 OBJECTIVES OF MONITORING WELLS.....................................................................................................4
2.2 TYPES OF MONITORING SYSTEMS.............................................................................................................4
2.3 CHOICE OF MONITORING SYSTEM .........................................................................................................7
2.4 MINIMUM CONSTRUCTION STANDARDS ................................................................................................8
2.4.1
Materials........................................................................................................................................8
2.4.2
Assembly and Installation...........................................................................................................9
2.4.3
Well Development.......................................................................................................................9
2.4.4
Record Keeping and Reporting .............................................................................................10
2.5 REFERENCES..............................................................................................................................................11
CHAPTER 3: LOCATIONS AND DEPTHS OF MONITORING WELLS .................................................................... 12
3.1 IMPORTANCE............................................................................................................................................12
3.2 APPROACH TO DETERMINING MONITORING LOCATIONS AND DEPTHS ........................................12
3.2.1
Ambient Monitoring ..................................................................................................................12
3.2.2
Compliance Monitoring...........................................................................................................12
3.2.3
Assessment Monitoring .............................................................................................................12
3.2.4
Remediation and Post-Closure Monitoring...........................................................................13
3.3 FACTORS IN DETERMINING TARGET ZONES FOR MONITORING........................................................13
3.3.1
Groundwater Movement.........................................................................................................15
3.3.1.1
Geological Factors...................................................................................................15
3.3.1.2
Groundwater Boundaries........................................................................................16
3.3.1.3
Karst Terrane ..............................................................................................................17
3.3.1.4
Deep Mined Areas ...................................................................................................20
3.3.2
Contaminant Distribution.........................................................................................................20
3.4 AREAL PLACEMENT OF WELLS................................................................................................................20
3.5 WELL DEPTHS, SCREEN LENGTHS, AND OPEN INTERVALS...................................................................22
3.5.1
Ambient and Compliance Monitoring..................................................................................22
3.5.2
Assessment and Remediation Monitoring ............................................................................24
3.6 NUMBER OF WELLS...................................................................................................................................24
3.7 WELL YIELD ................................................................................................................................................25
3.7.1
Fractured Rock...........................................................................................................................25
3.7.2
Heterogeneous Unconsolidated Formations .......................................................................26
3.7.3
Areas of Uniformly Low Yield....................................................................................................26
3.8 REFERENCES..............................................................................................................................................26

383-3000-001 / December 1, 2001 / Page iii
Page
CHAPTER 4: ANALYTE SELECTION AND MONITORING FREQUENCY............................................................... 29
4.1 IMPORTANCE OF ANALYTE SELECTION.................................................................................................29
4.2 DEVELOPMENT OF AN ANALYTE LIST.....................................................................................................29
4.2.1
Ambient Monitoring ..................................................................................................................29
4.2.2
Compliance Monitoring...........................................................................................................30
4.2.3
Assessment Monitoring .............................................................................................................30
4.2.4
Remediation Monitoring ..........................................................................................................30
4.2.5
Post-Closure Monitoring............................................................................................................30
4.3 DETECTION LEVELS AND METHODOLOGIES.........................................................................................30
4.4 DURATION AND FREQUENCY OF MONITORING PERIOD...................................................................31
4.4.1
Ambient Monitoring ..................................................................................................................32
4.4.2
Compliance Monitoring...........................................................................................................32
4.4.3
Assessment Monitoring .............................................................................................................33
4.4.4
Remediation Monitoring ..........................................................................................................33
4.4.5
Post-Closure Monitoring............................................................................................................34
4.4.6
Cessation of Monitoring ...........................................................................................................34
4.5 REFERENCE................................................................................................................................................34
CHAPTER 5: STATISTICAL ANALYSIS OF MONITORING DATA.......................................................................... 35
5.1 PURPOSE....................................................................................................................................................35
5.2 STATISTICAL APPROACHES......................................................................................................................36
5.2.1
Parametric Procedure Assumptions.......................................................................................36
5.2.2
Nonparametric Procedure Assumptions...............................................................................36
5.3 DATA AND SAMPLING CONSIDERATIONS............................................................................................36
5.3.1
Data Variability ..........................................................................................................................37
5.3.1.1
Spatial Variability......................................................................................................37
5.3.1.2
Temporal Variability .................................................................................................38
5.3.1.3
Seasonality................................................................................................................38
5.3.2
Significance Levels ....................................................................................................................38
5.3.3
Data Sufficiency and Limitations............................................................................................40
5.3.3.1
Independence..........................................................................................................40
5.3.3.2
Transformations and Distribution............................................................................40
5.3.3.3
Dataset Size ...............................................................................................................41
5.3.3.4
Outliers ........................................................................................................................42
5.3.3.5
Censored Data .........................................................................................................42
5.3.4
Network Design..........................................................................................................................42
5.4 STATISTICAL PROCEDURES ......................................................................................................................43
5.4.1
Graphical Procedures ..............................................................................................................45
5.4.1.1
Boxplots.......................................................................................................................45
5.4.1.2
Time Series Plots.........................................................................................................45
5.4.1.3
Control Charts ...........................................................................................................45
5.4.2
Summary Statistics .....................................................................................................................46
5.4.3
Interval Tests................................................................................................................................46
5.4.3.1
Statistical Intervals ....................................................................................................46
5.4.3.2
Tolerance Intervals ...................................................................................................46
5.4.3.3
Prediction Intervals...................................................................................................47
5.4.3.4
Confidence Intervals ...............................................................................................47
5.4.3.5
Two-Phase Retesting Strategies .............................................................................47
5.4.4 Well-to-Well Comparison Tests ................................................................................................47
5.4.4.1
Analysis of Variance (ANOVA)...............................................................................47

383-3000-001 / December 1, 2001 / Page iv
Page
5.4.4.2
Kruskal-Wallis Test ......................................................................................................48
5.4.4.3
Wilcoxon Rank Sum..................................................................................................48
5.4.4.4
t-test ............................................................................................................................48
5.4.5
Trend Tests ...................................................................................................................................48
5.4.5.1
Considerations ..........................................................................................................48
5.4.5.2
Parametric Trend Tests.............................................................................................49
5.4.5.3
Nonparametric Trend Tests.....................................................................................49
5.5 REFERENCES..............................................................................................................................................50
CHAPTER 6: GROUNDWATER SAMPLING TECHNIQUES ................................................................................... 52
6.1 IMPORTANCE OF SAMPLING TECHNIQUE ............................................................................................52
6.2 SAMPLE COLLECTION DEVICES .............................................................................................................54
6.3 SAMPLE COLLECTION PROCEDURES ....................................................................................................55
6.3.1
Protective Clothing ...................................................................................................................55
6.3.2
Water Levels................................................................................................................................55
6.3.3
Field Measurements ..................................................................................................................55
6.3.4
Purging.........................................................................................................................................55
6.3.4.1
Criteria Based on the Number of Bore Volumes.................................................56
6.3.4.2
Criteria Based on Stabilization of Indicator Parameters ...................................57
6.3.4.3
Special Problems of Low Yielding Wells................................................................57
6.3.4.4
Summary on Purging................................................................................................58
6.3.5
Management of Purge Water ................................................................................................58
6.3.6
Private Wells................................................................................................................................59
6.3.7
Filtering.........................................................................................................................................59
6.3.8
Sample Preservation .................................................................................................................59
6.3.9
Decontamination of Sampling Devices................................................................................61
6.4 REFERENCES..............................................................................................................................................61
CHAPTER 7: WELL ABANDONMENT PROCEDURES ........................................................................................... 62
7.1 INTRODUCTION.........................................................................................................................................62
7.2 WELL CHARACTERIZATION......................................................................................................................62
7.3 WELL PREPARATION .................................................................................................................................62
7.4 MATERIALS AND METHODS.....................................................................................................................63
7.4.1
Aggregate ..................................................................................................................................63
7.4.2
Sealants .......................................................................................................................................63
7.4.3
Bridge Seals.................................................................................................................................65
7.5 RECOMMENDATIONS..............................................................................................................................65
7.5.1
Casing Seal .................................................................................................................................67
7.5.2
Wells in Unconfined or Semi-Confined Conditions..............................................................67
7.5.3
Wells at Contaminated Sites ...................................................................................................67
7.5.4
Wells in Cavernous Rocks.........................................................................................................67
7.5.5
Multiple Aquifer Wells................................................................................................................68
7.5.6
Flowing Wells...............................................................................................................................68
7.5.7
Wells with Complicating Factors at Contaminated Sites ..................................................68
7.5.8
Monitoring Wells.........................................................................................................................68
7.6 EXISTING REGULATIONS AND STANDARDS...........................................................................................69
7.7 REPORTING................................................................................................................................................69
7.8 REFERENCES..............................................................................................................................................70

383-3000-001 / December 1, 2001 / Page v
Page
CHAPTER 8: QUALITY ASSURANCE/QUALITY CONTROL REQUIREMENTS ....................................................... 72
8.1 PURPOSE....................................................................................................................................................72
8.2 DESIGN.......................................................................................................................................................72
8.3 ELEMENTS...................................................................................................................................................72
8.4 REFERENCE................................................................................................................................................75
APPENDIX:
Decontamination procedures ................................................................................................................76
FIGURES:
1 Recommended construction of an open borehole well .....................................................................5
2 Recommended construction of a single screened well .....................................................................6
3 Example of a well cluster and a multiple screened well .....................................................................7
4 Examples of target zones .........................................................................................................................14
5 Monitoring well screens placed too deeply below the target zone to detect
contamination ...........................................................................................................................................16
6 Effect of fractures on the spread of contamination ..........................................................................17
7 Ineffective monitoring wells in a carbonate aquifer ..........................................................................19
8 Determination of monitoring locations at a landfill site based on assessment
reconnaissance studies ............................................................................................................................22
9 General guidance on selection of statistical procedures .................................................................44
10 Summary of procedures for well abandonment .................................................................................66
TABLES:
1
Procedures for handling nondetects .....................................................................................................43
2
Advantages and disadvantages of different sampling devices .....................................................54
3
Suggested procedure for the management of purge water from groundwater sampling ......60

383-3000-001 / December 1, 2001 / Page 1
INTRODUCTION
Pennsylvania’s groundwater is a critical resource that provides environmental benefits and
contributes to the well-being of the citizens and the economic growth of the Commonwealth.
Groundwater supplies the drinking water needs of nearly 50 percent of the population in the
state; in rural areas it represents the only practical source of water for domestic uses. High
quality groundwater is important to industry for various commercial and manufacturing
processes and to agriculture for irrigation and livestock watering. Additionally, groundwater is
critical to the protection of Pennsylvania’s surface streams since it provides the sustaining
baseflow to the Commonwealth’s thousands of miles of surface waters.
Adequate protection of Pennsylvania’s groundwater requires periodic monitoring of
groundwater quality. This document is intended to provide guidance on implementation of a
comprehensive statewide monitoring program consistent with the established principles and
objectives for protection and remediation of the Commonwealth’s groundwater resources.
Many state regulatory programs, such as the Land Recycling and Environmental Remediation
Standards Program, have specific monitoring requirements that have been established by
statute, regulation, or policy. This guidance manual does not supersede any of those
requirements.

383-3000-001 / December 1, 2001 / Page 2

Back to top


CHAPTER 1: OVERVIEW
1.1 INTRODUCTION
Monitoring of groundwater is an important component in many permit programs and in the
application of Act 2 of 1995, the Land Recycling and Environmental Remediation Standards
Act (Act 2) to abate unauthorized releases of contamination into groundwater.
The need for and level of monitoring depends upon a number of factors, including:
?? the type of permitted facility requiring groundwater monitoring
?? complexity of local hydrogeologic conditions
?? whether or not the activity is in an aquifer (defined as a geologic formation,
group of formations or part of a formation capable of a sustainable yield of
significant amount of water to a well or spring).
The five basic types of monitoring are:
1) Ambient monitoring (relating to determination of background conditions under
certain permit requirements and to Act 2 Background Standard)
2) Compliance monitoring (relating to determinations of unauthorized releases in
various permit programs)
3) Assessment monitoring (relating to documentation of groundwater pollution or
compliance monitoring results which indicate potential groundwater pollution from a
permitted facility)
4) Remediation monitoring (relating to determination of effectiveness of groundwater
clean-up activities and attainment of remediation goals under Act 2)
5) Post-closure monitoring (relating to determination of levels of contaminants at time of
cessation of certain permitted activities generally related to solid waste
management facilities)
All monitoring activities should incorporate quality control and quality assurance provisions
consistent with existing program regulations and policies (see Chapter 8).
1.2
AMBIENT MONITORING
Ambient monitoring is a relatively short term activity which is conducted to establish
background water quality conditions. The goal is to account for both natural variation
and any man-made impacts that may have influenced groundwater quality. These results
will form a basis against which future monitoring results will be compared to established
background values for specific substances of concern, develop groundwater quality
trend analyses, or determine permit compliance or remediation effectiveness under Act 2
when the Background Standard is selected.
1.3
COMPLIANCE MONITORING
When a regulated activity is authorized or permitted, certain activities may require
compliance monitoring to determine if groundwater has been impacted by an
unauthorized release.

383-3000-001 / December 1, 2001 / Page 3
Compliance monitoring is usually conducted at regularly established intervals during and
following a permitted activity. Compliance monitoring is usually achieved through a
combination of effluent, surface water and/or groundwater sampling.
1.4
ASSESSMENT MONITORING
If compliance monitoring results indicate an unauthorized release into groundwater,
assessment monitoring is usually initiated to determine if the permitted facility actually is
the cause of the groundwater impact prior to beginning any remediation operations
under Act 2. In some cases the assessment monitoring may lead to a determination that
sampling and/or analytical anomalies exist.
1.5
REMEDIATION MONITORING
Groundwater remediation may need to be initiated when an unauthorized release has
been documented through assessment monitoring. Remediation monitoring is
implemented concurrently with groundwater cleanup operations to determine the
effectiveness of these clean-up activities and attainment of remediation goals. All
groundwater remediation should be conducted in accordance with the provisions of Act
2 and associated regulations and guidances.
1.6
POST-CLOSURE MONITORING
In some situations, post-closure monitoring may be required for permitted activities. Post-
closure monitoring is conducted to determine any changes in groundwater quality after
the cessation of a regulated activity. Analytes to be included are those which were
monitored during compliance and/or remediation monitoring. Remediations conducted
under Act 2 do not require monitoring after approval of the final report.
Additional discussions on analytes to be monitored and the duration of monitoring periods
for these five monitoring categories can be found in Chapter 4.

383-3000-001 / December 1, 2001 / Page 4

Back to top


CHAPTER 2: MONITORING WELL TYPES AND CONSTRUCTION
2.1
OBJECTIVES OF MONITORING WELLS
Monitoring wells should be located and constructed to provide the controlled access
necessary to characterize the groundwater system. They must be constructed by a driller
who is licensed by the Commonwealth of Pennsylvania. (Drillers do not need to be
licensed to install lysimeters, temporary well points or in situ sampling probes, or for borings
used to measure the relationship of the water table to the subsurface portion of the
proposed structures.)
Monitoring wells should effectively achieve one or more of the following objectives:
1)
Provide access to the groundwater system for collection of water samples
2)
Measure the hydraulic head at a specific location in the groundwater flow system
3)
Provide access for conducting tests or collecting information necessary to
characterize the aquifer materials or their hydrologic properties
While achieving these objectives, the monitoring system should also preserve the
conditions of the subsurface that is penetrated but not monitored. For example, a well
designed to monitor a bedrock aquifer should be designed and installed with minimal
impact to the flow system in the unconsolidated material overlying the bedrock.
2.2
TYPES OF MONITORING SYSTEMS
Monitoring systems range from the simple to the complex. Each system has its own place
in the monitoring environment. Various types of monitoring systems are described below.
For more detailed descriptions of groundwater sampling devices and installations, see U.S.
EPA (1993) and Nielsen (1992). General recommendations for the construction of single
screened wells and open boreholes are shown in Figures 1 and 2. Site specific
circumstances may require modifications to the recommended construction details.
Open boreholes
- These are holes that are typically drilled into bedrock and left to monitor
groundwater. The overburden (unconsolidated material) is cased off. Recommended
installation details are shown in Figure 1.
Single screened wells
- These wells consist of a prefabricated screen of polyvinylchloride
plastic, stainless steel, etc. that is inserted into an open borehole. Clean sand or gravel is
placed around the annular space of the screen for the entire vertical distance of the
screen length. Recommended installation details are shown in Figure 2.
Well clusters
- Well clusters or a well nest consist of the construction of open boreholes or
screened monitoring wells in one particular location, with each well monitoring a different
depth or zone of groundwater. An example of a well cluster is shown in Figure 3.

383-3000-001 / December 1, 2001 / Page 5
Figure 1. Recommended construction of an open borehole well.
Multiple screened wells
- Wells that isolate specific zones of groundwater for sampling within one
oversized borehole are called multiple screened wells. Each zone is effectively isolated so that
only the desired interval can be accessed for monitoring. An example of a multiple screened
well is illustrated in Figure 3. When constructing multiple screened wells, note that the integrity of
the grout seal is of extreme importance and should be preserved at all times. Improper and
careless construction practices may ultimately create hydraulic communication between each
screened interval rendering each monitoring well unsuitable for monitoring purposes. Because
of the difficulty in constructing a properly sealed well of this type, multiple screened wells should
only be used in uncontaminated areas.

383-3000-001 / December 1, 2001 / Page 6
Figure 2. Recommended construction of a single screened well.
Well points
- Well points are usually short lengths (i.e. 1-3 feet) of screen attached to a hardened
metal point so that the entire unit can be driven, pushed, or drilled to the desired depth for
monitoring. (This method is usually limited to shallow, unconsolidated formations.)
Piezometers
- These are small diameter wells, generally non-pumping, with a very short well
screen or section of slotted pipe at the end that is used to measure the hydraulic head at a
certain point below the water table or other potentiometric surface.
Lysimeters
- A lysimeter is an example of a device used to collect soil moisture that passes
through the vadose zone. Lysimeters typically consist of a porous ceramic cup or caisson (where
water is collected), a sand pack, and collection tubes and vacuum lines. To collect a sample, a
vacuum is drawn on the lysimeter, causing moisture to be pulled into the caisson. The vacuum
lines convey the water to the surface for collection.

383-3000-001 / December 1, 2001 / Page 7
Figure 3. Example of a well cluster and a multiple screened well.
2.3
CHOICE OF MONITORING SYSTEM
The type of monitoring system chosen depends on the objectives of monitoring at the site.
In ambient and compliance monitoring, the monitoring system should offer widespread
opportunity for detection of contamination from the site, while minimizing monitoring costs
in terms of the number of wells to be drilled and the number of samples to be collected.
Once the target zones, or areal locations and depths that are the most likely to be
impacted by the facility are defined, ambient and compliance monitoring (prior to the
detection of any contamination from a site) is often accomplished adequately by using
open rock boreholes or single screened wells that monitor the entire saturated thickness or
a large portion of the target zone.
Where contamination has been detected and definition of vertical contaminant
stratification is desired, wells that monitor more discrete intervals of the target zone or
individual aquifers usually need to be constructed. In this case, well clusters such as shown
in Figure 3 will often be the construction of choice, although open holes that monitor a
short vertical interval or single water bearing zone also may find application. As the flow
beneath the site is better understood, the monitoring system typically will target more
specific depths and locations. This is more likely to occur as a site enters assessment,
remediation or post-closure monitoring. Then discrete zone monitoring may be most
appropriate. Post-closure monitoring can be simple detection or compliance monitoring
using the existing compliance system, if the site has caused no problems up to closure.

