Mo

onitorin
ng Optim
mization
n and Trrend An
nalysis T
Toolkit

USER
R’S GUID
DE

ESTCP Prroject ER-20 1209

Metthods for Min
nimization an
nd Managem
ment of Variaability in Lon
ng-Term Grooundwater
Monito
oring Resultss
t ti (L )
Concentration (Log)

Randoom
Variab
bility

Longg‐term
trend
d
C

Time

GSII Environmenttal, Inc.

2211 Norfolk, Suitte 1000, Houstton, Texas 770098-4054 713..522.6300
 TABLE OF CONTENTS

1.0 FREQUENTLY ASKED QUESTIONS ......................................................................... 3

2.0 EXECUTIVE SUMMARY .............................................................................................. 6

3.0 BACKGROUND ............................................................................................................... 8

4.0 MONITORING FREQUENCY OPTIMIZATION TOOLKIT: OVERVIEW .......... 9

4.1 Platform and Architecture .................................................................................................. 9
4.2 Basic Structure.................................................................................................................... 9
4.3 General Instructions and Navigation .................................................................................. 9
4.3.1 Instructions ................................................................................................................ 9
4.3.2 Navigation Panel...................................................................................................... 10
4.3.3 Cell Legend ............................................................................................................. 10
4.3.4 Help Buttons ............................................................................................................ 10

5.0 MAIN MENU .................................................................................................................. 11

6.0 MONITORING VARIABILITY TOOL ...................................................................... 13

6.1 Questions .......................................................................................................................... 13
6.2 Input.................................................................................................................................. 13
6.2.1 Data Entry ................................................................................................................ 13
6.2.2 Data Analysis and Results ....................................................................................... 13
6.2.3 Navigation ............................................................................................................... 13
6.3 Results .............................................................................................................................. 14
6.3.1 Data Entry ................................................................................................................ 14
6.3.2 Data Analysis and Results ....................................................................................... 14
6.3.3 Navigation ............................................................................................................... 15
6.3 Well Trends ...................................................................................................................... 14
6.3.1 Data Entry ................................................................................................................ 14
6.3.2 Data Analysis and Results ....................................................................................... 14
6.3.3 Navigation ............................................................................................................... 15

7.0 MONITORING OPTIMIZATION TOOL .................................................................. 17

7.1 Questions ......................................................................................................................... 17
7.2 Inputs ............................................................................................................................... 17
7.2.1 Accuracy .................................................................................................................. 18
7.2.2 Confidence ............................................................................................................... 18
7.3 Advanced Controls .......................................................................................................... 18
7.4 Calculate .......................................................................................................................... 18
7.5 Results ............................................................................................................................. 19
7.6 How Do I Decide between Sampling Options? .............................................................. 19\
7.7 Navigation ....................................................................................................................... 19

8.0 MISCELLANEOUS ....................................................................................................... 21

8.1 Unprotecting Individual Sheets and Viewing Hidden Interface ....................................... 21
8.2 Calculations for Advanced Users ................................................................................... 21

9.0 LIMITATIONS............................................................................................................... 22

10.0 REFERENCES ............................................................................................................... 23

2
1.0 FREQUENTLY ASKED QUESTIONS
a) Who should use this toolkit?

The primary purposes of this toolkit are i) to evaluate the overall source attenuation rate while
accounting for the effect of random variability in apparent attenuation rates in individual
monitoring wells and ii) to determine the amount of groundwater monitoring data required to
characterize the long-term source attenuation rate for sites undergoing long-term monitoring
(i.e., during an MNA process, either when MNA is a stand-alone remedy or when MNA is applied
as a polishing step after active remediation). This tool should be applied at sites where one
purpose of long-term monitoring is to characterize the rate of contaminant concentration over
time.

The toolkit will help the user select an optimal monitoring frequency and evaluate the
relationship between monitoring frequency and time required to characterize the long-term trend.

The user can select the optimal monitoring frequency through this process:

1) Select the primary goal of long-term monitoring: This optimization method is appropriate
when the long-term monitoring data will be used to either
i) demonstrate that constituent concentrations are decreasing OR
ii) estimate when concentrations will decrease to a clean-up goal
2) Identify the timeframe in which the monitoring goal should be met. In other words, how
quickly does the demonstration of decreasing concentrations or estimation of time to clean-up
need to be made?
3) Based on the monitoring goal and the decision timeframe, use the Excel spreadsheet toolkit to
determine the appropriate monitoring frequency.

b) What questions are addressed by the toolkit?