383-3000-001 / December 1, 2001 / Page 8
Well points or in situ sampling probes (such as the trade name Hydropunch) can be
valuable reconnaissance tools for preliminary site characterizations or for determining the
locations of permanent monitoring wells (see EPA, 1993). However, in situ sampling probes
can miss a light non-aqueous phase liquid (LNAPL) on the water table, and may have
problems penetrating coarse sands and gravel (where contamination may be located).
Other potential problems include very slow fill times in clayey sediments and significant
capture of fines in the sample.
Lysimeters may be a useful early warning tool. Because the unsaturated zone beneath an
impoundment or a land application (of sludge) site represents a buffer zone between
potential contamination and underlying aquifers, monitoring may provide for early
detection of migrating contaminants such as a landfill leachate. However, lysimeters are
not a substitute for groundwater monitoring wells.
Special well construction will be needed to monitor for certain types of contaminants. For
example, if an LNAPL is a concern, the well screen should be open to the top of the water
table and within the zone of fluctuation, so that the LNAPL contaminants will not be cased
off.
2.4
MINIMUM CONSTRUCTION STANDARDS
To properly meet the objectives listed in Section 2.1, monitoring wells should be designed
and constructed using minimum standards in each of the following categories.
1) Materials
2) Assembly and installation
3) Well development
4) Record keeping and reporting
Different standards and practices may be necessary depending upon the monitoring
objectives of an individual site. Monitoring wells constructed to meet multiple objectives
should employ the standards of the most rigorous objective. For instance, a well point
may be suitable for monitoring hydraulic head, but may not be optimum for collecting
samples. Therefore, a well proposed to monitor head and collect water samples should
be designed as a conventional screened well and not as a well point. In addition,
construction methods, materials, and well development of each point in the plan must not
compromise the objective of other downgradient monitoring wells.
2.4.1 Materials
Materials that are used in construction of a monitoring well should not
contaminate the groundwater being monitored. A list of materials should
include, but not be limited to the drilling tools and equipment, casing, riser pipe,
well screen, centralizers, annular sealant, filter pack, and drilling fluids or additives.
All materials should be of adequate size and of competent strength to meet the
objectives of the monitoring point. All materials introduced into the boring should
be free of chemicals or other contaminants that could compromise the
monitoring well or other downgradient wells. Practices must be employed to
minimize the potential for contamination of the materials during storage,
assembly, and installation. Specific cleaning procedures should be employed in
situations where the materials might introduce contaminants to the groundwater
system. Well screens and risers should be coupled using either water-tight flush-

383-3000-001 / December 1, 2001 / Page 9
joint threads or thermal welds. Solvent welded couplings are not recommended
for monitoring well construction.
2.4.2 Assembly and Installation
Equipment and techniques should be used that create a stable, open, vertical
borehole of large enough diameter to insure that the monitoring well can be
installed as designed while minimizing the impact on the zone(s) being
monitored. When material removed during construction will likely be
contaminated, procedures commensurate with the type and level of
contamination should be followed for the handling, storage, and disposal of the
contaminated material. Whenever feasible, drilling procedures that do not
introduce water or other liquids into the borehole should be utilized. When the
use of drilling fluids is unavoidable, the fluid should have as little impact on the
constituents of interest as possible. If air or other gas is used as the drilling fluid, the
compressor should be equipped with an oil air filter or an oil trap.
The well screen and riser assembly should be installed using procedures that insure
the integrity of the assembly. If water or other ballast is used, it should be of known
and compatible chemistry with the water in the boring. Unless designed
otherwise, the assembly should be installed plumb and in the center of the
boring. Centralizers of proper spacing and diameter can be used. Unless
otherwise approved, the riser should extend above grade and be capped to
prevent the entry of foreign material.
Installation of the filter pack, sealants, or other materials in the annular space
should be done using tremie pipes or other accepted practices. Protective
casing and locking well caps must be installed, and any other necessary
measures must be taken to insure that the monitoring well is protected from
vandalism and accidental damage. To reduce misidentification, all monitoring
wells constructed in developed areas, or in any location where they may be
mistaken for other structures (such as tank-fill tubes, drains, and breather tubes),
should have a locking cap conspicuously labeled "Monitoring Well" (preferably by
the well-cap manufacturer). In addition, locks for the monitoring wells should use
a key pattern different from locks on other structures at the site. It is also
advisable that the well identification number be placed on both the inside and
outside of the protective casing.
2.4.3 Well Development
Well development removes the fine-grained material to improve the hydraulic
efficiency of the well. Well development methods most often include
mechanical surging with bailing or pumping, over pumping, air lift pumping, and
jetting. Well development should proceed slowly and systematically to prevent
the movement of more material than the development method can effectively
remove. When it is likely that the water removed during development will be
contaminated, procedures commensurate with the type and level of
contamination should be employed for the handling, storage, and disposal of the
contaminated material. Development methods should minimize the introduction
of materials that might compromise the objective of the monitoring. If air is used,
the compressor should have an oil air filter or oil trap.

383-3000-001 / December 1, 2001 / Page 10
2.4.4 Record Keeping and Reporting
Because interpretation of monitoring data from a monitoring well is spatially
dependent on both the activity being monitored and other monitoring wells in
the system, records and samples of the materials used to construct and drill the
monitoring well should be kept. In addition, samples of geologic material such as
cuttings, cores, split-spoon samples, formation fluids, etc., should be collected
and preserved. Following construction, accurate horizontal and vertical surveys
should be performed. A permanent reference point should be made by
notching the riser pipe. If possible, all reference points should be established in
relation to an established National Geodetic Vertical Datum (NGVD). Monitoring
well locations should be surveyed to ? 1 linear foot, and monitoring well
elevations should be to the nearest .01 foot. Elevations of the protective casing
(with the cap off or hinged back), the well casing, and the ground surface should
be surveyed for each monitoring well (see Nielsen, 1991). DEP permitted facilities
are generally required to record the latitude and longitude for each monitoring
well.
A groundwater monitoring network report should be prepared. This report should
include copies of the well boring, test pit and exploratory borehole logs; details
on the construction of each monitoring point; maps, air photos or other
information necessary to fully describe the location and spatial relationship of the
points in the monitoring system; and a recommended decommissioning
procedure consistent with the applicable regulatory program and the well
abandonment procedures recommended in Chapter 7.
Monitoring well logs should be prepared and should describe, at a minimum, the
date of construction; the thickness and composition of the geologic units; the
location and type of samples collected; the nature of fractures and other
discontinuities encountered; the nature and occurrence of groundwater
encountered during construction, including the depth and yield of water bearing
zones; and the static water level upon completing construction.
A well completion plan should also be included in the monitoring network report.
Each plan should include information on the length, location, slot size, and nature
of filter pack for each screen; type, location and quantity of material used as
annular seals and filler; description of the type and effectiveness of well
development employed; and notes describing how the well, as constructed,
differs from its original design.
The reports described above do not relieve the driller from the obligation to
submit, for each well drilled, a Water Well Completion Report to the Department
of Conservation and Natural Resources, Bureau of Topographic and Geologic
Survey, as required by Act 610 (the Water Well Drillers License Act).

383-3000-001 / December 1, 2001 / Page 11
2.5
REFERENCES
ALLER, L., and others, 1989, Handbook of Suggested Practices for the Design and
Installation of Ground-Water Monitoring Wells, National Water Well Association.
This publication covers nearly all aspects of design and construction of monitoring wells.
Each of the eight chapters has an extensive reference list.
ANDERSON, K.E., 1993, Ground Water Handbook, National Ground Water Association. A
quick reference containing tables, formulas, techniques and short discussions covering,
among other things, drilling, well design, pipe and casing, and groundwater flow.
DRISCOLL, F.G., 1986, Groundwater and Wells, Second Edition, Johnson Filtration Systems,
Inc., St. Paul, Minnesota 55112, 1089 pp.
GABER, M.S., and FISHER, B.O., 1988, Michigan Water Well Grouting Manual, Michigan
Department of Public Health.
Very thorough coverage of grout and sealant formulation, characteristics, handling, and
placement.
NIELSEN, D.M., (Editor), 1991, Practical Handbook of Ground-Water Monitoring, NWWA,
Lewis Publishers, Inc., Chelsea, Michigan 48118, 717 pp.
Good general reference on the topic.
U.S. ENVIRONMENTAL PROTECTION AGENCY, 1993, Subsurface Characterization and
Monitoring Techniques-A Desk Reference Guide, Volume 1: Solids and Ground Water.
Largely 1 to 2 page thumbnail descriptions of methods and equipment. Copious
references.

383-3000-001 / December 1, 2001 / Page 12

Back to top


CHAPTER 3: LOCATIONS AND DEPTHS OF MONITORING WELLS
3.1
IMPORTANCE
The locations and depths of monitoring wells are the most important aspects of a
monitoring network. A monitoring point that is misplaced is of little use, and may
misrepresent the quality of the groundwater migrating to or from a site. On the other
hand, a properly positioned monitoring well that detects the earliest contamination could
save both time and money spent on cleanup of a site, and prevent extensive
contamination of groundwater.
3.2
APPROACH TO DETERMINING MONITORING LOCATIONS AND DEPTHS
Different approaches and efforts for determining the location and depth of wells may be
necessary based on the type of monitoring to be done. However, before well locations
are chosen for any type of monitoring, the existing data should be assessed. This can
reduce the costs of implementing the monitoring program, and can help to make
appropriate choices for three-dimensional monitoring locations.
Information may be obtained through site visits, site records and previous studies,
interviews with present and past workers, aerial photographs, publications on the local
and regional hydrogeology, geophysical surveys, borings, wells, aquifer tests, etc. If
enough information is available, the designer can determine the groundwater flow paths
and design a complete monitoring network. However, actual testing of aquifer
parameters provides the best information to evaluate placement of monitoring wells,
especially in newly established sites or facilities where little site information is available.
3.2.1 Ambient Monitoring
The determination of background water quality is paramount to understanding the
effect of an activity or site on groundwater quality. Often insufficient site
information is available so that initial well locations may depend on assumptions
regarding groundwater flow. If subsequent information shows that monitoring wells
are misplaced, new wells should be installed. Act 2 regulations (Chapter 250,
Subchapter G) provide requirements for establishing numerical values for regulated
substances.
3.2.2 Compliance Monitoring
Appropriately placed monitoring points are necessary to detect the spread of
contamination due to an unauthorized release from a permitted activity. The
more that is known about the (potential) contaminant flow paths and the site,
the more likely that compliance wells will be optimally placed to monitor the
impact of the permitted activity on groundwater quality. Monitoring well
locations should be concentrated in those areas that will first be impacted by the
facility, which typically will be located within or comprise the uppermost aquifer.
3.2.3 Assessment Monitoring
The greater the complexity of the hydrogeology and the spread of
contamination, the more monitoring may be necessary to assess the
contamination. Where a contamination plume of unknown dimensions exists,

383-3000-001 / December 1, 2001 / Page 13
various techniques (such as geophysics, soil vapor studies, etc.) can be used to
estimate the extent and magnitude of contamination. Such methods can be
used to focus the investigation and then properly position wells that will confirm
the studies and complete the assessment to determine if the permitted activity is
the cause of the unauthorized release.
3.2.4 Remediation and Post-Closure Monitoring
Further knowledge of groundwater flow directions, aquifer properties, and
contaminant distribution may be necessary to select appropriate well locations
where cleanup can be monitored and confirmed. Existing wells may be used for
remediation monitoring; however, the impact of the remediation method (such
as pumping the aquifer) on the groundwater flow paths should also be
considered. Appropriately placed wells will allow for an accurate assessment of
background water quality, which will be needed for determining the most
suitable cleanup standard under Act 2, and determination that the remediation
goals have been attained.
Well locations for post-closure monitoring are generally selected from existing
compliance monitoring wells. Where a source of contamination is removed prior
to impacting groundwater, post-closure monitoring should continue at locations
that will detect any residual contamination in the unsaturated zone that might
migrate to the groundwater.
3.3
FACTORS IN DETERMINING TARGET ZONES FOR MONITORING
The prime requirement for a successful monitoring system is to determine the "target"
zones - the areal locations and depths that are the most likely areas to be impacted by
the facility being monitored or site being investigated. The dimensions of target zones
depend on the vertical and horizontal components of flow in the aquifers being
monitored, the size of the facility being monitored, the potential contaminants, and the
distance that contamination may have traveled from the facility. Figure 4 shows how
different target zones could be formed based on these factors.
Horizontal and vertical components of groundwater flow are best determined by
constructing planar and cross-sectional flow nets based on the measurement of water
levels in piezometers. Where the vertical components of flow are negligible, wells, rather
than piezometers, drilled into the aquifer to about the same depth, will allow preparation
of a contour map of water levels representing horizontal flow. This should be adequate
to prepare a planar flow net and determine the target zone.

383-3000-001 / December 1, 2001 / Page 14
Figure 4. Examples of target zones.
With regard to upgradient wells, target zones (as defined above) do not exist.
Upgradient wells should be drilled to depths that are screened or open to intervals similar
to that of the downgradient wells, or to depths that yield water that is otherwise most
representative of the ambient quality of the water being monitored by the
downgradient wells. The latter is especially true for sites where no true upgradient flow
exists for the site.
The numerous site details to consider when establishing target zones may be grouped
into groundwater movement or the distribution of contamination.

383-3000-001 / December 1, 2001 / Page 15
3.3.1 Groundwater Movement
In what direction is groundwater flowing? If flow paths are not easily determined,
what will influence the direction of groundwater flow? The answers to these
questions are critical to selecting target zones and the optimal locations of
monitoring wells.
Using the groundwater levels from piezometers or wells at the site, the
groundwater gradient can also be determined. At least three monitoring points
are needed to determine the horizontal gradient; however, at some sites,
knowledge of the vertical component of flow may be important. This is best
accomplished by using well pairs of "shallow" and "deep" piezometers or short-
screened wells.
It may appear to be a simple task to place monitoring wells in downgradient
positions using a map of the groundwater elevation contours, or by anticipating
the gradient based on topography or discharge points. However, at many sites,
three-dimensional flow zones must be understood to install appropriate
monitoring points (see Section 3.5). Figure 5 shows how a well can miss the
vertical location of contamination at a site. Water level measurements,
piezometer and well construction, and groundwater gradient maps should be
reviewed carefully when assessing the dimensions of target zones.
3.3.1.1
Geological Factors
The geology of a site can complicate the selection of the target zones
for monitoring. Geological factors can produce aquifers that are
anisotropic. In an anisotropic aquifer, groundwater moves faster in one
direction than another, and oblique to the hydraulic gradient.
Anisotropy can result from various sedimentary or structural features
such as buried channels, bedding planes, folds, faults, and fractures.
In Pennsylvania, most of groundwater flow is through fractured rocks.
Fracture flow in bedrock (or hardened sediments) requires additional
considerations compared to flow in unconsolidated materials.
Consolidated materials may exhibit small effective porosities and low
hydraulic conductivities that impede groundwater flow. However, the
development of secondary porosity may allow substantial flow of
groundwater through fractures, joints, cleavage planes and foliations.
These features tend to be highly directional, exhibit varying degrees of
interconnection, and may produce local groundwater flow regimes
that are much different from the regional trends.
Geological factors influence the direction of groundwater flow by
controlling the transmissivity. For example, Figure 6 shows the effect of
fractures on the spread of contamination. Although the gradient
indicates flow to the north, groundwater also follows the major fractures
and spreads to the northeast. Monitoring wells "1" and "2" located to
the north of the site may detect contamination, but the lack of a
monitoring well to the northeast will miss an important direction of
migration. Common sedimentary bedding planes also could have a
similar effect on groundwater flow.

383-3000-001 / December 1, 2001 / Page 16
Figure 5. Monitoring well screens placed too deeply below the
target zone to detect contamination.
3.3.1.2
Groundwater Boundaries
The presence of hydrogeologic boundaries should also be considered
when locating wells or approving a groundwater monitoring system.
Important types of boundaries include the following:
Geologic faults
- Fault planes that contain gouge (soft rock material) or
bring rock bodies of widely differing hydraulic conductivity into
juxtaposition can influence groundwater flow direction and velocity.
Location of downgradient wells across fault zones or planes should not
be approved until the nature of the influence of the fault zone on
groundwater flow has been evaluated. One method of evaluating
fault zones is to conduct pumping tests with wells on either side of the
fault plane to evaluate the degree of hydraulic connection.
Dikes
- Diabase dikes, common in southeastern Pennsylvania, can
function as lithologic barriers to groundwater flow because of their very
low permeability. If a dike lies between a site and a proposed
downgradient well, the role of the dike should be evaluated prior to
approving the well’s location.

383-3000-001 / December 1, 2001 / Page 17
Figure 6. Effect of fractures on the spread of contamination.
Others
- Geologically "tight" layers or formations can function in a similar
way: they can create subsurface "dams" that cause groundwater to
flow in unexpected directions. Additional boundaries to flow can
include inclined confining beds, groundwater divides, and artesian
aquifers.
3.3.1.3
Karst Terrane
Limestone and dolomite are very often susceptible to the formation of
sinkholes, solution channels, and caverns. In Pennsylvania, almost all
carbonate rocks will exhibit some karst development. Resulting flow
patterns can be very complicated; flow depends on the degree of
interconnection of the joints, fractures, and solution openings (small and
large), the hydraulic gradient, and geologic barriers. Properly
monitoring a site in a karst area can be very difficult. Even a relatively
small cavernous opening with its connecting drainage paths can
control a significant amount of the flow from an area, and may
perhaps effectively carry all the groundwater that discharges from
underneath a site. In addition, karst geology has the potential to
rapidly transmit groundwater over a large distance.
Groundwater flow in a karst terrane can be highly affected by
precipitation events, and groundwater divides can be transient. To
determine monitoring locations in limestone and dolomite areas, the
monitoring designer should investigate the degree to which the rocks
are susceptible to dissolution. The more dissolution features that are
recognized, the more likely that conduit flow will occur.

383-3000-001 / December 1, 2001 / Page 18
Thus, it would seem that monitoring locations should be based on major
conduits of flow. However, Figure 7 shows how a monitoring well can
easily miss a primary conduit. It may be futile to attempt to establish
the locations of such flow zones, because they probably represent only
a small fraction of a site. However, several procedures can be used to
increase the odds of monitoring the facility of concern (Note that many
of the procedures discussed here also can be used in deep-mined
areas and other types of fractured rocks).
Tracer tests
- Tracer tests offer the best possibility of determining where
groundwater is flowing and discharging. They are conducted to
establish a hydraulic link between a downgradient monitoring point
and the facility of concern. Tracer tests should be combined with a
thorough inspection for local and regional springs that could serve as
discharge points for groundwater at the site. It also could be possible
that groundwater beneath a site could discharge to several springs or
that the flow directions could be different during flood stages. A
determination of the point of regional base flow also should be made
and possibly included as a monitoring point.
It is important to understand the potential chemical and physical
behavior of the tracer in groundwater. The objective is to use a tracer
that travels with the same velocity and direction as the water and does
not interact with solid material. For most uses, the tracer should also be
nontoxic. It should be easily detected, and be present in
concentrations well above natural background quality. The tracer
should not modify the hydraulic conductivity or other properties of the
medium being studied. Investigations using tracers should have the
approval of local authorities or the department and local citizens
should be informed of the tracer injections.
Various types of tracers are used including water temperature, solid
particles, ions, organic acids, and dyes. Fluorescent dyes are the most
common type of tracer used in karst areas. These dyes are used
because they are readily available, are generally the most practical
and convenient tracers, and they can be adsorbed onto activated
coconut charcoal or unbleached cotton. Fluorescent dyes can be
detected at concentrations ranging from one to three orders of
magnitude less than those required for visual detection of non-
fluorescent dyes. This helps to prevent the aesthetically unpleasant
result of discoloring a private or public water supply.
Fluorescein (CI Acid Yellow 73 - C
20
H
10
O
5
Na
2
) is one of the most widely
used water-tracers in karst terrane studies because of its safety,
availability, and ready adsorption onto activated coconut charcoal. It
is a reddish-brown powder that turns vivid yellow-green in water, is
photochemically unstable, and loses fluorescence in water with pH less
than 5.5.

383-3000-001 / December 1, 2001 / Page 19
Figure 7. Ineffective monitoring wells in a carbonate aquifer.
The toxicity of the dyes should also be considered, especially when
there is a chance of private or public water supplies being affected.
Smart (1984) presents a review of the toxicity of 12 fluorescent dyes.
Other excellent references include U.S.EPA and the USGS (1988) and
Davis and others (1985).
The mapping of outcrops and associated joints and faults may
distinguish directional trends that groundwater might follow. Fracture
trace analysis using aerial photographs may detect local and regional
trends in fractures, closed depressions, sinkholes, stream alignments,
and discharge areas. However, tracer tests are recommended to verify
where groundwater is flowing.
Additional site investigation techniques may be helpful in determining
flow paths. Geophysical methods such as self-potential (a surface
electromagnetic method) and ground penetrating radar can enhance
the understanding of karst systems.
Effort should be made to monitor at or near the site of concern, rather
than depend on springs that discharge away from the site. Wells sited
on fractures traces or other structural trends can be tested with tracers
to see if they intercept groundwater flowing from the site. A monitoring
network should not be solely dependent on water levels to establish the
locations of monitoring wells in such fractured rock settings. These
uncertainties and the potential traveling distances may cause
monitoring in karst areas to be involved and expensive.

383-3000-001 / December 1, 2001 / Page 20
3.3.1.4
Deep Mined Areas
When designing a groundwater monitoring program for a site in which
coal or noncoal deep mining has occurred, whether it is a mine site,
landfill, or industrial cleanup site, it is necessary to consider the
underlying mine.
Because of the extensive mine workings and the associated subsidence
fractures, the deep mine often acts as a large drain for the overlying
water bearing zones. Groundwater monitoring of this zone should be
considered because it is the first saturated zone available for
contaminant detection.
Saturated zones within deep mines may be characterized as a mine
pool, which is a body of water at a relatively stable elevation, or it may
be a pathway for channelized water. Because of these special
problems, a drilling plan should be devised that includes provisions for
drilling through the coal pillar, mine void or collapsed structures.
Several attempts should be made at each well location to intercept
the pool, saturated zone and/or mine void.
Well construction requires the placement of a grout basket or plug
attached to the riser pipe that is placed above the zone to be
monitored. This helps seal the bentonite grout.
3.3.2 Contaminant Distribution
In addition to normal groundwater flow (advection), the distribution of
contamination is critical to the correct placement of monitoring points. This
distribution is based on 1) the chemical characteristics that affect the migration
of the contaminant, and 2) its occurrence or source at the site. For example, the
density of a contaminant is one of the most important factors in its distribution in
the aquifer, and especially for determining the depth of a target zone (see
Section 3.5). Isoconcentration maps can be useful in plume interpretation and for
placement of groundwater recovery wells. Also, the designer of the monitoring
network should keep in mind the relationship of the flow lines with the activity's
location or potential sources of contamination.
3.4
AREAL PLACEMENT OF WELLS
For establishing the target zones, the monitoring system designer should consider the
topics of groundwater movement and contaminant distribution that were discussed
above. For the initial placement of wells at a site where little information is available, the
downgradient well position is typically assumed to be downslope. In apparent flat-lying
sites, drainage patterns can be used to estimate the gradient. The site boundary that is
closest to a body of water is a likely choice for downgradient well locations. An
upgradient well is typically placed upslope.
As more information is obtained about the site, groundwater gradients will be more
accurately defined. Upgradient and downgradient monitoring points may need to be
moved. However, even well-defined groundwater gradient maps should be evaluated
carefully when choosing the target zones for upgradient and downgradient wells.
Because of structural controls in fracture flow described in Section 3.3.1, groundwater
can move obliquely to the regional gradient. Some monitoring points may need to be
moved as target zones are refined.

383-3000-001 / December 1, 2001 / Page 21
In general when comparing sites, intervals between monitoring wells probably should be
closer for a site that has:
?? a small area
?? complicated geology such as folding, faulting, closely spaced fractures, or solution
channels
?? heterogeneous lithologies and hydraulic conductivities
?? steep or variable hydraulic gradient
?? high seepage velocity
?? had liquid contaminants
?? buried pipes, trenches, etc.
?? low dispersivity potential
Sites without these features may have well interval distances that are greater. See also
Section 3.6 on the number of wells.
For assessment monitoring, reconnaissance tools and screening techniques such as
surface geophysical techniques and soil gas studies can help to locate plumes before
wells are drilled and thus help to determine optimal well locations. Methods for selecting
sample locations range from random picks to probability sampling (such as a grid
pattern). Random sampling is very inefficient. When selecting many monitoring points in
an area where little is known, such monitoring points should be placed in a grid or
herringbone pattern.
When selecting the areal locations for wells, the chosen sites should monitor any major
branches of flow of the target zones. Figure 8 depicts a landfill site that straddles a
groundwater divide. Here an assessment has been required. The groundwater gradient
indicates flow to the north and south, but a terrain conductivity study and data results
from existing monitoring wells indicate the spread of contamination in at least three
major directions. The migration is strongly affected by northeast and northwest trending
fractures. These three major areas (A, B, and C) should be included in the assessment
and remediation monitoring. In addition, groundwater divides, site boundaries, and
objectives should be considered.

383-3000-001 / December 1, 2001 / Page 22
Figure 8. Determination of monitoring locations at a landfill site
based on assessment reconnaissance studies.
3.5
WELL DEPTHS, SCREEN LENGTHS, AND OPEN INTERVALS
The first zone of saturation is typically an unconfined or water table aquifer, which is
recharged from direct infiltration of precipitation. Impacts to the aquifer under
unconfined conditions are more easily evaluated than under confined conditions. The
shallowest aquifer should be the target zone for chemicals and substances that are less
dense than water.
Sites with confined aquifers that have the potential to be impacted will need to be
evaluated in combination with the unconfined aquifer. Such a situation would require
more detailed vertical and discrete zone monitoring.
3.5.1 Ambient and Compliance Monitoring
Once the subsurface geometry of the monitoring target zone is determined,
decisions can be made with respect to the depth and screen lengths of
individual wells that will be used for ambient and compliance monitoring.
Ambient and compliance monitoring networks should monitor the entire
saturated thickness of the target zone or a very large percentage of it. If large
vertical intervals of the target zone are unmonitored, chances are increased that
groundwater contamination may go undetected, or be underestimated if
detected.

383-3000-001 / December 1, 2001 / Page 23
Choosing the length of the open interval in a monitoring well is in many respects
a balancing act. Shorter open intervals or screen lengths provide better accu-
racy in determining hydraulic head at a specific point in the flow system. If a
sufficient number of shorter well screens or open intervals are stacked or clustered
vertically so that the entire saturated thickness of the target zone is adequately
monitored, they will, when taken together, provide better resolution of the
vertical distribution of any contamination that may be detected. In addition, the
possibility of cross-contamination is minimized. Disadvantages of shorter intervals
include reduced water volume from each well, and the increased cost of
installing, sampling, analyzing, and interpreting the data from the more numerous
sampling points, which can be considerable.
Some disadvantages also are likely for longer screen lengths or open intervals.
Resolution of hydraulic head distribution in the aquifer decreases, contamination
entering the well at a specific point may be diluted by other less contaminated
water, and there is less certainty regarding where water is entering the well.
It would be preferable from a strictly technical point of view to monitor the entire
saturated thickness of any target zone with a number of individual, shorter
screened wells drilled to different depths that together monitor the entire target
zone. However, the hydrogeologist designing the project must decide if the
increased cost over single, longer screened wells is justified for ambient and
compliance monitoring. The goal is to establish screens and open intervals that
will detect as quickly as possible any contamination emanating from any portion
of the facility.
In many cases, ambient and compliance monitoring can be and is accom-
plished by using relatively long screen lengths or gravel packed intervals, or open
intervals in open rock holes. Exceptions to this include sites where different aquifer
systems are being monitored (such as unconsolidated deposits overlying a
bedrock formation) or where strong enough vertical gradients exist to lead to
concerns about introducing any contamination that might occur into uncon-
taminated aquifers.
Compliance (detection) monitoring should not be conducted by using a single
short screened well that only monitors a small percentage of the target zone.
Care should also be taken when monitoring target zones in bedrock formations.
In this case, by geologic necessity, the portion of the target zone which is moni-
tored will be determined by the location and number of water producing frac-
tures that are intercepted by the well. Care must be taken not to drill wells too
deeply below the target zone in search of a water-producing fracture.
An exception to the goal of monitoring the entire saturated thickness would be in
the case of an aquifer that was underlain by an unsaturated zone such as a mine
opening. Here a well drilled through the aquifer into a mine opening would drain
the aquifer. In such cases, well construction should prevent dewatering of the
aquifer.
Where multiple aquifers exist, such as an unconsolidated aquifer overlying a
bedrock aquifer, or where two permeable aquifers are separated by an aquitard,
the target zones within each aquifer should be monitored separately.

383-3000-001 / December 1, 2001 / Page 24
The specific gravity of a contaminant and whether it will most likely be introduced
to the environment as a free phase or in a dissolved phase also will influence how
a well is constructed. In conducting monitoring for an LNAPL (light non-aqueous
phase liquid) contaminant, such as gasoline, wells should be constructed with
screens or open intervals that intercept the water table surface at all times of the
year. Then, LNAPL can accumulate into a distinct layer and flow into the
monitoring well. For materials that exhibit specific gravities greater than water
(such as many chlorinated solvents), it is desirable, though not always possible to
locate subsurface boundaries on which such contaminants might accumulate if
released to the environment in a free phase.
3.5.2 Assessment and Remediation Monitoring
The major purpose of assessment monitoring is to determine the vertical and
horizontal extent and magnitude of contamination that has been detected
during compliance monitoring. In most cases this will require the installation or
modification of wells so that they are screened or open to relatively short vertical
intervals within each target zone. This work is desirable for resolving any
stratification of contamination and to establish the maximum depth of
contamination. This information will be useful in targeting remediation options to
those portions of the aquifer that are most contaminated and that serve as
significant sources of contamination to other portions of the flow system.
If the assessment reveals a significant vertical component of flow, then the
possibility exists for long-screened wells to act as conduits for current or future
contamination of previously uncontaminated portions of the target zone. These
detection wells should be grouted or their construction modified to prevent this
outcome.
Remediation monitoring will most likely be conducted in wells that have been
drilled for the compliance or assessment phases. In some cases wells will be
drilled for the recovery of groundwater. Obviously these will be designed and
drilled at locations to maximize their effectiveness in capturing contaminated
groundwater. As long as these wells are pumping and recovering groundwater,
concerns with their construction are minimal; nevertheless, if their use as recovery
wells ceases for any extended period of time prior to restoration of the aquifer to
appropriate Act 2 cleanup standards, and adequate justification for not sealing
the wells cannot be provided, they should be properly abandoned without delay
in accordance with the procedures described in Chapter 7.
3.6
NUMBER OF WELLS
The number of wells needed depends on site-specific factors. Compliance monitoring
may need only one downgradient well for a small site such as an underground storage
tank. In general, the spacing of background or upgradient wells should be adequate to
account for any spatial variability in the groundwater quality. Downgradient wells should
be positioned to adequately monitor the activity and any other variability of the
groundwater quality. The estimate of the separation distance will depend on the extent
and type of activity, the geology, and the potential contaminants (see also Section 3.4
on the Areal Placement of Wells, and Section 5.3.4 on Network Design).
For ambient and compliance monitoring, the monitoring well network should cover most
of the site. It is recommended that at least 85 percent of the site be monitored. The

383-3000-001 / December 1, 2001 / Page 25
percentage can be adjusted based on the knowledge of the site, groundwater flow,
and the potential or existing contaminants. For example, it might be reasonable to
require the well network to cover 95 percent of a site where little information is available.
Percentages can be estimated using computer models such as the MEMO model (Wilson
and others, 1992). In the absence of models to estimate the coverage of the monitoring
well network, best professional judgment should be used to assure the network would
detect a plume of reasonable size.
As the monitoring network is refined, active monitoring wells can be established. In some
cases, wells can be dropped from active monitoring if it can be shown that the target
zones are being monitored completely. For remediation or post-closure monitoring, a
lower percentage of coverage may be appropriate after that information on the site has
been obtained.
3.7
WELL YIELD
Monitoring wells should produce yields that are representative of the formation being
drilled. Wells that are located in anomalously low yielding locations are undesirable for
several reasons. First, flow lines tend to flow around rather than through low permeability
areas. This in effect results in contaminants bypassing low permeability areas and failing
to be detected in representative concentrations. In addition, by the time a contaminant
shows up in a very low yielding well that is unrepresentative of the formation, other
contamination may have traveled extensively downgradient beyond the monitoring
well. Therefore, in settings where well yields are variable, the best monitoring wells will be
those that are open to the highest permeability flow lines that are potentially able to be
contaminated by the site.
The best information regarding representative yield for the target zones selected for a
particular site should come from the wells and borings used in the investigation to
determine the groundwater flow system for the site. Borehole geophysics can be a
valuable tool for determining the location of yielding zones and the presence of
contaminants. For more detailed descriptions of borehole geophysical techniques and
devices, see EPA (1993) Chapter 3 - Geophysical Logging of Boreholes, and Nielsen
(1991). Additional regional hydrogeologic information may be obtained from:
?? The Pennsylvania Bureau of Topographic and Geologic Survey (BTGS)
?? The United States Geological Survey (USGS)
Water Resource Reports have been published by the USGS and BTGS for selected
counties and areas in Pennsylvania. They are available through the State Bookstore.
In Pennsylvania, there are three general hydrogeologic settings that merit special
discussion from a standpoint of well yield.
3.7.1 Fractured Rock
In aquifers composed of fractured bedrock, groundwater flow is generally
restricted to the fractures. If a well fails to intersect any fractures or a very few
small fractures, the well will not detect contamination, or will be inefficient in
detecting contamination. For this reason, wells that fail to intersect fractures in
the target zone that are representative of the formation should be approved with
caution and wells that are essentially dry are not acceptable. Such wells should

383-3000-001 / December 1, 2001 / Page 26
be relocated nearby and another attempt made to obtain a better yield, when it
is determined that it is likely that more representative yields can be obtained.
Likewise, wells drilled below the proper target zone, strictly in an effort to obtain
yield, should not be used for site characterization.
3.7.2 Heterogeneous Unconsolidated Formations
Low permeability clay-rich formations with interbedded or lenticular, higher
permeability sand or gravel units can present a significant challenge to designers
and installers of monitoring wells. Wells need to be located so that they are open
to any high permeability zones within the target zone that are hydraulically
connected to the site being monitored. These wells will be higher yielding than
their counterparts that are drilled exclusively into the clay-rich portions of the site.
3.7.3 Areas of Uniformly Low Yield
Certain geologic formations and hydrogeologic settings are characterized by
naturally low permeability over a wide area. Other geologic formations may
exhibit low permeability locally in certain settings such as ridge tops, steeply
dipping strata, or slopes. In such settings, a permanent or seasonal perched
water table or shallow flow system may develop on the relatively impermeable
bedrock that may or may not be hydraulically connected to the bedrock system.
Depending on the permeability of the soils and unconsolidated material overlying
the solid, slowly permeable bedrock, the shallow groundwater flow can express
itself as a rather rapid "subsurface storm flow" or a more sluggish, longer lasting
saturation in poorly drained soils.
It is important to be sure that the shallow systems are part of the target zone of
the site being monitored. In these cases the shallow system may constitute the
most sensitive target zone for monitoring a facility. While wells drilled into the
bedrock system may be needed to monitor for vertical flow of contaminants, the
importance of sampling monitoring wells or springs in the shallow intermittent flow
system should not be underestimated, although the usual periodic monitoring
schedules may not be appropriate in these settings. If the systems are
intermittent, one will have to become aware of when they are active (e.g. in the
spring, after significant precipitation) and be prepared to monitor the systems at
that time. Monitoring can be conducted in wells, springs that are properly
developed, or in some cases, by sampling man-made underdrain systems that
are constructed to collect the shallow flow system.
3.8
REFERENCES
DAVIS, S.N. and others, 1985, Ground Water Tracers, through the U.S. EPA. Cooperative
Agreement CR-810036.
DOBRIN, M.B., 1965, Introduction to Geophysical Prospecting, 3rd ed., McGraw-Hill, New
York, 583 pp.
EVERETT, L.G., 1980, Groundwater Monitoring, General Electric Company, 440 pp. Note
Section 2 ("Groundwater Monitoring Methodology") and Section 4 ("Monitoring in the
Zone of Saturation").

383-3000-001 / December 1, 2001 / Page 27
FERGUSON, COLIN, June 1992, The Statistical Basis for Spatial Sampling of Contaminated
Land, Ground Engineering, pp. 34-38.
FETTER, C.W. Jr., Fall 1981, Determination of the Direction of Ground Water Flow, Ground
Water Monitoring Review, pp. 28-31. Discusses anisotropy in groundwater flow.
GIDDINGS, TODD and SHOSKY, D.J. Jr, Spring 1987, Forum - What is an Adequate Screen
Length for Monitoring Wells? Ground Water Monitoring Review, pp. 96-103. Pros and cons
of screen lengths.
GRANT, F.S. AND WEST, G.F., 1965, Interpretation Theory in Applied Geophysics, McGraw-
Hill, New York, 583 pp.
KURTZ, DAVID and PARIZEK, R., 1986, "Complexity of Contaminant Dispersal in a Karst
Geological System," in Evaluation of Pesticides in Ground Water, American Chemical
Society, Symposium Series, vol. 315, pp. 256-281.
NIELSEN, D.M., (Editor), 1991, Practical Handbook of Ground-Water Monitoring, NWWA,
Lewis Publishers, Inc., Chelsea, Michigan 48118, 717 pp.
Note especially Chapter 2 on "Ground-Water Monitoring System Design" by Martin Sara.
PA DEPARTMENT OF ENVIRONMENTAL PROTECTION, July 1995, The Land Recycling
Program Technical Guidance Manual, 2530-BK-DEP1853.
PFANNKUCH, H.O., Winter 1982, Problems of Monitoring Network Design to Detect
Unanticipated Contamination, Ground Water Monitoring Review, pp. 67-76.
Discusses contamination release, propagation, and monitoring stages of unexpected
releases.
QUINLAN, J.F., 1990, Special problems of ground-water monitoring in karst terranes:
Ground Water and Vadose Zone Monitoring, ASTM Special Technical Paper 1053, pp.
275-304.
QUINLAN, J.F., 1989, Ground-Water Monitoring in karst terranes: Recommended protocols
and implicit assumptions: US EPA, EPA/600/X-89/050, 78 pp. [Draft: final version to be
published in 1995].
SAINES, M., Spring 1981, Errors in Interpretation of Ground Water Level Data, Ground
Water Monitoring Review, pp. 56-61. Identifies common errors.
SMART, D.L., 1984, A review of the toxicity of twelve fluorescent dyes used for water
tracing: National Speleological Society Bulletin, v. 46, no. 2, pp. 21-33.
U.S. EPA, September 1986, RCRA Ground Water Monitoring Technical Enforcement
Guidance Document.
Note Chapter 2, "Placement of Detection Monitoring Wells."
U.S. EPA and the USGS, October 1988, Application of Dye-Tracing Techniques for
Determining Solute-Transport Characteristics of Ground Water in Karst Terranes, EPA
904/6-88-001.