Monitoring Variability

Question 1: When will this site meet the groundwater clean-up goal?

Question 2: Do any wells appear to be attenuation more slowly than the source as a whole?

Monitoring Optimization

Question 1: How much monitoring data do I need to determine a site’s long-term source
attenuation rate with a defined level of accuracy or confidence?

Question 2: What are the trade-offs between monitoring frequency and time required for trend
identification?

c) What data is required for the implementation of this toolkit?

For a site-specific evaluation of the source attenuation rate and short-term variability, the tool
requires historical monitoring data for at least four monitoring wells and at least two years’
worth of monitoring events (recommended at least five monitoring events) for each well. Using

3
more historical data will result in more accurate results. For sites with time dependence in
short-term variability, monitoring records of six years or longer will yield the most accurate
results.

If historical monitoring data is not available, for Monitoring Optimization, the user can enter
estimated values for short-term variability and long-term attenuation rate (see Section 7.1 of this
Guide).

The data evaluation methods used in this tool are based on an assumption of first order source
attenuation. The tool should not be applied to monitoring records where first order attenuation
is clearly not applicable. For example, the tool should not be applied to monitoring records
covering an initial period of plume expansion followed by stabilization and decrease. For sites
where short-duration active remediation has been applied (e.g., chemical oxidation, thermal
treatment), the first order source attenuation concept is applicable to the polishing period after
active remediation is complete. In other words, the method described in this manual is for long
periods where the effects of active remediation or natural attenuation processes are relatively
constant. Monitoring records before- and after-aggressive remediation should not be combined.

d) How do I choose between accuracy and confidence?

The user should choose “Confidence” if a primary objective is collecting enough data to
demonstrate (with a defined level of confidence) that constituent concentrations in the
monitoring well are actually decreasing over time.

The user should choose “Accuracy" if a primary objective is collecting enough data to estimate
the time required to for constituent concentrations in the monitoring well to decline to a
defined numerical standard based on the observed source attenuation rate

e) What are the key results?

Monitoring Variability:

For Question 1: The tool tells you the overall attenuation rate and short-term variability of the
source area, as well as a range of when you can expect the site to obtain the clean-up goal.

For Question 2: The tools tells you if the number of wells with observed “slower” attenuation
rates or with observed “increasing” concentration trends are within the amount expected due to
short-term variability or if they actually represent slower attenuation or increasing trends. These
results allow the user to evaluate whether there are true differences in attenuation rates between
monitoring wells at a site.

Monitoring Optimization:

For Question 1: The tool tells you your options for sampling in the future to determine the
source attenuation rate for a defined level of accuracy or confidence.

For Question 2: The tool tells you the total number of future monitoring events (and estimated
cost) to characterize the long-term attenuation rate. Different results are provided for different
monitoring frequencies (e.g., quarterly, semi-annual, annual).

4
5
2.0 EXECUTIVE SUMMARY
The Monitoring Optimization and Trend Analysis Toolkit optimizes groundwater monitoring
frequency for your site based on site-specific values for the long-term source attenuation rate and
the magnitude of short-term variability. The Toolkit includes the following:

Main Menu:

1) Choose Monitoring Variability Tool if you have historical data for at least four wells and
at least two years’ worth of monitoring events (recommended at least five monitoring
events).
2) Choose Monitoring Optimization Tool if you don’t have historical monitoring data from
your site.

Monitoring Variability Tool:

1) Enter at least two years’ worth of most recent monitoring results (recommended at least
five monitoring events) for 4 to 20 monitoring wells. Use of more monitoring results will
increase the accuracy of the results.
2) After entering data, click “Calculate Results” (may take a while) and then “Review
Results”. The Results page presents a number of pieces of information to answer the
questions for the Monitoring Variability Tool.
a. To answer the first part of Question 1, the page shows the Long-Term
Attenuation Rate and Short-Term Variability for the source area overall.
b. To answer the second part, enter the Representative Source Area Concentration
(default is the maximum Recent Concentration) and Groundwater Clean-Up Goal
you wish to be used to calculate when the goal will be met. The tool then
calculates the clean-up time using the 25th and 75th percentile attenuation rates
from the wells.
c. To answer Question 2, the page shows two summary tables detailing the Glide
Paths and Concentration Trends at the site. Review these tables to determine if
the observed “slower” Glide Paths and “increasing” Concentration Trends are
actually indicative of “slow” Glide Paths or “increasing” Trends or not.
d. Next, review the table that presents, for each well, the Recent Concentration
(average of 5 most recent values), the Long-Term Attenuation Rate, and the
Concentration Trend direction. This will help you decide if the wells with slower
attenuation rates or increasing trends will delay clean-up attainment.
e. After reviewing all of this, choose click “Show Well Trends”.
3) The Well Trends page shows the two summary tables from the Results page, with graphs
below for each well giving a more in depth view of the trends for each well. Review these
tables for more detailed information on the each well’s effect on clean-up attainment.
Then, click on “Go To Optimization”.