383-3000-001 / December 1, 2001 / Page 28
Note especially Chapter 2: "Hydrogeology of Karst Terrane."
U.S. EPA, May 1993, Subsurface Characterization and Monitoring Techniques - A Desk
Reference Guide, Volume 1: Solids and Ground Water. EPA/625/R-93/003a. Complete
description of geophysical techniques and their advantages and disadvantages is
included. Also, aquifer tests and sampling methods are presented.
WILSON, C.R., EINBERGER, C.M., JACKSON, R.L., and MERCER, R.B., 1992, Design of
Ground-Water Monitoring Networks Using the Monitoring Efficiency Model (MEMO),
Ground Water, v.30, No.6, pp. 965-970.

383-3000-001 / December 1, 2001 / Page 29

Back to top


CHAPTER 4: ANALYTE SELECTION AND MONITORING FREQUENCY
4.1
IMPORTANCE OF ANALYTE SELECTION
There are practical reasons for predetermining a set of analytes for which to monitor.
These include the purpose(s) of the monitoring effort, and the amount of resources (time
and money) available for sample collection, analysis and data interpretation or program
regulatory requirements. Unless a sampling plan is focused, significant data can become
lost in a sea of analytical results.
4.2
DEVELOPMENT OF AN ANALYTE LIST
In developing an analyte list, it is necessary to establish and define the objectives of the
monitoring activity (i.e. ambient, compliance, assessment, remediation monitoring, or
post-closure monitoring). The analytes selected should be those of concern based on
their generation (including reaction products, etc.) at the facility or from historical uses
that may be identified in an assessment of a site.
In addition, it may be beneficial to monitor other analytes beyond the obvious analytes
of concern. The general chemistry of an aquifer can be used to monitor changes and
differences in the hydrogeologic system. For example, the monitoring of basic analytes
can be used to evaluate the mobility of chemical species, correlate recharge and flow
zones with the water quality, assess the chemical equilibrium and kinetics of groundwater
reactions, and to develop analyte contour maps and graphical plots. Such evaluations
can be excellent tools to understand the flow and quality of the groundwater system.
Because of the potential for gathering valuable information at a relatively inexpensive
price, monitoring programs should include appropriate and basic chemistry species and
analytes. This is especially true of ambient and compliance monitoring at larger sites that
have the potential to be complicated in terms of contaminants, neighboring facilities,
and groundwater flow paths.
4.2.1 Ambient Monitoring
Ambient monitoring is a pre-operational or pre-remediation, relatively short term
activity that is conducted to establish groundwater quality conditions at a site
prior to any operation of a proposed activity, or to establish groundwater quality
that has not been affected by site activities. The goal is to account for both non-
site related variation (background), and any impacts from the site that may have
influenced groundwater quality. These results will form a basis against which
future monitoring results will be compared to determine permit compliance or
remediation attainment when the Act 2 Background Standard is selected or is the
controlling factor.
The ambient monitoring analytes should include all those that are being
generated by an existing activity or will be generated by a proposed activity.
Potential degradation products of known contaminants should also be
considered. In addition, parameters that may have been generated by past or
adjacent activities should also be monitored.

383-3000-001 / December 1, 2001 / Page 30
Analytes for existing or proposed activities can be obtained from permits or from
permit applications. Analytes of concern from past practices for permitted
activities can be obtained from official records including Toxics Release
Inventories (TRIs), National Pollutant Discharge Elimination System (NPDES) permits,
Bureau of Land Recycling and Waste Management permits, Agriculture
Applicators permits, and federal notification and state storage tank registration
records. Information on past practices can also be determined by interviewing
local residents, past and present employees of a facility, local officials and the
owners of the property. In addition, aerial photographs can be a valuable
source of information.
Finally, an on-site inspection can provide additional evidence of past activities
that may have impacted groundwater quality. This could be in the form of waste
disposal sites, discarded or abandoned containers, or abandoned facilities. This
on-site inspection can be a critical link between historical practices and a
sampling plan.
4.2.2 Compliance Monitoring
Analytes for compliance monitoring should include, at a minimum, all those that
are generated by the activity. In some cases, it may be necessary to include
additional analytes if they are determined to be of concern due to past
practices, potential contamination from activities in upgradient areas, or
potential degradation products. Reasons for including such additional analytes
should be documented along with actions to be taken in the event of permit
violations.
4.2.3 Assessment Monitoring
Analytes to be included in assessment monitoring include those that have been
found to be of concern through the compliance monitoring.
4.2.4 Remediation Monitoring
Analytes to be included in remediation monitoring are those that are being
remediated, indicator parameters of such, or any parameters that may indicate
physical or chemical conditions within the aquifer that could affect the
remediation processes being carried out on the site (i.e. pH, Eh, dissolved oxygen,
temperature). During active remediation, marker compounds that will monitor
the efficiency of the remediation effort and attainment of remediation goals
should be selected.
4.2.5 Post-Closure Monitoring
Analytes to be included in post-closure monitoring are those that were monitored
during compliance monitoring, as discussed above. Indicator compounds may
be used in certain cases.
4.3
DETECTION LEVELS AND METHODOLOGIES
The term detection limit is most often associated with MDL, Method Detection Limit and
PQL, Practical Quantitation Limit. MDL is defined by the EPA as the lowest concentration

383-3000-001 / December 1, 2001 / Page 31
that can be determined, with 99 percent confidence, to be greater than zero in a given
matrix. Approved EPA analytical methods and detection limits for many toxic substances
can be found in Table 2 of DEP's Water Quality Toxics Management Strategy - Statement
of Policy (Title 25, Chapter 16). A detailed description of the theory and its
implementation can be found in 40 CFR Part 136, Appendix B. By definition, MDLs do not
address the accuracy of measurements made at or near this concentration. In an effort
to provide more meaningful information to the users of analytical data, the terms PQL,
Practical Quantitation Limit, and RL, Reporting Limit, have been utilized. PQLs are
normally defined as 5 or 10 times the corresponding MDL and should represent the lowest
concentration a laboratory has confidence quantitating. An RL is defined by the Bureau
of Laboratories as the concentration in the sample that is equivalent to the lowest
standard routinely run for that analyte by a given method/instrument combination. It is
important to remember that all of these limits may be adversely affected by the sample
matrix. Sampling plans should include necessary analytical quality control measures to
determine if matrix effects are occurring. Under Act 2, PQLs are the minimum detection
level in remedial activities. A list of PQLs is published in Appendix A of the Act 2 Technical
Guidance Manual.
Methods chosen for monitoring purposes need to have undergone method validation
with regard to the proposed application. Methods published by EPA, ASTM, and
Standard Methods generally meet this standard.
The Bureau of Laboratories can provide guidance in selecting appropriate analytical
test methods if the following information is available:
a) What analytes are to be monitored?
b) Is there a requirement for the determination of non-target analytes? If so,
what types of analytes need to be included?
c)
Are there regulatory, or trigger levels that determine the needed detection
levels, and what are they?
d) Are there regulatory requirements for high accuracy?
e)
What matrices will need to be analyzed?
4.4
DURATION AND FREQUENCY OF MONITORING PERIOD
The frequency and duration of monitoring may vary within activities and specific sites. In
many cases the duration and frequency of monitoring are established by DEP regulation.
For remediation projects, the monitoring requirements are determined on a site-specific
basis by permit, DEP order, by agreement with the party responsible for the monitoring, or
by a demonstration in a Final Report that an Act 2 standard is met. Most remediation
monitoring will be conducted in accordance with the provisions of Act 2 and associated
regulations and guidances. Consideration should be given to the statistical aspects of
the sampling frequency and duration. See especially Section 5.3.3, Data Sufficiency and
Limitations.

383-3000-001 / December 1, 2001 / Page 32
4.4.1 Ambient Monitoring
The duration of pre-activity or pre-remediation ambient monitoring is determined
by the need to establish a credible database. This is especially critical if the Act
2 Background Standard has been selected for remediation. This duration may
conflict with the operator's economic needs to start the operation or activity as
soon as possible. Ideally, the duration of monitoring should be sufficient to
detect any problem or important characteristic at the site before the ambient
monitoring is stopped.
The frequency of monitoring also will be influenced by the proposed use of the
data. For example, if a valid statistical, ambient database is desired, an
adequate number of samples will need to be taken in the pre-activity monitoring
period.
In some cases, pre-activity monitoring is required by DEP regulation. In the
absence of specific regulations, at least four samples that would reflect any
seasonal influences should be collected from each sampling point prior to any
activity at the site that could affect groundwater quality.
4.4.2 Compliance Monitoring
For most compliance sites and activities, monitoring will continue as long as the
activity exists. The frequency of compliance monitoring is often established by
DEP regulation or policy. However, if specific policy or regulations do not exist,
then the frequency of monitoring may be established on a case-by-case basis.
The interval between sampling events should be short enough to allow the
operator to respond and correct a problem before any significant, widespread,
or permanent damage is done to the environment. That is, the periodic sampling
should be able to detect contamination migrating from the source before (for
example) it migrates offsite undetected.
The groundwater velocity and the contaminant characteristics are two main
factors that can affect the determination of a sampling frequency. Consider a
case where a sampling frequency must be determined for a well that is 500 feet
upgradient from the property line. The operator wants to be alerted to any
problem before it reaches the property line. If the groundwater velocity has been
determined to be 5 feet per day, then it would be necessary to sample the well
no less than once every 100 days (500 feet/5 feet/day), minus the number of days
required for analysis.
Sampling at this frequency would alert the owner to contamination before it
crossed the property line. If the owner wanted to allow enough time for some
pre-planned actions, then the interval would have to be an appropriate number
of days less than the calculated sampling frequency.
The vulnerability of a site to pulses of contaminant migration during rain events
may affect the desired frequency of sampling. This may occur in deep mined
areas and karst conditions. In such cases, the chosen sampling frequency may
be irregular and keyed to significant rainfall events rather than regular intervals.

383-3000-001 / December 1, 2001 / Page 33
Characteristics of the contaminant also can influence the duration and
frequency of sampling. For example, the toxicity or health risk of the site's
potential contaminants such as carcinogens may be much greater than other
substances, and may require more intensive sampling.
Specific and tangible reasons must exist in order to change established sampling
frequencies based on groundwater flow and contaminant transport. This may be
very difficult to determine and the subsequent sampling frequency should be
conservative. However, tangible evidence could include tracer studies and
pumping test data which indicate that sampling should occur on a more or less
frequent basis.
Also, an indicator analyte might be sampled more often than other analytes (see
Section 4.2), or certain wells might be sampled more often. For example, shallow
wells may require more attention where the potential contaminants of a
permitted facility are less dense than water.
4.4.3 Assessment Monitoring
Assessment monitoring may be required based on sampling data obtained from
one or more sampling points during compliance or post-closure monitoring.
Assessment monitoring is conducted as a response to either 1) the indication of
contamination, or 2) exceeding an action or "trigger" level. Assessment
monitoring typically involves resampling of wells to confirm contamination and/or
performing an assessment of the extent of contamination.
Resampling should be accomplished as soon as possible after the initial analyses
are available. In some programs, the resampling is required within 10 days of
completion of the initial analyses that triggered the resampling obligation.
If resampling confirms the initial indication of contamination or exceeding of a
permit level, an assessment plan should be implemented. The assessment
investigation should be completed as soon as possible so that a thorough
remedial action plan can be quickly designed and implemented. New wells that
are installed for the assessment purposes are usually sampled two or three times
at a frequency of once every two to four weeks.
4.4.4 Remediation Monitoring
The duration and frequency of monitoring during remediation should be
designed with two major purposes in mind. Monitoring should be able to
demonstrate whether the groundwater contamination is spreading or is
adequately contained; and it should be able to objectively document the
degree of effectiveness of any remediation effort to achieve one of the three
alternatives of Act 2 at the point of compliance.
Remediation monitoring may be more frequent in the early stages of a project
than in later stages. This is because understanding and predictability increase as
experience with the project accumulates. For example, as a remediation project
proceeds, the contaminant concentrations may approach steady levels. Once
an asymptotic or steady state condition is verified, the contaminant

383-3000-001 / December 1, 2001 / Page 34
concentrations remaining should be compared to the three remediation
standards found in Act 2. Remediation systems may be stopped at this point.
Monitoring should then continue for some period to assure that contaminants do
not return in exceedance of the Background or Statewide Health standard levels
specified in Act 2. Seasonal variations in precipitation and the water table levels
may affect the migration rates and paths of residual contaminants. A rising water
table can collect contamination residue in the unsaturated zone and produce
higher concentrations in groundwater during subsequent sampling.
4.4.5 Post-Closure Monitoring
In most cases, post-closure monitoring requirements are established by
department regulation.
In general, a potential source of groundwater contamination should be
monitored in accordance with post-closure case requirements or until cleanup
standards have been met and liability has been released under Act 2.
4.4.6 Cessation of Monitoring
Monitoring should continue as long as there is a clear and practical reason for
continuing it. However, monitoring can be conducted at a reduced frequency,
or possibly discontinued when any of the following occur:
1.
When allowed by regulations, permit, administrative order, or other
agreement with DEP.
2.
When the goals of a monitoring program have been achieved, such as
when cleanup standards have been met.
4.5
REFERENCE
BARCELONA, M.J., WEHRMANN, H.A., SCHOCK, M.R., SIEVERS, M.E., and KARNY, J.R.,
September 1989, Sampling Frequency for Ground-Water Quality Monitoring, EPA Project
Summary, EPA/600/S489/032.

383-3000-001 / December 1, 2001 / Page 35
CHAPTER 5: STATISTICAL ANALYSIS OF MONITORING DATA
5.1
PURPOSE
The purpose of this chapter is to provide guidance for the use of statistics in assessing
groundwater quality. It is not intended to address statistical procedures in detail, but
rather to provide a framework and define the key concepts, terms, and references for
proper statistical analysis. The references cited and standard texts should be used to
perform the procedures as appropriate, or if necessary, professional services should be
obtained. The two EPA publications (Statistical Analysis of Ground-Water Monitoring Data
at RCRA Facilities - Interim Final Guidance, 1989; and Addendum to Interim Final
Guidance, 1992) should serve as default guidances where this chapter or existing
regulations do not address an issue or procedure. See Section 5.5 for information on
obtaining these documents.
In some cases, regulations such as the Hazardous Waste Regulations for treatment,
storage, and disposal, require statistical analyses. These are summarized below:
Chapter
Topic
264.96
Record keeping and reporting
264.97
General groundwater monitoring requirements
264.98
Detection monitoring program
264.99
Compliance monitoring program
264.278
Unsaturated zone monitoring
264.280
Closure and post-closure care
264
Appendix C
265.91
Groundwater monitoring system
265.93
Preparation, evaluation and responses
On occasion, the recommendations here will contrast with the Bureau of Land Recycling
and Waste Management's regulations dealing with statistics. It should be noted that the
regulations do allow for alternate approaches when appropriate (Section 264.98 (c)(2)).
Also, the resampling approach discussed in this manual (Section 5.4.3.5) may be useful
when determining "whether the facility has caused the [statistically significant] change"
(Section 264.98 (e)). Where possible, these procedures should be used because they
represent updated methods and approaches. It is anticipated that additional guidance
on statistical issues will be provided in the future. Specific rulemaking associated with
legislation also may provide guidance on statistical methods used in achieving cleanup
standards.
Some of the statistical procedures recommended in this chapter will be easier to
implement with computer software. The program should be compatible with the
database, and should be able to handle data transformations, and both parametric and
nonparametric procedures.
Often, however, adequate statistical analysis with small databases can be accomplished
through exploratory data analysis using a calculator and time series graphs. The depth of
statistical analysis will typically depend on the amount and condition of the data, and the
ultimate purpose and goals for monitoring.

383-3000-001 / December 1, 2001 / Page 36
5.2
STATISTICAL APPROACHES
Both graphical and mathematical methods are used to analyze groundwater monitoring
data. Graphical procedures summarize data concisely and aid interpretation of data
but are not used to determine confidence levels or probabilities. Confidence levels and
probabilities are determined by applying mathematical methods such as parametric
and nonparametric statistical procedures.
The first step in the statistical assessment of data should be exploratory data analysis
(EDA). This includes the use of graphical techniques and calculation of summary
statistics. Graphical methods include probability plots, box charts, and time-series plots to
visually review the data for trends or drifts. EPA and most statistical texts recommend that
time-series data should be graphed. This visual approach allows for a quick assessment
of the statistical features of the data. Calculation of summary statistics are typically done
to characterize the data and make judgments on the central tendencies, symmetry,
presence of outliers, etc. EDA is critical to selecting additional appropriate
mathematical procedures.
For mathematical methods of analysis, both nonparametric and parametric procedures
can be used depending on the distribution of the data, the percentage of nondetects,
and the database size. However, both procedures have assumptions that must be met to
be considered valid analyses. Note also that "from a statistical point of view, all water
quality constituents are considered random variables" (Harris and others, 1987).
5.2.1 Parametric Procedure Assumptions
Assumptions of parametric procedures include a specific data distribution such as
normal (also know as Gaussian or the bell-shaped curve) or lognormal (normality
achieved by logtransforming the data), and data variances that are similar. In
addition, the data are assumed to be independent (see Section 5.3.3.1).
5.2.2 Nonparametric Procedure Assumptions
Assumptions for nonparametric tests also are important. Nonparametric
procedures assume equal variances and that the type (shape) of distribution of
the population is the same at each well. This is different from assuming a normal
distribution when using parametric statistics. In other words, nonparametric
methods do not require a specific type of data distribution.
Nonparametric procedures may be preferred because they:
1.
are free from normal distribution assumptions thereby eliminating the need
for normality tests and data transformations;
2.
are resistant to effects of outliers; and
3.
are usable when censored (i.e. less-than detection values) data are present.
5.3
DATA AND SAMPLING CONSIDERATIONS
Three of the primary purposes of statistics are to 1) estimate the characteristics of the
data, 2) compare datasets, and 3) quantify the uncertainty of data. These purposes
include such activities as measuring the central tendency of data, comparing data from
upgradient and downgradient wells or with established standards, and quantifying the
probability of statistical conclusions.

383-3000-001 / December 1, 2001 / Page 37
The following topics typically should be considered when using statistics (especially when
assessing groundwater data for permitted activities):
?
data variability
?
significance levels
?
data sufficiency and limits
?
monitoring network design
5.3.1 Data Variability
All groundwater analyte measurements contain a dimension of random variability
caused by a variety of factors. Effects of heterogeneous geology, groundwater
flow, in situ processes, and errors in sampling and analysis all contribute to this
variability.
5.3.1.1
Spatial Variability
Spatial variability is a significant factor when comparing upgradient
groundwater quality with data from compliance wells. EPA
recommends considering the inclusion of wells that are not necessarily
physically upgradient to understand the spatial variability of the
groundwater quality. This is because testing a number of downgradient
wells against one upgradient well does not consider the upgradient
spatial variability. Statistical tests under these conditions often will
produce "false positives."
The accompanying effect of using more upgradient wells is that there is
a better chance to produce "false negatives." These two concerns must
be offset by considering the level of significance of the statistical tests
(Section 5.3.2) and the statistical procedures (Section 5.4). See also
Network Design (Section 5.3.4).
However, it must be emphasized that if the variability of the upgradient
data is not considered through multiple upgradient wells, it may
invalidate the statistical procedures that follow or lead to erroneous
judgments regarding trends of the data. It is to the advantage of both
the regulated and the regulator to adequately characterize the data
variability of the upgradient groundwater quality.

383-3000-001 / December 1, 2001 / Page 38
5.3.1.2
Temporal Variability
Temporal variability is variability that occurs at a fixed point with the
passage of time. For example, a time series plot for a monitoring well
that is sampled through time will typically show varying concentrations.
Samples that are taken close together may show serial correlation,
which is also called autocorrelation (see Section 5.3.3.1).
5.3.1.3
Seasonality
Seasonality, cyclical behavior in water quality due to annual seasons,
can contribute to data variability. Davis and McNichols (1994, Part I)
state that seasonality is "rarely a concern in analyzing groundwater
data... in our experience." However, analytes in "shallow, highly
permeable aquifers" and those affected by fluctuating water levels
may show seasonal variability (see Montgomery, Loftis, and Harris, 1987).
Major factors that may cause seasonality include land use practices
(especially agricultural) and the rise and fall of the water table.
Changing elevations of the water table can significantly affect the
concentrations of contaminants such as those associated with
petroleum hydrocarbons.
The patterns of seasonality can be highly variable because of
complicating factors such as variable recharge and permeability
(Pettyjohn, 1982). Activities at a site also may override any evidence of
seasonality. If seasonality patterns are detected that are on the order
of several years, the impact on quarterly sampling will tend to be
minimal.
The EPA cautions against using corrections for seasonality, noting that
they "represent extrapolation into the future." Recommendations for
assessing seasonality in upgradient wells are presented in Gilbert (1987,
Chapter 17) and in Montgomery, Loftis, and Harris (1987). Graphical
methods should be used for the initial detection of seasonality. See also
Dataset Size in Section 5.3.3.3.
5.3.2 Significance Levels
A statistical test is typically performed under a confidence level such as 90
percent, 95 percent, or 99 percent. The confidence level is the chosen
probability of accepting the hypothesis (of no contamination). The confidence
level has a corresponding significance level. The significance level of a test is
called alpha (?), and corresponding values to the percentages above would be
.1, .05, or .01.
The significance level is equal to the false positive rate. This is the rate at which
the test indicates contamination when there is none. (This is a Type I error: the
hypothesis is incorrectly rejected. Our hypothesis is that there is no
contamination.) For example, if ? is .05, 1 out of 20 (5 percent) statistical tests
would generate a false positive (an incorrect rejection of the hypothesis).

383-3000-001 / December 1, 2001 / Page 39
When a confidence level of a test is raised (e.g. from 95 percent to 99 percent),
then the ? value is lowered (.05 to .01) and the following occurs:
?? the test is less likely to generate a false positive (now only 1 out of 100), but
?? the test is more likely to miss detecting the contamination, because the
hypothesis is harder to reject.
The overall chance of having a false positive over an entire site is cumulative - it
increases with the number of statistical tests performed. This effect is called the
Facility Wide False Positive Rate (FWFPR).
The formula for determining the FWFPR is:
1 - (1- ?.)
# of tests
where the number of tests is the number of wells times the number of parameters.
For example, at ? = 0.01, and 20 wells with 15 parameters:
FWFPR = [1 - (1 - 0.01)
300
] * 100 percent
FWFPR = 95.1 percent
This indicates that the chance for generating a false positive at the site is over 95
percent. This high percentage indicates that statistical tests for a facility are likely
to indicate an instance of contamination when there really is no contamination.
The facility would be subjected to all the actions and expenses that a statistical
test might trigger.
Section 5.1 of the EPA Addendum (1992) discusses how to lower the FWFPR,
without compromising the test's power. The power of the test is its ability to detect
contamination when it exists. Section 5.1 also discusses how to graphically assess
the power of a test.
For any individual comparison at a permitted facility, the recommended
significance level is at a minimum .01. Large monitoring networks are judged
differently because a balance must be sought between the FWFPR and the
power of the test. For such sites, the FWFPR should be approximately 5 percent
while a certain level of statistical power is held (Section 5.1, EPA Addendum,
1992).
These recommendations should be considered when reviewing or setting
significance levels for statistical tests at permitted facilities. Davis and McNichols
(1994, Parts I and II) offer suggestions to improve the implementation of the
recommendations. EPA expects to issue further guidance in the future.

383-3000-001 / December 1, 2001 / Page 40
5.3.3 Data Sufficiency and Limitations
5.3.3.1
Independence
All of the statistical methods discussed in this section assume sample
independence. A groundwater sample must have a sufficient
sampling interval to ensure independence. That is, the well must be
allowed to equilibrate with the surrounding aquifer and enough time
be permitted to pass for changes to occur in the aquifer before the
sampling of a distinct volume of water can occur. For example, a
second groundwater sample taken one week later may be highly
correlated (serially) to the first. This can be understood by
considering the calculation of the average temperature of the day.
Two thermometer readings five minutes apart may give little
information on the mean and variance of the daily temperature.
Duplicate samples should not be treated as independent
groundwater samples. Independence also can be affected by field
instruments and analytical laboratories. See the EPA Interim Final
Guidance (1989, Section 3 - Choosing a Sampling Interval). This
reference shows how to calculate the minimum acceptable
sampling interval.
Generally, quarterly sampling programs can continue as normal;
however, sample independence should be considered if a greater
sampling frequency is required, or groundwater flow at a site is very
slow. Moderate to fast groundwater flow typically ensures the
independence of a sample. See Davis and McNichols (1994, Part I).
5.3.3.2
Transformations and Distribution
The EPA Addendum (1992, Section 1) recommends methods for
addressing the assumptions of statistical tests. The following is a
summary of the approach recommended by EPA.
For data transformations, it is recommended that all data be logged,
because the lognormal distribution appears to be most appropriate
as the default statistical model. The logged data are then checked
for normality. If the test for normality is not rejected, further testing is
done with the logged data and not the original data.
Note that when the dataset is smaller than 20 - 30 observations, all
normality tests do "at best a fair job of rejecting non-normal data." As
more data become available, normality assumptions should be
revisited.
Two main methods are recommended for checking normality.
Referencing is done by using site historical data or data from similar
hydrogeologic settings to assume what the distribution of the data is.
As more data become available, the assumptions would be verified
by standard tests. Probability plotting is the only test recommended
by the EPA without any qualifiers. Probability plots are used to look for
irregularities in the data. Normally distributed data will plot as a
straight line.

383-3000-001 / December 1, 2001 / Page 41
Other tests for data normality such as the Coefficient-of-Variation test
and the Chi-squared test may give tenuous results and should be
backed up by an additional test. See Section 1 of the EPA
Addendum (1992) for specific procedures and use of the tests, and
for limitations on traditional normality tests such as Coefficient-of-
Variation. See also the EPA Interim Final Guidance (1989, Section 4.2
- Checking Distributional Assumptions).
Unequal variances between groups of data (for example,
upgradient wells and compliance wells) can invalidate a statistical
test; this is particularly true for the parametric analysis of variance
(ANOVA). See the EPA Interim Final Guidance (1989, Section 4.3 -
Checking Equality of Variance: Bartlett's Test) and the EPA
Addendum (1992, Section 1.2 - Testing for Homogeneity of Variance).
Davis and McNichols (1994, Part I) argue that nonparametric ANOVA
tests also will be invalidated by unequal variances. The EPA
Addendum (1992) also reviews the use of box plots for assessing the
homogeneity of variance, and the use of Levene's test as an
additional check.
5.3.3.3
Dataset Size
In most cases, the necessary dataset size (i.e. the number of samples
needed per well) will be a function of the purpose of the test, the
power of the test needed to detect the minimum contamination,
and the type of test.
Minimum sample sizes of groups for nonparametric ANOVA tests must
be a little larger than those for parametric ANOVA tests. For the
Wilcoxen Rank Sum test, the EPA recommends having at least four
samples for both groups. For the Kruskal-Wallis procedure, EPA gives a
rule of thumb that there should be "a minimum of three well groups
with at least four observations per group" (EPA Addendum, 1992,
Sections 3.1 and 3.2).
Possible scenarios for choosing the number of upgradient samples
and the number of downgradient samples for two-phase (retesting)
strategies are contained in Section 5.2 of the EPA Addendum (1992).
The power of a statistical test can be increased by increasing the
number of samples. The power should be comparable to the EPA
reference power curve (see EPA Addendum, 1992 - Section 5.2 and
Appendix B). The EPA reference power curve does not depend on
the number of wells in the network, but it does depend on the
number of upgradient samples that are used to construct the upper
prediction limit.
The exception to the benefits of increasing the number of samples is
in long-term trend tests (greater than five years) where a greater
number of data points (i.e. less time between sampling) may
increase problems with serial correlation.
Procedures for seasonality tests (the Seasonal Kendall test) will
generally require continuous monthly samples over years (at least 3)

383-3000-001 / December 1, 2001 / Page 42
for adequate characterization of seasonal variation (see Gilbert,
1987, Chapter 17). Seasonality tests for quarterly sampling programs
are not recommended.
5.3.3.4
Outliers
Outlier values often result from identifiable analytical or transcription
errors. These need to be corrected through careful review of the
data. True outlier values need no special treatment and should not
be deleted arbitrarily. This is especially true with the lognormal nature
of groundwater data, which by definition may contain a few very
large values. Resampling can be used as a tool to judge outliers. In
addition, good quality assurance/quality control procedures will
include data validation (see Section 8.3, item Q).
The EPA Addendum (1992, Section 6.2) lists a test procedure for
normal or lognormal data that can be used, typically for outliers that
are orders of magnitude above the rest of the data.
5.3.3.5
Censored Data
Censored data refers to less-than detection or less-than reporting
values. The EPA Addendum (1992, Section 2) details methods for
handling nondetects. The approach depends on the percentage of
nondetects (see Table 1).
5.3.4 Network Design
Monitoring network design typically results from prior knowledge or assumptions
regarding site location, hydrology, water quality, and economics. Lacking prior
information, Gilbert (1987) presents a thorough review of methods available to
develop a statistically sound network.
Wilson and others (1992) present an additional method that attempts to quantify
monitoring efficiency. A Monitoring Efficiency Model (MEMO) is used to calculate
the area of detection compared to the total area of the site. The EPA Interim
Final Guidance (1989, Appendix A) recommends that wells are sited so that "at
least one of the wells should intercept a plume of contamination of reasonable
size."

383-3000-001 / December 1, 2001 / Page 43
Table 1. Procedures for handling nondetects (from the EPA Addendum, 1992).
Percentage of
Nondetects
Method
Less than 15
percent
Replace nondetects by one-half the detection limit and
proceed with parametric ANOVA or interval tests.
Over 15 percent
For comparing upgradient with downgradient wells, use a
nonparametric test. For all two-group comparisons, use the
Wilcoxon Rank-Sum test. For more than two, use Kruskal-
Wallis test.
When using statistical interval tests (Confidence, Tolerance,
or Prediction limits), use parametric approach along with
Cohen’s Procedure or Aitchisons’s test to adjust the mean
and standard deviation. The decision for Cohen’s or
Aitchison’s test should be made based on the use of
probability plots. See the EPA Addendum (1992) for more
details.
Over 50 percent
With nondetects >50 percent or if tests cannot be justified,
use nonparametric interval or nonparametric ANOVA tests
(EPA Addendum, 1992, Chapter 4).
Over 90 percent
Use Poisson distribution model (EPA Addendum, 1992,
Sections 2.2.3 and 2.2.4).
See also Chapter 3 of the EPA Interim Final Guidance (1989, Location and Depths
of Monitoring Wells). As discussed under Data Variability (Section 5.3.1),
adequately characterizing the upgradient groundwater quality is very important.
Nielsen (1991) recommends that multiple (at least four) upgradient wells be
constructed at a site. For small sites or those with roughly homogeneous geologic
conditions, two upgradient wells could be sufficient to account for spatial
variability.
5.4
STATISTICAL PROCEDURES
The specific statistical procedures used depend on the goals and quality of the
monitoring data. A general flow chart for statistical analysis is shown in Figure 9. The
methods selected should be consistent with the goals of the monitoring. For example,
monitoring data analysis at a permitted site may include trend analysis, well-to-standards
and well-to-well comparisons, or the use of intervals. Monitoring data collected to
characterize an aquifer may consist mainly of graphical displays and summary statistics.

383-3000-001 / December 1, 2001 / Page 44
Figure 9. General guidance on selection of statistical procedures.
GROUNDWATER
MONITORING
DATA
WILCOXEN
RANK-SUM
BETWEEN
2 GROUPS
KRUSKAL
WALLIS
MORE THAN
2 GROUPS
NONPARAMETRIC
ANOVA
PARAMETRIC
ANOVA
ASSESS NONDETECTS
See Table 1
WELL TO WELL
COMPARISON
Prediction intervals can be
combined with retesting
(see EPA Addendum,
1992, Section 5.2.3)
CONFIDENCE INTERVAL
(Not typically used -
see EPA Addendum,
1992, Section 4.3)
PREDICTION
INTERVAL
TOLERANCE
INTERVAL
NONPARAMETRIC
TESTS
An alternative to
ANOVA: combine tests
and use as a two-phase
testing strategy
CONFIDENCE INTERVAL
(Not typically used -
see EPA Addendum,
1992, Section 4.3)
PREDICTION
INTERVAL
TOLERANCE
INTERVAL
PARAMETRIC
TESTS
ASSESS NONDETECTS
See Table 1
INTERVAL
TESTS
ADDITIONAL
ANALYSIS
AND/OR
TREND TESTS
(graphical, regression,
Mann Kendall)
EXPLORATORY
DATA ANALYSIS
(box plots, time series plots,)
summary statistics, and other tests

383-3000-001 / December 1, 2001 / Page 45
Graphical and parametric statistical procedures discussed here are included in many
introductory statistics textbooks (e.g. Iman and Conover, 1983 and Ott, 1988) and are
available in many computer statistics packages.
The EPA Interim Final Guidance (1989) and EPA Addendum (1992) describe and provide
examples for both the parametric and nonparametric methods. See additional
discussions in Helsel and Hirsch (1992), Conover (1980), Gilbert (1987), and Davis and
McNichols (1994, Parts I and II).
5.4.1 Graphical Procedures
Refer to Cleveland (1993) for a general reference on graphical procedures. The
use of boxplots is described in the EPA Addendum (1992).
5.4.1.1
Boxplots
A boxplot summarizes a data set by presenting the percentile
distribution of the data. The “box” portion indicates the median and
interquartile range (IQR). IQR is the middle 50 percent of data.
Difference in the size of box halves represents data skewness.
Normal and symmetrical distributions will have equal size box halves.
Extreme outliers are displayed as individual points that are recognized
easily. Boxplots can be constructed by hand; however, many
computer statistical packages will do them.
The boxplot of a lognormal distribution will have noticeably different-
sized box halves. Lack of IQR overlap for different data sets will indicate
a probable significant difference. Boxplots of seasonally grouped data
can be used to detect data seasonality.
5.4.1.2
Time Series Plots
A time series plot displays individual data points on a time scale. A
monthly scale can help to identify seasonal variation. A yearly scale
also can identify possible trends. Superimposing data from multiple
sampling locations may provide additional information. Improved trend
information is often available with data smoothing. One smoothing
procedure showing movement of the ‘center’ of data over time is
LOWESS. This procedure is most helpful with data having substantial
variability and a long period of record. LOWESS requires computer
software.
5.4.1.3
Control Charts
Control charts are used to define limits for an analyte that has been
monitored at an uncontaminated well over time. This procedure is a
graphical alternative to prediction limits.
A common technique is the Shewhart-CUSUM control chart that plots
the data on a time scale. Obvious features such as trends or sudden
changes in concentration levels could then be observed. With this
method, if any compliance well has a value or a sequence of values

383-3000-001 / December 1, 2001 / Page 46
that lie outside the control limits for that analyte, it may indicate
statistically significant evidence of contamination.
The control chart approach is recommended only for uncontaminated
wells, a normal or lognormal data distribution with few nondetects, and
for a dataset that has at least eight independent samples over a one-
year period. This baseline is then used to judge the future samples. See
the EPA Interim Final Guidance (1989, Section 7) and the EPA
Addendum (1992, Section 6.1) for procedures.
5.4.2 Summary Statistics
Basic summary statistics can be used to characterize groundwater monitoring
data. Summary statistics include median, interquartile range (IQR), mean,
standard deviation, and range. Median and IQR are determined from
percentiles. Median is 50th percentile and IQR is 25th to 75th percentile. Median
indicates “center” of data values. The mean is another measure of center but
only if data are normally or symmetrically distributed. However, most water
quality data are not. Mean and standard deviation are required values with
parametric procedures. Range is the minimum to maximum values. Procedures
for such summary statistics are found in introductory statistics texts.
5.4.3 Interval Tests
5.4.3.1
Statistical Intervals
Statistical interval tests can be used in combination with well to well
comparisons, or independently. Statistical intervals include three main
types: tolerance intervals, prediction intervals, and confidence intervals.
Which ones are used depend on the goals of the data analysis (see
Section 4 of the EPA Addendum, 1992 for procedures).
5.4.3.2
Tolerance Intervals
Tolerance intervals will typically be the most useful interval test. They
are used to determine the extent of data that is specified to be within a
standard or ambient level. For example, the tolerance limit with a
population coverage of 95 percent and a 95 percent confidence level
calculated from compliance data should not exceed a standard (like
an MCL). This tolerance limit will ensure that at least 95 percent of the
population values will not exceed a standard with a 95 percent
confidence level.

383-3000-001 / December 1, 2001 / Page 47
5.4.3.3
Prediction Intervals
Prediction intervals are used to determine if the next one or more
samples are within the existing data distribution at a certain confidence
level. The prediction interval contains 100 * (1- ? value) percent of the
distribution. A smaller value will include a larger range of data.
Prediction intervals are used for intrawell (single well) comparisons, and
for comparison of a compliance well with an upgradient well.
5.4.3.4
Confidence Intervals
Confidence intervals contain a specified parameter of the distribution
(such as the mean of the data) at a specified confidence level.
Confidence intervals do not address extreme values.
5.4.3.5
Two-Phase Retesting Strategies
Once a tolerance interval is established for upgradient data, data from
downgradient compliance wells can be compared to the upper limit of
the interval. Such an interval can be combined with a prediction
interval to use with the next sample, or it can be combined with
resampling. A resampling strategy is used when an analyte exceeds
the upper tolerance level. The well is retested for the parameter of
concern and the value is compared to the upper limit of a prediction
interval. These two-phase testing strategies can be very effective tools
for controlling the Facility Wide False Positive Rate while maintaining a
high power of detecting contamination.
See Sections 5.2.2 and 5.2.3 of the EPA Addendum (1992) which
describe the procedures to use along with recommended coverage
and confidence levels.
5.4.4 Well-to-Well Comparison Tests
The following tests are outlined in the EPA Interim Final Guidance (1989) and the
EPA Addendum (1992). These are the recommended tests for analysis of
groundwater data between upgradient and downgradient well groups,
downgradient wells and a health-based standard, or of intrawell (single well)
comparisons. This does not include all potentially satisfactory statistical tests, but
are the preferred tests.
5.4.4.1
Analysis of Variance (ANOVA)
ANOVA includes a group of procedures used for comparing the means
of multiple (three or more) independent groups such as upgradient
wells and downgradient wells. The ANOVA methods are used to
determine if there is statistically significant evidence of contamination
at downgradient wells compared to an upgradient well, or groups of
wells.

383-3000-001 / December 1, 2001 / Page 48
The one-way ANOVA method is described with examples in Section 5.2
of the EPA Interim Final Guidance (1989). This is the EPA recommended
procedure for comparing data that do not violate the assumptions of
normal distribution and approximately equal variances.
However, as the number of wells (or groups) increases at a site, the
power of ANOVA to detect individual instances of contamination
decreases. For this reason, tolerance and prediction intervals with
retesting provisions are often much better procedures to use.
5.4.4.2
Kruskal-Wallis Test
If assumptions of the one-way ANOVA test are "grossly" violated, the
nonparametric Kruskal-Wallis test is used for more than two
independent groups of data. It can be used for comparison of
upgradient water quality to water quality from many downgradient
wells in one procedure. Alternatively, if the wells are grouped by some
characteristic (e.g. depth, geology, location, season), comparisons
among other groups can be made.
If the null hypothesis (no change) is rejected by Kruskal-Wallis (i.e. the
test statistic exceeds the tabulated critical value), then pairwise
comparisons should be made to determine what wells are
contaminated (see Gilbert, 1987, Section 18.2.2; the EPA Addendum,
1992, Section 3.1; and the EPA Interim Final Guidance, 1989, Section
5.2.2).
5.4.4.3
Wilcoxon Rank Sum
This procedure (also known as Mann-Whitney) is a nonparametric test
for differences between two independent groups such as upgradient
and downgradient water quality samples, baseline and compliance
samples, and assessment and remediation samples. See Section 3.2 of
the EPA Addendum (1992).
5.4.4.4
t-test
The t-test is a parametric, ANOVA type of test used to assess differences
in means of two independent groups. This test assumes normal
distributions and equal variances for both groups. The t-test is best
limited to situations where the data sets are too small to use
nonparametric procedures. For example, if upgradient groundwater
quality is limited to two or three samples, the t-test can be used to test
for differences between upgradient and compliance data.
5.4.5 Trend Tests
5.4.5.1
Considerations
When monitoring data have been collected over several years or more,
trend tests allow the determination of the change in distribution of data
over time. In addition to water quality trends, a time series of
monitoring data may contain characteristics of seasonality and serial
correlation. Other complicating factors include changes in laboratories
or procedures involving the sampling and analysis of the analyte.