Monitoring Optimization Tool:

1) Use the Short-Term Variability value and Long-Term Attenuation Rate calculated from
your site data or enter typical values (See Section 6).
2) Set the Historical Monitoring Frequency as the most common type of sampling frequency
for the data already collected, and enter the number of Years of Historical Monitoring
Data Collected.

6
3)) Choose “C Confidence” if the primaary monitoriing goal is tto demonstraate decreasinng
concentrattions. Choose “Accuracy” if the prrimary monittoring goal is to estimatte
remediatioon timeframess.
4)) Choose yo our desired level of confiddence or accurracy for the llong-term trennd that will bbe
measured by future monitoring.
m “Medium” confidence or accuracyy represents a
reasonablee minimum am mount of dataa required to ssupport decisiion making.
5)) If you wou uld like moree exact contro
ol over the coonfidence or aaccuracy, cheeck “Advanceed
Controls” and enter thee specific leveel of confidennce or accuraccy and the Peercent of Wellls
that will meet
m that param meter.
6)) Click “Callculate and Show Results””. The tool w will calculate a set of seveen options that
provide different combiinations of monitoring
m freqquency and tiime required to characterizze
the long-teerm attenuatio
on rate with th
he specified l evel of accurracy or confiddence.

Enter Concentration vs. Tim me Choosse a

Monitoring Well
W Data for a desired
d CON FIDENCE Click
Param meter LEVEL (Mediu um or High) “CA
ALCULATE
(or enter typiical site-wide or a AN
ND SHOW
variability andd attenuation des
sired ACCUR RACY LEVEL RE
ESULTS”
rate
es) (Medium oor High)

Fig
gure E1: Sam
mple Monitorring Optimizzation Tool R
Results table

7
3.0 BACKGROUND
The Monitoring Optimization and Trend Analysis Toolkit provides a method to distinguish
between random variations in attenuation rates and true spatial differences in remedy
performance and provides recommendations for optimized groundwater monitoring frequency
based on a site-specific evaluation of the magnitude of short-term variability and the long-term
attenuation rate. The Toolkit uses site-specific data to determine clean-up horizons and possible
problem wells and simulated monitoring datasets that match the short-term variability and long-
term source attenuation rate characteristics of your site in order to determine the number of
monitoring events required to characterize the long-term trend with a defined level of accuracy or
statistical confidence.

This User’s Guide focuses on the practical aspects of using the Toolkit. The development of the
Toolkit is documented in the report “Task 2 Report: Methods for Minimization and Management
of Variability in Long-Term Groundwater Monitoring Results, ESTCP Project ER-201209”
(McHugh et al., 2015).

8
4.0 MONITORING OPTIMIZATION AND TREND ANALYSIS
TOOLKIT: OVERVIEW
4.1 Platform and Architecture

The toolkit is built on the Microsoft Excel platform with a user-friendly interface. The tool opens
on the Main Menu page, which contains a flow chart that directs the user to either the Monitoring
Variability Tool or the Monitoring Optimization Tool.

The Monitoring Variability Tool uses historical groundwater monitoring data from individual
wells in order to calculate the site’s overall short-term variability and long-term attenuation rate.
Subsequently, the user is directed to the Monitoring Optimization Tool.

The Monitoring Optimization Tool uses the short-term variability and long-term attenuation rate
to determine options for sampling frequencies needed to capture the true attenuation rates of the
groundwater constituents at the site.

4.2 Basic Structure

The basic structure of the Toolkit is as follows. Each square indicates what the user will find at
each step, as well as the name of the page it is found on in parenthesis.