383-3000-001 / December 1, 2001 / Page 49
Seasonality and serial correlation interfere with trend tests either by
reducing the power to detect trends or giving erroneous probabilities.
Correction for seasonality is available for tests presented here. Serial
correlation exists if a data point value is at least partially dependent on
nearby data point values. For a given data set, serial correlation
decreases with increasing temporal distance between samples. Harris
and others (1987) reported difficulty detecting serial correlation in 10
years or less of quarterly groundwater data. Therefore, correction is not
recommended for quarterly data. Serial correlation correction is
available for Seasonal Kendall trend test (Hirsch and Slack, 1984), but
has reduced power with small data sets and not recommended for a
monthly time series that is less than five years.
5.4.5.2
Parametric Trend Tests
Parametric trend tests are based on regression methods and allow
compensation for exogenous effects (outside influences). Regression
analysis between two variables can be used to calculate the
correlation coefficient (r). The closer r is to one, the closer the
relationship is between the two variables. A t-test of correlation can be
done on r to see if it is significant (see Davis, 1987, Chapter 2).
Mixed (i.e. parametric and nonparametric methods) methods also are
available when removing the effects of exogenous variables. Helsel
and Hirsch (1992) present a thorough review of trend analysis. Methods
for detecting trends also are presented in Chapter 16 of Gilbert (1987).
Because regression techniques are based on the assumption of a
normal distribution of the data, a nonparametric approach may have
to be used.
5.4.5.3
Nonparametric Trend Tests
The Mann-Kendall trend test is a nonparametric test for monotonic
(steadily upward or downward) trend. This method is used on a dataset
of less than 40 in number, but can be used for as few samples as ten if
the dataset does not contain many “ties.” For greater than 40 samples,
the normal approximation test is used (Gilbert, 1987).
This test requires constant variance in data. Non-constant variance
may be changed to constant variance with a power transformation.
Logarithm transformation is usually most appropriate. This
transformation does not affect the test statistic. Decision rules, exact
test tables, normal approximation formulas, and correction for ties can
be found in Helsel and Hirsch (1992); Gilbert (1987) and many
introductory statistics texts. When a trend is present, the slope of fitted
line can be estimated using Sen's estimator (see Gilbert, 1987).
The Seasonal Kendall trend test is a seasonally corrected Mann-Kendall
trend test. This should be applied when time series graphs or boxplots of
data indicate the presence of seasonal variation. See Chapter 17 of
Gilbert (1987).

383-3000-001 / December 1, 2001 / Page 50
5.5
REFERENCES
BARCELONA, M.J., WEHRMANN, H.A., SCHOCK, M.R., SIEVERS, M.E., and KARNY, J.R.,
September 1989, Sampling Frequency for GroundWater Quality Monitoring, EPA Project
Summary, EPA/600/S489/032.
CLEVELAND, W.S., 1993, Visualization Data, Hobart Press, Summit, N.J., 360 pp.
CONOVER, N.J., 1980, Practical Nonparametric Statistics, 2nd ed., John Wiley and Sons,
New York, 493 pp.
CRAWFORD, C.G., SLACK, J.R., and HIRSCH, R.M., 1983, Nonparametric Tests for Trends in
Water-quality Data Using the Statistical Analysis System, U.S. Geological Survey Open-File
Report 83550, 102 pp.
DAVIS, C.B., and MCNICHOLS, R.J., 1994, Ground Water Monitoring Statistics Update: Part
I: Progress Since 1988, in Ground Water Monitoring and Remediation, pp. 148-158.
DAVIS, C.B., and MCNICHOLS, R.J., 1994, Ground Water Monitoring Statistics Update: Part
II: Nonparametric Prediction Limits, in Ground Water Monitoring and Remediation, pp.
159-175.
DAVIS, J.C., 1986, Statistics and Data Analysis in Geology, 2nd ed., Kansas Geological
Survey, John Wiley & Sons, New York, 646 pp.
GIBBONS, R.D., 1994, Statistical Methods for Ground-Water Monitoring, John Wiley and
Sons, New York.
GILBERT, R.O., 1987, Statistical Methods for Environmental Pollution Monitoring, Van
Nostrand Reinhold, New York, 320 pp.
HARRIS, J., LOFTIS, J.C., and MONTGOMERY, R.H., 1987, Statistical Methods for
Characterizing Ground Water Quality, Ground Water, v.25, no.2, pp. 185-193.
HELSEL, D.R., and HIRSCH, R.M., 1992, Statistical Methods in Water Resources, Elsevier, New
York, 522 pp.
HIRSCH, R.M., and SLACK, J.R., 1984, A Nonparametric Trend Test for Seasonal Data with
Swerial Dependence, in Water Resources Research, v.20, no.6, pp. 727-735.
IMAN, R.L., and CONOVER, W.J., 1983, A Modern Approach to Statistics, John Wiley and
Sons, New York, 497 pp.
MONTGOMERY, R.H., LOFTIS, J.C., and HARRIS, J., 1987, Statistical Characteristics of
Ground-Water Quality Variables, Ground Water, v.25, no.2.
NIELSEN, D.M., (Editor), 1991, Practical Handbook of Ground-Water Monitoring, NWWA,
Lewis Publishers, Inc., Chelsea, Michigan 48118, 717 pp.
Note especially Chapter 13 on "Organization and Analysis of Water Quality Data" by
Martin Sara and Robert Gibbons.
OTT, L., 1988, An Introduction to Statistical Methods and Data Analysis, 3rd ed., PWS-KENT
Publishing, Boston, 835 pp.

383-3000-001 / December 1, 2001 / Page 51
PETTYJOHN, W.A., 1982, Cause and effect of Cyclic Changes in Ground Water Quality,
Ground Water Monitoring Review, pp. 43-49.
U.S. ENVIRONMENTAL PROTECTION AGENCY, April 1989, Statistical Analysis of Ground-
Water Monitoring Data at RCRA Facilities: Interim Final Guidance, Office of Solid Waste.
U.S. ENVIRONMENTAL PROTECTION AGENCY, 1992, Statistical Analysis of Ground-Water
Monitoring Data at RCRA Facilities: Addendum to Interim Final Guidance, Office of Solid
Waste, EPA/530-R-93003.
Draft. Currently distributed as part of Statistical Training Course for Ground-Water
Monitoring Data Analysis. Available through the RCRA Docket, (202) 260-9327.
WILSON, C.R., EINBERGER, C.M., JACKSON, R.L., and MERCER, R.B., 1992, Design of
Ground-Water Monitoring Networks Using the Monitoring Efficiency Model (MEMO),
Ground Water, v.30, No.6, pp. 965-970.

383-3000-001 / December 1, 2001 / Page 52

Back to top


CHAPTER 6: GROUNDWATER SAMPLING TECHNIQUES
6.1
IMPORTANCE OF SAMPLING TECHNIQUE
Proper sampling procedures that will allow a representative measure of groundwater
quality are critical to any monitoring program. The potential accuracy of the sample
analysis in the laboratory depends on the sampling methodology in the field. A
laboratory cannot generate reliable data if the sample was collected improperly.
Therefore, taking precautions and selecting the correct sampling methods are
imperative to ensure accurate and representative analyses.
Groundwater samples may not be representative of the aquifer for the following reasons:
1. The sample was taken from stagnant water in the well
. Water standing in a well and
exposed to the atmosphere may undergo a gas exchange (oxygen and carbon
dioxide) allowing chemical reactions to occur. Biological organisms capable of
driving reactions might also be introduced. Obviously, such waters will no longer be
representative of the water within the aquifer and should be purged prior to
sampling.
2. The sample was not collected at the appropriate time
. The sample should be
collected as soon as possible after purging is completed. This reduces the possibility
of chemical reactions occurring because of gas exchange and temperature
variations. In addition, if the well is pumped too long, the sample may represent
water so far from the well site that the groundwater chemistry is not representative of
the site being monitored.
3. The sample contained suspended or settleable solids
. Because of the natural filtering
action and slow velocity of most aquifers, groundwater is generally free of suspended
solids. However, even properly constructed monitoring wells will often fail to produce
samples that are free of sediment or settleable solids (turbidity). When samples
containing suspended solids are analyzed for metals, this sediment is digested
(dissolved) in the laboratory prior to performing the analysis. Consequently, any of
the metals present in the sediment (primarily iron, manganese, and aluminum) will be
included in the results of the analysis made on the water that includes these metals.
The analysis of the water samples containing sediment will result in certain analytes
such as these metals being reported at higher levels than they actually occur in
groundwater.
In addition to these common metals, numerous other metals that occur in the earth's
crust in trace amounts such as lead, chromium, and cadmium also will show up in the
analysis. The sediment content of the monitoring wells will often vary across a site,
and even samples collected from the same well at different times will vary in
sediment content. This problem can make analysis of monitoring well data for metals
where samples have not been filtered to remove turbidity an almost futile exercise.
4. The sample was not collected at the proper depth interval
. Sampling of domestic
water supply wells or wells with similar construction may encounter this problem.
Assume that an open well bore penetrates more than one aquifer with one of the
aquifers supplying contaminated water near the bottom of the well. If the sampling
interval is near the top of the well, the sample may reflect only the uncontaminated

383-3000-001 / December 1, 2001 / Page 53
aquifer with the contaminated water remaining near the bottom of the well in the
unsampled interval. However, because of mixing, it will not be possible to obtain a
true sample from the individual aquifers with this type of well construction.
5. Release of carbon dioxide during pumping increased the pH, allowing many metallic
ions to come out of solution (i.e. iron, manganese, magnesium, cadmium, arsenic,
selenium, and boron). Pumping can also cause volatilization of VOCs. This
emphasizes the importance of conducting field measurements such as pH, specific
conductance, temperature, etc. within the well before the sample is brought to the
surface.
6. Chemical changes occurred from oxidation of the sample during sampling
.
Dissolved oxygen is usually very limited within aquifers. Bringing the sample to the
surface allows oxygen to dissolve within the water sample. Oxidation also can occur
in the pump or can be caused by water cascading into a well installed in tight
formations. Depending on the chemical makeup of the sample, the addition of
dissolved oxygen may allow chemical reactions to proceed. Some of the changes
that can be expected include oxidation of 1) organics, 2) sulfide to sulfate, 3) ferrous
iron and precipitation of ferric hydroxide, 4) ammonium ion to nitrate, and 5)
manganese and precipitation of manganese dioxide or similar hydrous oxide. In
cases where oxidation would be expected to impact on chemical quality,
precautions should be employed to reduce oxidation potential (e.g. minimize
agitation during purging and sample collection, minimize the length of time the
sample is exposed to air, fill the sample container completely to the top, and
promptly chill the sample).
7. The sample was not preserved correctly
. Increases in temperature will allow certain
chemical reactions to occur. Certain metals, especially iron, may coat the insides of
the sample container. If the sample is not properly preserved for shipment to the
laboratory, the sample which arrives at the lab may be quite chemically different
from the sample which was collected in the field.
8. The sample was contaminated by residues in sampling equipment
. Residues may
cling to the sampling equipment if it is not properly cleaned or decontaminated.
Those residues may become mobile in succeeding samples, yielding unreliable
results. This becomes critical when the analytes being sampled are in the parts per
billion or parts per trillion range.

383-3000-001 / December 1, 2001 / Page 54
6.2
SAMPLE COLLECTION DEVICES
The most common devices available for the collection of water from monitoring wells
include bailers, suction-lift pumps, air-lift samplers, bladder pumps, and submersible
centrifugal pumps. Each has its advantages and disadvantages as shown in Table 2.
These should be considered before selecting the sample collection device.
Table 2. Advantages and disadvantages of different sampling devices.
ADVANTAGES
DISADVANTAGES
Bailer
Portable
Simple to use
Difficult to ascertain where within the water
column the sample is collected
Allows for oxidation of the sample
Disturbance of the water column by the
sampler
Impractical for removing large volumes of
water
Suction-lift
Pump
Allows sample to contact only
Teflon
Simple to use for shallow
applications
Limited to shallow groundwater conditions
Causes sample mixing, oxidation, and allows
for degassing
Air-lift
Sampler
Suited for small diameter wells
Causes extreme agitation
Significant redox, pH, and specie
transformations
Plastic tubing source of potential
contamination
Bladder
Pump
Provide a reliable means for
highly representative sample
Mixing and degassing minimized
Portable
Somewhat more complex than other samplers
Submersible
Centrifugal
Pump
Higher extraction rates
Considerable agitation and turbulent flow
Potential to introduce trace metals from the
pump materials

383-3000-001 / December 1, 2001 / Page 55
6.3
SAMPLE COLLECTION PROCEDURES
The following are general procedures that should serve as a framework for sampling
groundwater. These procedures should be modified where necessary for each situation
encountered in the field and to conform to monitoring objectives. In addition,
appropriate health and safety measures should always be taken before, during, and
after sampling.
6.3.1 Protective Clothing
Wear protective clothing if the nature of the contaminants dictate that it should
be worn. Different types of protective clothing are appropriate for different
contaminants.
6.3.2 Water Levels
If possible, determine and record the static water level of the well. This level
should be determined prior to well purging, which creates drawdown.
6.3.3 Field Measurements
In most cases, field measurements should be accomplished before and during
the sampling to gauge the purging of the well, and to measure any changes
between the time the sample is collected and when it is analyzed in the
laboratory. The following measurements and observations are often determined
in the field:
?? pH
?? Eh
?? water level
?? temperature
?? specific conductance
?? dissolved oxygen
?? acidity/turbidity
?? weather conditions
?? time of sampling
The specific techniques for obtaining each of these measurements depend upon
the instruments used. The operator should carefully read and follow the
manufacturer's instructions, including those for equipment maintenance and
calibration. A record of the calibration and maintenance checks should be kept.
6.3.4 Purging
The purpose of purging a well prior to sampling is to remove stagnant water from
the well bore and assure that the sample is representative of the groundwater in
the geologic formation being sampled. Stagnant water in the well bore results
from the water's contact with the casing and atmosphere between sampling
events. What would seem to be a relatively simple and straightforward
procedure, purging technique has, in fact, been the subject of considerable
scientific investigation and discussion.

383-3000-001 / December 1, 2001 / Page 56
There are two basic approaches to purging a well. The first is to use dedicated
equipment in which the water is pumped from a fixed position in the well. This
technique eliminates the possibility of cross-contamination, but tends to purge
only the well section or screen section that is opposite the purge pump. (This is
especially a concern when purge rates are much lower than the yield of the
water-bearing zone supplying water to the purge pump.)
The second basic approach is to use a transportable pump and purge from the
water surface, or preferably by gradually lowering the pump in the well as
stagnant water is evacuated. This technique is considered to be more reliable in
terms of evacuating the entire well bore. However, the disadvantage is that the
equipment must be decontaminated between wells, which in turn increases the
potential for cross-contamination.
An excellent summary of purging methods and techniques is given by Herzog
et al., (in Nielsen, 1991). The following discussion is based in part on that summary.
Four techniques for determining the volume of water to be purged from a well
are discussed. These include criteria based on:
1. Numbers of well bore volumes
2. Stabilization of indicator parameters
3. Hydraulic and chemical parameters
4. Time series sampling
By far, the most common choices have been to base the purging volume on
either a certain number of well volumes or stabilization of chemical analytes, or
some combination of these two.
6.3.4.1
Criteria Based on the Number of Bore Volumes
The purging of three well volumes has become so unquestionably
accepted and ingrained in monitoring practice as to be practically
elevated to the status of a scientific law. However, Herzog et al.
provide references from numerous studies that variously conclude that
from less than one to more than 20 bore volumes be purged from wells
prior to sampling. They conclude:
"It is obvious that it is not possible to recommend that a specific
number of bore volumes be removed from monitoring wells
during purging. The range of suggested volumes is too large
and the cost of improper purging is too great to permit such a
recommendation."
DEP recommends that if the borehole volume technique is going to be
used, the number of borehole volumes required for each well should
have a technical or scientific basis, such as stabilization of indicator
parameters (see following section) conducted at least once for each
well during initial sampling events, rather than being based on some
arbitrary criterion such as "three well volumes."
When purging is based on some set number of borehole volumes, the
borehole volume calculation should take into account the entire

383-3000-001 / December 1, 2001 / Page 57
original borehole diameter, corrected for the porosity of any sand or
filter pack, and not be based just on the innermost casing diameter.
6.3.4.2
Criteria Based on Stabilization of Indicator Parameters
Stagnant water in a well bore differs from formation water with respect
to many parameters. Field measurement of indicator parameters such
as temperature, pH, specific conductance, dissolved oxygen, and Eh
has been used as the criteria for determining the amount of water to
purge and when to sample a well. These parameters are measured in
the purge water during purging until they stabilize. DEP encourages the
use of this method.
DEP recommends that all of the above indicators should be measured
during the initial first few sampling events of the monitoring well. The
data should then be reviewed to determine which indicator parame-
ters are the most sensitive in indicating when stagnant water has been
evacuated from the well. The most sensitive parameters will be those
that show the greatest changes and longest times for stabilizing. During
the initial sampling, the purging time should probably be extended
beyond what initially appears to be stabilization as a check to assure
that the parameter stability is maintained. Where this technique is
chosen, DEP believes that the wells should be measured for the
indicator parameters suggested above and the three most sensitive
should be used.
6.3.4.3
Special Problems of Low Yielding Wells
Low yielding wells present a special problem for the sampler in that
they may take hours or even days to recover after purging to the extent
that there is enough water to take a sample. This waiting period not
only increases the cost of sampling but also allows changes in water
quality, especially with regard to volatile constituents, to occur
between the time the sample water enters the casing and the time it is
collected.
In practice, very low yield wells are commonly pumped dry and
sampled the following day if necessary. This practice is believed to
result in less than truly representative water being sampled from the well
due to the loss of volatiles and oxygenation of the water during the
waiting period, and as a result of pumping the well dry and exposing
the formation to the atmosphere. While there does not appear to be
any uniformly agreed upon method for eliminating these concerns, the
following considerations are suggested:
?? Purge in such a way that the water level does not fall below the
well screen.
?? Evaluate the use of larger diameter wells that may deliver the
required amount of sample water sooner than small diameter wells.

383-3000-001 / December 1, 2001 / Page 58
?? If full recovery cannot be achieved within two hours, collect the
required amount of water as it becomes available, collecting
samples for parameters in order of decreasing volatility.
6.3.4.4
Summary on Purging
The following general statements can be made with respect to purging:
?? Every groundwater monitoring plan should contain a section
dealing with how wells will be purged.
?? It is often desirable to use the same device for sampling that was
used for purging. In this case the purge pump can be set within the
screened section of the well, or across from the yielding zone being
monitored.
?? If different devices are used for purging and sampling, purging
should begin at the static water surface and the device should be
lowered down the well at a rate proportional to water stored in the
well bore. Because of the better mixing of water in wells with
multiple yielding zones, this technique is considered preferable for
sampling wells with multiple yielding zones where a composite
sample of water in the yielding zones is desired (see Section 3.5 on
Well Depths, Screen Lengths, and Open Intervals).
?? Where the same device is used to sample and purge a well, it must
be established that the sampling device will not change the quality
of the groundwater it comes into contact with.
?? In sampling for some analytes, such as volatile organics, it is critical
that the discharge be reduced to approximately 100 ml/minute to
minimize degassing and aeration (Barcelona et al., 1984). Flow
control should be by means of an electric current using a rheostat
rather than by valving or other flow restrictors.
?? Purging should be completed without lowering the water level in
the well below the well screen or water bearing zone being
sampled.
?? Never purge a well at such a rate or in such a way that water
cascades into the well bore resulting in increased degassing and
volatilization.
6.3.5 Management of Purge Water
Groundwater removed during purging should be handled in a way that is
environmentally compatible with the type and concentration of the suspected
contaminant in the aquifer. Monitoring instruments such as photoionization meters
should be used when appropriate to periodically screen the groundwater. A
procedure that can be used is outlined in Table 3. The goal of handling

383-3000-001 / December 1, 2001 / Page 59
potentially contaminated groundwater is to safely discharge the purge water
while avoiding pollution of another part of the environment, such as surface
water bodies or another aquifer.
6.3.6 Private Wells
If the well is a private water supply, sample as close to the well as physically
practical and prior to any treatment or filtering devices if possible and practical.
If collection has to be made from a holding tank, allow water to flow long
enough to flush the tank and the lines. If a sample that passes through a
treatment tank must be taken, the type, size, and purpose of the unit should be
noted on the sample data sheet and in the field log book.
6.3.7 Filtering
When possible, avoid collecting samples which are turbid, colored, cloudy or
contain much suspended matter. Exceptions to this include when the sample site
has been pumped and flushed, or has been naturally flowing for a sufficient
amount of time to confirm that these conditions are representative of the aquifer
conditions.
Unless analysis of unfiltered samples for "total metals" is specifically required by
program regulation or guidance, all samples for metals analysis should be field-
filtered through a 0.45-micron filter prior to analysis.
6.3.8 Sample Preservation
Perform sample preservation techniques on-site as soon as possible after the
sample is collected. Complete preservation of samples is a practical impossibility.
Regardless of the nature of the sample, complete stability for every constituent
can never be achieved. For this reason, samples should be analyzed as soon as
possible. However, the ongoing chemical and biological changes in the sample
may be slowed significantly by proper preservation techniques.
Chemical changes generally happen because of a shift in the physical
conditions of the sample. Under a fluctuation in the reducing or oxidizing
conditions, the valence number of the cations or anions may change; other
analytes may volatilize or dissolve; metal cations may form complexes or
precipitate as hydroxides, or they may adsorb onto surfaces.
Biological changes can also alter the valence of a constituent. Organic
processes may bind soluble material into the cell structure, or cell material may
be released into solution.

383-3000-001 / December 1, 2001 / Page 60
Table 3. Suggested procedure for the management of purge water
from groundwater sampling.
TYPE OF
GROUNDWATER
ACTION
Uncontaminated
Groundwater
Disposal may proceed with normal precautions (i.e. avoid erosion,
stream discharge, and hazards associated with freezing conditions;
also, monitor for adverse changes in water quality)
Contaminated
Groundwater
a) Convey directly into an on-site treatment plant
b) Take to off-site treatment
c) Place into an area where it will drain to a collection site for on or
off site treatment (i.e. leachate collection systems)
d) Discharge to ground surface if it is determined that based on the
type and concentration of the contaminant , and the volume of
the discharge, that the discharge will not impact any surface
water body or cause environmental harm
e) Other methods approved by DEP
Groundwater
of Uncertain
Quality
Discharge to ground if all of the following conditions are met*:
a) The well is not associated with a corrective action or a
remediation project
b) The groundwater shows no obvious sign of contamination, such as
odor, color, or visually-apparent material in the water, or readings
from monitoring instruments
c) The discharge will not impact any surface water body or cause
environmental harm
*If these conditions cannot be met, then the water is classified as
contaminated groundwater and must be handled as indicated
above.
Methods of preservation are relatively limited and are intended to generally 1)
retard biological activity, 2) retard hydrolysis of chemical compounds and
complexes, 3) reduce the volatility of constituents, and 4) reduce sorption effects.
Preservation methods are generally limited to pH control, chemical addition,
refrigeration, freezing, and selecting the type of material used to contain the
sample.
The best overall preservation technique is refrigeration at or about 4?C.
Refrigeration mainly helps to inhibit bacteria. However, this method is not always
applicable to all types of samples.
Acids such as HNO
3
and H
2
SO
4
can be used to prevent precipitation and inhibit
the growth of bacteria. Preservation methods are specified in the Bureau of
Laboratories Users Guide for each of the standard analyses listed.

383-3000-001 / December 1, 2001 / Page 61
6.3.9 Decontamination of Sampling Devices
The methods used will vary depending upon the analytes which are being
sampled. This is extremely important when sampling for constituents thought to
be present in the parts per billion or parts per trillion range. Decontamination
procedures are described in Appendix A.
6.4
REFERENCES
BARCELONA, M.J., HELFRICH, J.A., GARSKE, E.E., and GIBB, J.P., 1984, A Laboratory Evaluation
of Groundwater Sampling Mechanisms, Groundwater Monitoring Review, v.4, No.2, pp. 32-
41.
DRISCOLL, F.G., 1986, Groundwater and Wells, Second Edition, Johnson Filtration Systems,
Inc., St. Paul, Minnesota 55112, 1089 pp.
GIBB, J.P., SCHULLER, R.M., and GRIFFIN, R.A., 1981, Procedures for the Collection of
Representative Water Quality Data from Monitoring Wells, Cooperative Groundwater
Report 7, Illinois State Water Survey and Illinois State Geological Survey, Champaign, IL.
NIELSEN, D.M., (Editor), 1991, Practical Handbook of Ground-Water Monitoring, NWWA,
Lewis Publishers, Inc., Chelsea, Michigan 48118, 717 pp.

383-3000-001 / December 1, 2001 / Page 62

Back to top


CHAPTER 7: WELL ABANDONMENT PROCEDURES
7.1
INTRODUCTION
Unsealed or improperly sealed wells may threaten public health and safety, and the
quality of the groundwater resources. Therefore, the proper abandonment
(decommissioning) of a well is a critical final step in its service life.
Act 610, the Water Well Drillers License Act, includes a provision for abandonment of
wells. This legislation makes it the responsibility of a well owner to properly seal an
abandoned well according to the rules and regulations of the Department of
Conservation and Natural Resources (DCNR). In the absence of more stringent
regulatory standards, the procedures outlined in this section represent minimum
guidelines for proper abandonment of wells and borings. These procedures may be
applicable for, but not limited to, public and domestic water supply wells, monitoring
wells, borings or drive points drilled to collect subsurface information, test borings for
groundwater exploration, and dry wells (drains or borings to the subsurface).
Proper well abandonment accomplishes the following: 1) eliminates the physical hazard
of the well (the hole in the ground), 2) eliminates a pathway for migration of
contamination, and 3) prevents hydrologic changes in the aquifer system, such as the
changes in hydraulic head and the mixing of water between aquifers. The proper
decommissioning method will depend on both the reason for abandonment and the
condition and construction details of the boring or well.
7.2
WELL CHARACTERIZATION
Effective abandonment depends on knowledge of the well construction, geology, and
the hydrogeology. The importance of a full characterization increases as the complexity
of the well construction, site geology, and the risk of aquifer contamination increases.
Construction information for wells drilled since 1966 may be available from DCNR, Bureau
of Topographic and Geologic Survey's (BTGS) Water Well Inventory System database.
Additional well construction data and information describing the hydrologic
characteristics of geologic formations may be available from reports published by BTGS
and the United States Geological Survey (USGS). Site or program records also may exist.
The well should be positively identified before initiating the abandonment. Field
information should be compared with any existing information.
Water levels and well depths can be measured with a well sounder or weighted tape
measure. In critical situations, well construction details and hydrogeology can be
determined with borehole geophysics or a downhole camera. For example, a caliper
log, which is used to determine the borehole diameter, can be very helpful in locating
cavernous areas in open hole wells.
7.3
WELL PREPARATION
If possible, the borehole must be cleared of obstructions prior to abandonment.
Obstructions such as pumps, pipes, wiring, and air lines must be pulled. Well preparation
also may involve fishing obstacles out of the borehole. An attempt should be made to
pull the casing when it will not jeopardize the integrity of the borehole. Before the casing

383-3000-001 / December 1, 2001 / Page 63
is pulled, the well should be grouted to near the bottom of the casing. This will at least
provide some seal if the well collapses after the casing is pulled.
The presence of nested or telescoped casing strings complicates well abandonment.
Inner strings should be removed when possible, but only when removal will not jeopardize
the abandonment of the well. If inner strings cannot be removed and sealing of the
annular space is required, then the inner string should be vertically split (plastic cased
wells) or cut (metal-cased wells) at intervals necessary to insure complete filling of the
annular space.
Damaged, poorly constructed or dilapidated wells may need to be redrilled in order to
apply proper abandonment techniques. Also, in situations where intermixing of aquifers
is likely, the borehole may need to be redrilled.
7.4
MATERIALS AND METHODS
7.4.1 Aggregate
Materials that eliminate the physical hazard and open space of the borehole, but
do not prevent the flow of water through the well bore, are categorized as
aggregate. Aggregates consist of sand, crushed stone or similar material that is
used to fill the well. Aggregates should be uncontaminated and of consistent size
to minimize bridging during placement.
Aggregate is usually not placed in wells smaller than two inches in diameter.
Nominal size of the aggregate should be no more than 1/4 of the minimum well
diameter through which it must pass during placement. Because aggregate is
usually poured from the top of the well, care must be taken to prevent bridging
by slowly pouring the aggregate and monitoring the progress with frequent
depth measurements.
Aggregates may be used in the following circumstances: 1) there is no need to
penetrate or seal fractures, joints or other openings in the interval to be filled, 2) a
watertight seal is not required in the interval to be filled, 3) the hole is caving, 4)
the interval does not penetrate a perched or confined aquifer, and 5) the
interval does not penetrate more than one aquifer. If aggregate is used, a
casing seal should be installed (see Section 7.5.1). The use of aggregate and a
casing seal must be consistent with the future land use.
7.4.2 Sealants
Sealants are used in well abandonment to provide a watertight barrier to the
migration of water in the well bore, in the annular spaces or in fractures and
openings adjacent to the well bore. Sealants usually consist of Portland cement
based grouts, "bentonite" clay, or combinations of these substances. Additives
are frequently used to enhance or delay specific properties such as viscosity,
setting time, shrinkage, or strength.
Sealing mixtures should be formulated to minimize shrinkage and ensure
compatibility with the chemistry of the groundwater in the well.
A grout pump and tremie pipe are preferred for delivering grout to the bottom of
the well. This method insures the positive displacement of the water in the well,
and will minimize dilution or separation of the grout.

383-3000-001 / December 1, 2001 / Page 64
If aggregate is to be placed above sealant, a sufficient amount of curing time
should pass before placing the aggregate above the seal. Curing time for grout
using Type 1 cement is typically 24 - 48 hours, and 12 hours for Type III cement.
General types of sealants are defined as follows:
Neat cement grout
: Neat cement grout is generally formulated using a ratio of
one 94-pound bag of Portland cement to no more than 6 gallons of water. This
grout is superior for sealing small openings, for penetrating any annular space
outside of the casings, and for filling voids in the surrounding rocks. When applied
under pressure, neat cement grout is strongly favored for sealing artesian wells or
those penetrating more than one aquifer. Neat cement grout is generally
preferred to concrete grout because it avoids the problem of separation of the
aggregate and the cement. Neat cement grout can be susceptible to shrinkage
and the heat of hydration can possibly damage some plastic casing materials.
Concrete grout
: Concrete grout consists of a ratio of not more than six gallons of
water, one 94-pound. bag of Portland cement, and an equal volume of sand.
This grout is generally used for filling the upper part of the well above the water
bearing zone, for plugging short sections of casings, or for filling large-diameter
wells.
Concrete grout, which makes a stronger seal than neat cement, may not
significantly penetrate seams, crevices or interstices. Grout pumps can handle
sand without being immediately damaged. Aggregate particles bigger than this
may damage the pump. If not properly emplaced, the aggregate is apt to
separate from the cement. Concrete grout should generally not be placed
below the water level in a well, unless a tremie pipe and a grout pump are used.
Grout additives
: Some bentonite (2 to 8 percent) can be added to neat cement
or concrete grout to decrease the amount of shrinkage. Other additives can be
used to alter the curing time or the permeability of the grout. For example,
calcium chloride can be used as a curing accelerator.
High-solids sodium bentonite
: This type of grout is composed of 15-20 percent
solids content by weight of sodium bentonite when mixed with water. To
determine the percentage content, the weight of bentonite is divided by the
weight of the water plus the weight of the bentonite. For example, if 75 pounds
of powdered bentonite and 250 pounds of granular bentonite were mixed in 150
gallons of water (at 8.34 pounds per gallon), the percentage of high-solids
bentonite is approximately 20 percent (325/(1251+325)). High-solids bentonite
must be pumped before its viscosity is lowered. Pumping pressures higher than
those used for cement grouts are usually necessary. Hydration of the bentonite
must be delayed until it has been placed down the well. This can be done by 1)
using additives with the dry bentonite or in the water, 2) mixing calcium bentonite
(it expands less) with sodium bentonite, or 3) using granular bentonite, which has
less surface area.
In addition, positive displacement pumps such as piston, gear, and moyno
(progressive cavity) pumps must be used because pumps that shear the grout
(such as centrifugal pumps) will accelerate the congealing of the bentonite. A
paddle mixer is typically used to mix the grout. A high-solids bentonite grout is not
made from bentonite that is labeled as drilling fluid or gel.

383-3000-001 / December 1, 2001 / Page 65
Chip Bentonite
: Chip (coarse grade) or pelletized bentonite can form adequate
seals. This type of bentonite is poured directly down the borehole. The size of the
bentonite chips also should be no more than 1/4 of the minimum well diameter
through which it must pass during placement. Because of the potential for
bridging, this material may not be suitable for deep wells or borings where
positive displacement is necessary to seal the well.
When coarse bentonite is placed above the water level, water must be added
frequently to hydrate the bentonite. Care must be taken with chip or pelletized
bentonite to not overload the interval to be sealed. Rapidly swelling bentonite
could result in incomplete hydration and a heterogeneous seal containing lumps
of dry bentonite. The level of the bentonite should be checked often to make
sure that bridging of the chips does not occur.
7.4.3 Bridge Seals
A bridge seal can be used to isolate cavernous sections of a well, to isolate two
producing zones in the well, or to provide the structural integrity necessary to
support overlying materials (and thus protect underlying aggregate or sealants
from excessive compressive forces). Bridge seals are usually constructed by
installing an expandable plug made of wood, neoprene, or a pneumatic or other
mechanical packer. Additional aggregate can be placed above the bridge.
7.5
RECOMMENDATIONS
The complexity of the abandonment procedure depends primarily on the hydrogeology,
geology, well construction, and the groundwater quality. Four principal complicating
factors have been identified; they include 1) artesian conditions, 2) multiple aquifers, 3)
cavernous rocks, and 4) the threat or presence of contamination. The recommended
procedures for abandoning wells will be more rigorous with the presence of one or more
complicating factors. The procedures may vary from a simple casing seal above
aggregate to entirely grouting a well using a tremie pipe after existing casing has been
ripped or perforated. Figure 10 summarizes the general approach to well abandonment.

383-3000-001 / December 1, 2001 / Page 66

Back to top


Figure 10. Summary of procedures for well abandonment.
1.
Casing seal
2.
Aggregate
3.
Sealant
4.
Possibly use bridge seal or plug
5.
Possibly cut existing casing
Note
:
Program requirements must
be followed where applicable.
Wells in unconfined or semi-confined
conditions (may include test wells,
borings, and domestic water supply wells)
1,2
NO
Wells in
cavernous
rock
1,2,3,4,5
Wells with
multiple
aquifers
1,2,3,4,5
Artesian
wells
1,2,3,4,5
NO
Wells in
contaminated
aquifers
1,3,5
NO
Wells in
cavernous
rock
1,3,4,5
Wells with
multiple
aquifers
1,3,4,5
Artesian
wells
1,3,4,5
YES
Is there an
additional
complicating
factor?
YES
Is there existing
or potential ground-
water contamination?
YES
Is there a complicating factor:
existing or potential groundwater
contamination; artesian pressure;
multiple aquifers; or cavernous rock?

383-3000-001 / December 1, 2001 / Page 67
7.5.1 Casing Seal
The transition from well casing to open borehole is the most suspect zone for
migration of water. In order to minimize the movement of water (contaminated
or otherwise) from the overlying less consolidated materials to the lower
waterbearing units, this zone must be sealed. Generally this can be
accomplished by filling at least the upper 10 feet of open borehole and the lower
five feet of casing with sealant. The length of open borehole sealed should be
increased if extenuating circumstances exist. Such circumstances would include
a history of bacterial contamination, saprolitic bedrock, or possibly deep fracture
zones. Waterbearing zones reported in the upper 20 feet or so of open borehole
are indications of fractures and would warrant additional sealant. Casing that is
deteriorated should be sealed along its entire length. If the casing is to be pulled
the sealant used should remain fluid for a period of time adequate for removal of
the casing.
If the casing is to remain, then whenever feasible, it should be cut off below land
surface. After the casing seal discussed above achieves adequate strength, the
open casing should at a minimum, be filled with aggregate. It is strongly
suggested that a sealant be used in the upper 2 to 5 feet of casing.
7.5.2 Wells in Unconfined or Semi-Confined Conditions
These are the most common type of wells in Pennsylvania. The geology may
consist of either unconsolidated or consolidated materials. When applicable,
unconfined wells in non-contaminated areas may be satisfactorily abandoned
using aggregate materials up to 10-15 feet below the ground surface. This would
apply mainly to domestic wells, and test borings or wells not covered by existing
regulations. Monitoring wells that are not covered by specific regulatory
programs and are located at sites with no known contamination, might be
abandoned in this manner. Above the aggregate, the casing seal should be
installed. A sealant may be used over the entire depth.
7.5.3 Wells at Contaminated Sites
An abandoned, contaminated well often mixes contaminated groundwater with
uncontaminated groundwater. Complete and uniform sealing of the well from
the bottom to the surface is required. Therefore, proper well preparation (Section
7.3) must be done before the well is sealed with a proper sealant (Section 7.4.2).
7.5.4 Wells in Cavernous Rocks
Problems can arise when filling wells that penetrate cavernous rock. Although
such wells are usually located in carbonate terrain, voids can also occur in areas
that have been deep mined. Care must be taken to insure that aggregates and
sealants are of a size and consistency to prevent their removal by water flowing in
the void. Large voids or high flow velocities warrant placement of a bridge in
competent rock over the void. Aggregate and sealants can then be placed
above the bridge.