Monitoring
Main Menu Optimization Tool
(Results: Optimized Sampling
Monitoring Variability Frequency)
Tool
(Results: Short-Term Variability,
Long-Term Attenuation Rate,
and Clean-Up Attainment Info.)

4.3 General Instructions and Navigation

Each screen has four main components that help in using the tool and navigating between
screens:
i) Instructions
ii) Navigation Panel
iii) Cell Legend
iv) Help Buttons

4.3.1 Instructions

Instruction boxes are located in the top left corner of all screens and contain specific steps or
instructions the user should follow in order to operate the page effectively. Each section of the
page also has the particular Instruction number (i.e., 1,2,3,4) next to the section title.

9
 Figure
F 1: Exa
ample Instru
uction Box

4.3.2 Navigation
n Panel

Naviggation panels are located ini the top rigght corner of each page annd contain buuttons that wiill
advan
nce the user to different screens
s (Figu
ure 2). They also containn buttons to ggo to previouus
pagess, as well as to
o review or ed
dit inputs.

Fig
gure 2: Inputt page Navig ation Panel

4.3.3 Cell Legen

nd

d as either an Input or Calcculated cell (Figure 3). Ussers are able tto
Cells of interest arre categorized
enter project or site-specific infformation into
o the Input ceells (colored w
white). Somee of these Inpuut
cells will
w have dro op-down lists, while otherrs will requiree manual enttry of data. C
Calculated cellls
(tan colored)
c contaain calculated values that th
he user is unaable to changee.

Figure 3: Cell Legeend

4.3.4 Help Butto
ons

Help buttons
b are av
vailable to thee right of man
ny of the Tooolkit’s cells, tiitles, and butttons (as show
wn
below
w). The user can
c click thesse to get moree information about the cloosest topic.

Figure 4: Help Buttton

10
5.0 MAIN MENU
M
The Main
M Menu page contain ns a flowchaart (Figure 55) that leadss the user thhrough several
questiions, and direects them to th
he appropriatee Tool.

Start
Here

Endd
Herre

Fiigure 5: Toollkit Outline F

Flowchart

The flowchart
fl quesstions are as follows:
f

(1) Do
D you have historical
h welll data for your site?

Historrical site dataa can be used in the Monito oring Variabillity Tool to dderive site-speecific estimatees
of shoort-term variaability and lo ong-term atten nuation ratess. Data for aat least 4 wellls and up to a
maximmum of 20 wellsw with at least two yeaars of recent sampling eveents (recomm mended at leaast
five sampling even nts) for each well
w are requiired to run thiis tool. Chooose wells wheere the primarry
contamminant of con ncern is deteccted for at leasst 80% of moonitoring evennts.

If Yes: Pro
oceed to Quesstion 2 below..

If No: Go to
t Monitoring
g Optimizatio
on Tool

(2) Iss data “cleanup” required

d?

The Monitoring
M Vaariability Too
ol requires datta in the folloowing format::

 Dates in ch
hronological order
o

11
  Dates in MM/DD/YYYY format
 Concentrations as numerical values (i.e., no flags)
 Non-detects replaced with the appropriate numerical values
 All concentration data must be in the same units (either mg/L or ug/L)

As seen in the table below, typical historical groundwater monitoring data can be available in
many formats and styles, and may require “clean-up” before being entered into the toolkit.
Various scenarios of original data, and the final required format are presented below.

Table 1: Examples of Clean-Up Scenarios for Historical Monitoring Data

Original “Cleaned-Up” Format
Monitoring Data Required for Tool
Scenario Notes
Sample TCE Sample TCE
Date Conc. Date Conc.
June 10, Dates in MM/DD/YYYY
1 250 06/10/2001 250
2001 format
Non-detects replaced with
2 3/15/1995 ND 3/15/1995 1
numerical values
5 Non-detects replaced with
3 8/12/2004 <5 8/12/2004
(or 2.5) numerical values
Dates in MM/DD/YYYY
4 14/1/2003 125 1/14/2003 125
format
Concentrations as numerical
5 3/14/1989 45 J 3/14/1989 45
values

If Yes: Go to a Data Management Tool to clean-up your data (e.g., MAROS).

A tool such as MAROS can help the user manage data and meet the requirements listed
above. After completion of this step, navigate back to the Monitoring Variability Tool.