383-3000-001 / December 1, 2001 / Page 68
7.5.5 Multiple Aquifer Wells
The main goal in sealing wells that extend into more than one aquifer is to
prevent the flow of groundwater from one aquifer to another. If no appreciable
movement of water is encountered, and there is no threat of groundwater
contamination, sealing with concrete, neat cement, grout, or alternating layers of
these materials and aggregate will prove satisfactory. When groundwater
velocities are high, the procedures for wells with artesian flow (see the next
section) are recommended. If alternating plugs (or bridges) and aggregate
layers are used, the plugs should be placed in known nonproductive horizons or, if
locations of the nonproductive horizons are not known, at frequent intervals.
7.5.6 Flowing Wells
The sealing of artesian wells requires special attention. The flow of groundwater
may be sufficient to make sealing by gravity placement of concrete, cement
grout, neat cement, clay or sand impractical. In such wells, large stone
aggregate (not more than 1/4 of the diameter of the hole), well packers
(pneumatic or other), or wooden plugs will be needed to restrict the flow and
thereby permit the gravity placement of sealing material above the zone where
water is produced. If plugs are used, they should be several times longer than the
diameter of the well to prevent tilting. Seals should be designed to withstand the
maximum anticipated hydraulic head of the artesian aquifer.
Because it is very important in wells of this type to prevent circulation between
water yielding zones, or loss of water to the surface or to the annular spacing
outside of the casing, it is recommended that pressure grouting with cement be
done using the minimum volume of water during mixing that will permit handling.
In wells in which the hydrostatic head producing flow to the surface is low, the
movement of water may be stopped by extending the well casing to an
elevation above the artesian pressure surface.
7.5.7 Wells with Complicating Factors at Contaminated Sites
Wells with one or more of the above complicating factors that are to be
abandoned in areas with contaminated groundwater or in areas where the
groundwater is at a high risk for future contamination, require the most rigorous
abandonment procedures. In general, the entire length of these wells should be
sealed.
When the threat of contamination has been established, the elimination of a
potential flowpath is critical. For example, a contaminated well in a karst terrain
must be carefully sealed to avoid worsening the situation. In general, the entire
lengths of these wells should be sealed. In some situations, a bridge seal may
have to be installed, and casing may have to be perforated. In each case, a
prudent method should be selected that will eliminate all potential vertical
flowpaths.
7.5.8 Monitoring Wells
Monitoring wells should be abandoned in accordance with the rules and
regulations of the program under which they were installed and operated.

383-3000-001 / December 1, 2001 / Page 69
Monitoring wells which do not fall under the jurisdiction of a regulatory program,
or fall under a program that has no rules or regulations for abandonment, should
be abandoned under the following guidelines.
Monitoring wells that were installed and continue to function as designed, can
usually be abandoned in place. Exceptions would include wells whose design
precludes complete and effective placement of sealant and wells in locations
subject to future disturbance that could compromise the abandonment. In such
instances all tubing, screens, casings, aggregate, backfilling, and sealant should
be cleaned from the boring and the hole should be completely filled with an
appropriate sealant.
Monitoring wells that are abandoned in place should be completely filled with
sealant. Screened intervals can be backfilled with inert aggregate if sealant will
alter the groundwater chemistry and thereby jeopardize ongoing monitoring at
the facility. Intervals between screens, and between the last screen and the
surface, must be filled with sealant. Generally, sealant must be emplaced from
the bottom of the interval being sealed. Protective casings, riser pipes, tubing,
and other appurtenances at the surface which could not be removed should be
cut off below grade after the sealant has properly set. When the abandonment
will be completed below the finished grade, the area of the boring should be
covered with a layer of bentonite, grout, concrete, or other sealant before
backfilling to grade.
7.6
EXISTING REGULATIONS AND STANDARDS
The Water Well Drillers License Act requires that the owner or consultant who is to
abandon the well notify the DCNR, Bureau of Topographic and Geologic Survey of the
intent to decommission a well at least 10 days before the well is sealed or filled.
Individual DEP bureaus may have specific regulations or guidelines.
The Bureau of Oil and Gas Management regulates the plugging of oil and gas wells.
Plugging provisions for oil and gas wells in coal and non-coal areas are established in
§ 210 and § 211 of Act 223, and § 78.91 - 78.97 of Chapter 78. These sections describe
methods that would stop any vertical flow of fluids or gas within the well bore. Alternate
methods of plugging also are allowed if they would afford the same level of protection.
Alternate methods must be approved before the plugging is initiated.
The Bureau of Mining and Reclamation regulates the abandonment of borings and wells
associated with the mining of coal. Coal exploration holes must be abandoned
according to the § 87.93 for surface mining of bituminous coal, § 88.83 for anthracite
coal mining, § 89.54 for deep mining of bituminous coal, and § 90.93, coal refuse
disposal.
The Bureau of Water Supply and Wastewater Management uses the AWWA Standard for
Water Wells for abandonment of public water supply wells. This standard is referenced in
Part II of the Public Water Supply Manual.
7.7
REPORTING
All abandoned wells shall be reported to BTGS, along with any bureau that requires a
report, on forms required by BTGS (and any other forms). If available, the original driller's

383-3000-001 / December 1, 2001 / Page 70
log should be included along with the details of the well abandonment procedure. A
photograph should be taken of the site, and a reference map should be made to locate
the abandoned well. It also may be appropriate to survey the exact location of the well.
This is especially important for wells associated with contaminated sites.
7.8
REFERENCES
AMERICAN WATER WORKS ASSOCIATION, 1990, Abandonment of Test Holes, partially
completed wells and completed wells: AWWA Standard for Water Wells, pp. 25-26.
DRISCOLL, F.G., 1986, Groundwater and Wells, 2nd ed., Johnson Filtration Systems, Inc., St.
Paul, Minnesota 55112, 1089 pp.
NYE, J.D., September 1987, Abandoned Wells - How One State Deals with Them, Water
Well Journal, pp. 41-46
RENZ, M.E., May 1989, In Situ Decommissioning of Ground Water Monitoring Wells, Water
Well Journal, pp. 58-60.
U.S. ENVIRONMENTAL PROTECTION AGENCY, 1975, Manual of Water Well Construction
Practices, Office of Water Supply, EPA-570/9-75001.

383-3000-001 / December 1, 2001 / Page 71

Back to top


WELL ABANDONMENT FORM
CONTRACTOR/AGENT:
REGISTRATION NO.
DATE:
TYPE OF SITE OR PROGRAM:
1. WELL LOCATION: (Show sketch of location on back of this form.)
Municipality:
County
Quadrangle
(Road, community, subdivision, lot number)
Latitude
Longitude
2. OWNER AND ADDRESS:
3. TOPOGRAPHY: (Circle) hilltop,
slope,
stream terrace,
valley,
stream channel,
draw,
local depression, flat
4. USE OF WELL:
5. DEPTH OF WELL:
WELL DIAGRAM: sketch a diagram showing
depths of well, casing (if present), grouting
materials, perforations, etc.
DIAMETER OF WELL
6. AMOUNT OF CASING REMOVED
DIAMETER:
Neat cement
Sand cement
7. SEALING
Bags (94 lbs.):
MATERIAL
gals of water:
yds of sand:
OTHER MATERIAL
amount:
8. EXPLAIN METHOD OF EMPLACEMENT OF MATERIAL:
9. CERTIFICATION: We hereby certify that this well abandonment record is true and exact, and was
accomplished on
day of the month of
,
with our active
with our active participation and that we are qualified to participate in such abandonment actions.
Signature of Participant:
Signature of Participant:
Address:
Address:
Date:
Date:

383-3000-001 / December 1, 2001 / Page 72

Back to top


CHAPTER 8: QUALITY ASSURANCE/QUALITY CONTROL REQUIREMENTS
8.1
PURPOSE
A Quality Assurance/Quality Control Plan (QA/QC Plan) is a detailed account of methods
and procedures that are used in data collection (i.e. monitoring) activities. This plan,
when properly developed and implemented, ensures that adequate control and
documentation procedures are utilized from initiation to completion of the monitoring so
that the data generated are of the highest quality and can be used for the intended
purpose with confidence. A QA/QC plan is also an effective tool in assessing and
assuring the completeness and adequacy of the basic monitoring plan.
8.2
DESIGN
A QA/QC plan should be designed to satisfy the objectives of the monitoring project.
Although the elements of each QA/QC plan, as described below, will be basically similar,
the intended uses of the collected data will determine the requirements associated with
the monitoring activity. For example, data collection to achieve objectives of an
ambient monitoring project will require different QA/QC procedures for some QA/QC
plan elements than data collection for compliance monitoring for a permitted activity or
remediation monitoring activity. In addition, in most cases, there will be sufficient
differences within these monitoring categories for each project to require a specific
QA/QC plan.
The following paragraphs describe the basic elements of a QA/QC plan. In most cases,
the proper development and adherence to this format will be sufficient to ensure that
the data collection meets the objectives of a project. However, in some cases it may be
necessary to include additional considerations that may be unique to a specific site
and/or project. The Bureau of Laboratories should be consulted on any questions or
problems which may arise in developing any QA/QC plan.
8.3
ELEMENTS
A. Project Name or Title
: Provide the project identification and location.
B. Project Required by
: Provide the reason(s) or requirement(s) for the project, such as
(name) permit compliance monitoring or (name) remediation monitoring.
C. Date of Requirement
: Provide date the project was required either by permit or by
legal or other order.
D. Date of Project Initiation
: Provide date that the project was implemented.
E. Project Officer(s)
: Provide name(s) of individual(s) responsible for managing or
overseeing the project.
F. Quality Assurance Officer(s)
: Provide name(s) of individual(s) responsible for
development of and adherence to the QA/QC plan.
G. Project Description
: Provide 1) an objective and scope statement which
comprehensively describes the specific objectives and goals of the project, such as
determining permit compliance, treatment technology effectiveness or remediation
effectiveness for specific parameters, 2) a data usage statement that details how the

383-3000-001 / December 1, 2001 / Page 73
monitoring data will be evaluated including any statistical or other methods, 3) a
description of the location of monitoring stations and reasons or such locations,
including geological, hydrogeological or other considerations, and 4) a description of
the monitoring analytes and frequency of collection including for each analyte the
number of samples expected to be collected, the sample matrix (i.e. water) the
exact analytical method, reason for selection of analytes, reference, and sample
preservation method and holding time.
H. Project Organization and Responsibility
: Provide a list of key personnel and their
corresponding responsibilities, including the position and/or individual in charge of
the following functions: field sampling operations, field sampling QA/QC, laboratory
analyses, laboratory analyses QA/QC, data processing activities, data processing
QA/QC and overall project coordination.
I. Project Fiscal Information
: Provide an estimate in work days of the project time
needed for data collection, laboratory support, data input, quality assurance and
report preparation.
J. Schedule of Tasks and Products
: Provide a projected schedule for completing the
various tasks and developing the products associated with the project, such as
sample collections (monthly, quarterly, etc.), data analysis/reports (quarterly, annual,
biennial, etc.).
K. Data Quality Requirements and Assessments
: Provide a description of data accuracy
and precision (consult the Bureau of Laboratories for this information), data
representativeness, data comparability, and data completeness.
L. Sampling Procedures
: Provide a description of the procedures used to collect
samples from monitoring wells or other sites, including sampling containers and field
preservation and transport procedures.
M. Sampling Plan
: A sampling plan should provide necessary guidance on the number
and types of sampling quality controls to be used. The following is a list of common
sample quality control types and a recommended minimum frequency if used. It is
important to remember that all of these quality control samples must be treated with
the same dechlorination and/or preserving reagents as the field samples they are
associated with.
1)
Trip Blanks
: These are appropriate sample containers filled with laboratory
quality reagent water that are transported to and from the sampling site(s) and
are shipped with the samples to the laboratory for analysis. These samples are
intended to determine if there was any cross contamination that occurred
during the shipping process. They will also validate that the sampling
containers used were clean. Each sampling event that uses this type of quality
control should have a minimum of one trip blank for each container type used.
2)
Field Blanks
- These are appropriate sample containers that are filled at the
sampling site(s) with laboratory quality reagent water and are shipped with the
samples to the laboratory for analysis. These samples are intended to
determine if there was any cross contamination that occurred during the
sampling process due to ambient conditions. They will also validate that the
sampling containers used were clean. Each sampling event that uses this type
of quality control should have a minimum of one field blank for each sampling

383-3000-001 / December 1, 2001 / Page 74
site of each container type used. This type of sampling quality control is most
useful when sampling for volatile compounds.
3)
Rinsate Blanks
- These are samples of laboratory quality reagent water used to
rinse the collection device, including filtration devices and filters, coming into
contact with the same surfaces as the sample. The quality control samples(s)
are then submitted with the field samples for analysis. This type of quality
control sample helps to determine if the sample collection device is
contributing any detectable material to the sample. The minimum number of
this type of sampling quality control, if utilized, is dependent on operational
considerations. If multiple samples are being collected with the same
collection device, with field decontamination, you should, at a minimum,
submit two of this type of quality control, one before sampling and one at the
end of sampling. If you are using disposable sample collection devices or
multiple pre-cleaned devices, then a single representative sample should
suffice.
4)
Split/Duplicate Samples
- This is a single large sample that has been
homogenized, split into two or more individual samples, and each sample
submitted independently for analysis. This quality control determines the
amount of variance in the entire sampling/analysis process. This type of quality
control is not recommended for samples analyzed for analytes that would be
adversely affected by the homogenization process (i.e. volatile organics). The
minimum number of this type of sampling quality control, if utilized, is one per
sampling event with a rate of 5 percent to 10 percent commonly used.
5)
Replicate Samples
- These are two or more samples that are collected from the
same source, in a very short time frame (i.e. minutes), and each sample
submitted independently for analysis. This quality control, like the
split/duplicate sample, determines the amount of variance in the entire
sampling/analysis process. The amount of variance determined by this type of
quality control may be larger than that of a split/duplicate sample. The use of
this type of quality control also presumes that the samples material is already
homogenous. This type of quality control is recommended for samples
analyzed for analytes that would be adversely affected by an external
homogenization process (i.e. volatile organics). The minimum number of this
type of sampling quality control, if utilized, is one per sampling event with a
rate of 5 percent to 10 percent commonly used.
6)
Known Samples
- These are reference materials that have been characterized
as to the acceptable range of values for the analytes of concern. These
materials are available from commercial sources. This type of quality control
helps determine if the analytical work has adequate accuracy. It must be
noted that improper handling or storage of this type of reference material can
invalidate the materials characterization. The minimum number of this type of
quality control, if used, would be one per subject.
7)
Spiked Samples
- These are split/duplicate or replicate samples that have been
fortified with the analytes of concern. This quality control is intended to
determine if there have been changes in concentration due to factors
associated with the sample, or shipping and analysis process. This type of
quality control is very difficult to use in a field environment and routinely is done

383-3000-001 / December 1, 2001 / Page 75
as part of the analysis process. If this type of quality control is determined to be
needed, the minimum would be one per project.
N. Sample Custody Procedures:
Provide information which describes accountability for
sample chain of custody including sample collector identification, sample location
identification, sample number, date and time of collection, parameters monitored,
preservatives and fixatives, identification of all couriers, identification of laboratory
and receiver, time and date of receipt at laboratory, laboratory analyzer, and time
and date of analysis. The Bureau of Laboratories should be consulted to ensure that
this section is consistent with current policy and procedure.
O. Calibration Procedures and Preventative Maintenance:
Equipment maintenance
and calibration should be performed in accordance with manufacturers instructions.
Calibration and maintenance sheets should be maintained on file for all equipment.
P. Documentation, Data Reduction, and Reporting:
Provide discussion on where field
data are recorded, reviewed, and filed.
Q. Data Validation:
Provide a discussion on or reference to the protocols for validation
of chemical data and field instrumentation and calibration. Describe procedure for
validating database fields (i.e. through error checking routines, automatic flagging of
data outside of specified ranges, and manual review and spot checking of data
printouts against laboratory analytical results).
R. Performance and Systems Audits:
Provide description of how field staff performance
is checked and data files are verified for accuracy and completeness.
S. Corrective Action:
Provide discussion on making corrections when errors are found,
and actions taken to prevent recurrence of errors.
T. Reports:
Provide list of types and frequency of reports to be generated (i.e.
performance and systems audits, compliance analyses, remediation effectiveness,
etc.).
8.4
REFERENCE
U.S. ENVIRONMENTAL PROTECTION AGENCY, May 1984, Guidance for Preparation of
Combined Work/Quality Assurance Project Plans for Environmental Monitoring, (OWRS
QA-1), USEPA Office of Water Regulations and Standards.

383-3000-001 / December 1, 2001 / Page 76

Back to top


APPENDIX
DECONTAMINATION PROCEDURES
Equipment Cleaning
The following guidelines provide methods for cleaning sampling equipment used for sampling
wells.
(1)
Portable Pump
Disassemble the pump by unscrewing the discharge-hose adaptor and removing
the inlet screen; then, remove the four Phillips screws which secure the
pumphead to the motor case (Do not remove the single screw at the base of the
motor housing).
Clean the Teflon rotors and the stainless stator in warm detergent solution, using a
stiff bristled brush to clean the parts. The hose adaptor, pumphead, screen and
screws should also be cleaned and, if deemed necessary, they too may be
scoured with a brush.
Rinse all parts with tap water.
Rinse all parts with 10 percent nitric acid solution.
Rinse all parts with distilled water (Note: Procedure to be detailed later).
Reassemble pump and rinse internally and externally with ASTM Type IV or better
reagent grade water.
(2)
Submersible Pump
Since these pumps usually remain dedicated to one well, cross contamination is
not a problem. If for any reason the pump is removed or relocated to another
well, the decontamination procedures described above should be followed,
except that the pump need not be disassembled.
(3)
Bailers
Clean bailer and rope or wire line with warm detergent solution.
Rinse with tap water.
Rinse bailer with 10 percent nitric acid solution.

383-3000-001 / December 1, 2001 / Page 77
Rinse bailer and rope twice with distilled water, once with ASTM Type IV or better
reagent grade water, drain, and air dry in an uncontaminated area.
Place clean bailers and ropes in clean transportation tubes or wrap in clean
aluminum foil.
(4)
Field Filtration Apparatus
Non-Disposable Filtration Apparatus
- Disassemble permanent filtration kit parts,
then wash all parts in warm detergent solution, then rinse all parts with tap water,
then rinse all parts with 10 percent nitric acid solution, then rinse all parts twice
with distilled water, once with ASTM Type IV or better reagent grade water, and
then air dry.
Disposable Filtration Apparatus
- No cleaning is required for the disposable
apparatus since it is only used one time and then disposed. Care should be
taken to properly dispose of the apparatus.
However, the plastic hose connecting the vacuum pump to the filtration
apparatus should be cleaned, as described above, whenever hose
contamination is suspected.
(5)
Sample Bottles
All sample bottles should be pre-cleaned at the central laboratory (bacteria
bottles, special organic bottles, etc.) and do not require cleaning. The disposable
plastic sample bottles should be used directly as obtained from the laboratory,
but should receive a field rinse with sample water prior to actual sample
collection.
(6)
Water Level Indicator
This device should be cleaned in the office with detergent solution followed by
tap water rinse and a final distilled water rinse. Between wells, the level indicator
should be rinsed with distilled water.
SPECIAL NOTE: Whenever sampling equipment is severely contaminated with organics such as
oil, special decontamination procedures should be followed. A detergent solution should be
used first, if visible contamination remains on the equipment, it may be necessary to use a
solvent rinse. Laboratory grade hexane will usually be sufficient to remove most organics. The
equipment should then be subjected to the normal cleaning procedure once the solvent
residue has been air-dried. Normally, this cleaning procedure will be done in the lab.

Back to top


Bureau of Watershed Management

Back to top


P.O. Box 8555

Back to top


Harrisburg, PA 17105-8555

Back to top


717-787-5259
383-3000-001
An Equal Opportunity Employer
This and related environmental information are available electronically via Internet. For more information, visit us
through the PA PowerPort at http://www.state.pa.us or visit DEP directly at http://www.dep.state.pa.us (directLINK
“Drinking Water Publications”).
www.GreenWorks.tv - A web space dedicated to helping you learn how to protect and improve the
environment. The site features the largest collection of environmental videos available on the
Internet and is produced by the nonprofit
Environmental Fund for Pennsylvania, with financial
support from the Pennsylvania Department of Environmental Protection,
877-PA-GREEN.

Back to top