If No: Go to Monitoring Variability Tool

12
6.0 MONITORING VARIABILITY TOOL
6.1 Questions

The Monitoring Variability Tool answers two questions:

Question 1: When will this site meet the groundwater clean-up goal?

Question 2: Do any individual wells appear to be attenuating more slowly than the source as a
whole?

Going through the Tool will provide the answers to these questions.

6.2 Input

6.2.1 Data Entry

On the Input page the user should enter the following information:

 Site Info
 Site Name
 Contaminant of Concern
 Concentration Units (Note: This is a dropdown menu of either mg/L or ug/L)

Historical groundwater monitoring data should be entered next as follows:

 Well Name: This is text (i.e., MW-2)

 Date: Arranged chronologically in MM/DD/YYYY format.
 Concentration: Numerical values (e.g., no text strings) in consistent units (mg/L or µg/L).

The Toolkit allows for the data entry and analysis of between 4 and 20 wells, with data from at
least two years of sampling events (and a recommended minimum of five events). If more than 25
sampling events exist at a site, it is recommended to select the 25 most recent sampling events.

6.2.2 Data Analysis and Results

No results are displayed on the Input page.

6.2.3 Navigation

The navigation options are as follows:

 Go Back to Main Menu - Takes the user back to the Main Menu page
 Clear Data - Clears all data input thus far
 Calculate Results - Calculates both individual well results and overall (i.e., for the entire
site) results, which can be seen on the Results page. Results must be re-calculated any
time the data Input page is edited.
 Review Results - Takes the user to the Results page

13
6.3 Results

6.3.1 Data Entry

On the Results page, the user should enter the following information:

 Representative Source Area Concentration: Numerical value that defaults to the

maximum well Recent Concentration
 Groundwater Clean-Up Goal: Numerical value

6.3.2 Data Analysis and Results

The first data are answers to the first part of Question 1:

Question 1a: What is the overall attenuation rate for the key contaminant at my site?

These answers are:

 Long-Term Attenuation Rate: The overall attenuation rate is the median of the individual
well attenuation rates. A positive value indicates a decreasing concentration trend.
 Short-Term Variability: For the overall dataset, the short-term variability is estimated as
the standard deviation of the natural log of the normalized residual from the attenuation
rate regression analysis. However, this standard deviation value is optimized to reduce
the influence of outlier values in the dataset.

The Tool then uses the previously entered information to answer the second part of Question 1:

Question 1b: When will this site meet the groundwater clean-up goal?

The Tool gives a range of years that the user can expect the groundwater clean-up goal to be
achieved, based on the 25th and 75th percentile attenuation rates from the wells.

From here, the Tool shows two summary tables that answer Question 2. The first – Attenuation
Rate Summary – shows how many wells can be expected to have Faster, Within Range, or Slower
Glide Paths (based of Short-Term Variability) and how many wells actually do fall into those
categories. The second – Concentration Trend Summary – shows the same information for
Increasing and Decreasing Concentration Trends. Based on these results, the Tool provides two
“Yes” or “No” answers to Question 2.

Below these tables, data for individual wells is presented, which gives a more detailed look at
these wells. The Tool gives the user these instructions:

What to do next: If the answer to Question 2 is “Yes”, evaluate the individual wells with slower
attenuation rates or increasing concentrations. Consider whether these wells will delay attainment
of the overall site clean-up goals.

In order to do this, the Tool presents these data for each well:

 Recent Concentration: The average of the most recent five concentration values.
 Long-Term Attenuation Rate

14
 o Rate: For individual wells, the long-term attenuation rate is estimated using least
squares regression and a first-order (exponential) decay model. A positive value
indicates a decreasing concentration trend (Newell et al., 2002).
o Glide Path: Identifies whether the attenuation rate glide path for each individual
well is slower or faster than the overall rate for the site. If the glide path is
“Within Range”, then the difference between the attenuation rate for the well and
the overall attenuation rate for the site is consistent with the effects of short-term
monitoring variability.
o Accuracy: The attenuation rate accuracy is based on the 95% confidence interval
for the attenuation rate as follows.
 High Accuracy: confidence interval <±0.05 or <±25% of
attenuation rate;
 Medium Accuracy: confidence interval <±0.1 or <±50% of
attenuation rate;
 Low Accuracy: confidence interval >0.1 and >50% of
attenuation rate
When the accuracy for a well is “Low”, the user should be cautious in using the
observed attenuation rate for predicting future contaminant concentrations.
 Concentration Trend
o Direction: Whether or not the long-term trend shows increasing or decreasing
constituent concentrations
o Confidence: The statistical confidence that the true attenuation rate is not zero.
This is based on the p-value of the attenuation rate at this well.
 High Confidence: p<0.05;
 Medium Confidence: 0.05<p<0.1;
 Low Confidence: p>0.1

6.3.3 Navigation What is p-value?

The navigation options are as follows: The p-value is the probability that
the apparent pattern in the data
(e.g., a decreasing concentration
 Edit Well Data - Takes the user back to the
trend) is due to random chance. In
Input page
other words, when the p-value is
 Show Well Trends - Takes the user to the Well
less than 0.1, there is a greater than
Trends page
90% chance that the observed
decreasing concentration trend is
6.4 Well Trends real.
6.4.1 Data Entry

No inputs are entered on the Well Trends page.

6.4.2 Data Analysis and Results

On this page, the Tool shows the same two summary tables as on the Results page that answer
Question 2. The first – Attenuation Rate Summary – shows how many wells can be expected to
have Faster, Within Range, or Slower Glide Paths (based of Short-Term Variability) and how
many wells actually do fall into those categories. The second – Concentration Trend Summary –
shows the same information for Increasing and Decreasing Concentration Trends. Based on these
results, the Tool provides two “Yes” or “No” answers to Question 2.

15
Also as in the Results page, below these tables data for individual wells is presented, which gives
a more detailed look at these wells. The Tool gives the user these instructions:

What to do next: If the answer to Question 2 is “Yes”, evaluate the individual wells with slower
attenuation rates or increasing concentrations. Consider whether these wells will delay attainment
of the overall site clean-up goals.

On this page, these data are presented as graphs that show, for each well, the Well Trend in red
and the Site (Median), Upper 95%, and Lower 95% Trends in black. This shows the user how the
individual well related to the rest of the site, to determine if it will delay attainment.

Below each graph, the Tool shows whether the well has an Increasing or Decreasing Trend and
whether the Attenuation Rate is Slower, Faster, or Within Range of the Site 95% Confidence
Trends.

A key is shown at the top of this page to explain this to the user.

6.4.3 Navigation

The navigation options are as follows:

 View Results - Takes the user back to the Results page

 Go To Optimization - Takes the user to the Optimization page

16
7.0 MONITORING OPTIMIZATION TOOL
7.1 Questions

The Monitoring Optimization Tool answers two questions:

Question 1: How much monitoring data do I need to determine a site’s long-term source
attenuation rate with a defined level of accuracy or confidence?

Question 2: What are the trade-off between monitoring frequency and time required for trend
identification?

Going through the Tool will provide the answers to these questions.

7.2 Inputs

The Monitoring Optimization Tool page requires values for: i) Short-Term Variability and ii)
Long-Term Attenuation Rate. As such, if the user provided historical monitoring data and ran the
Monitoring Variability Tool, the two parameters will have been calculated and will appear on this
page. If no historical monitoring data is available, the user should input typical values for short-
term variability and long-term attenuation rates. Typical values for these parameters are as
follows:

Table 2: Typical Values for Short-Term Variability and Long-Term Attenuation Rate
Constituent Short-Term Variability Long-Term Attenuation
Rate
Petroleum Hydrocarbons 1.0 to 2.0 0.35 yr-1
Chlorinated Solvents 0.5 to 1.0 0.12 yr-1
For characterization of short-term variability and long-term attenuation rates at 20 sites, see Section 4.1 of Task 2
Report, Methods for Minimization and Management of Variability in Long-Term Groundwater Monitoring Results,
ESTCP Project ER-201209.

Information on the historical groundwater monitoring data should be entered next, if applicable:

 Historical Monitoring Frequency: The sampling frequency for monitoring data already
collected at the site. The user can select from pre-defined options, including: quarterly,
semiannually, etc. If different wells had various monitoring frequencies, please select the
most common type.

 Years of Historical Monitoring Data Collected: The number of years of monitoring data
already collected. If different wells have been monitoring for different time intervals,
choose a value representative of a typical well.

The user must then calibrate the Tool so the results will be calculated to the desired detail. As
such, the user must choose between “Accuracy” and “Confidence”, at a “Medium” or “High”
level.

Choose “Confidence” if a primary goal of monitoring is to demonstrate that contaminant

concentrations are actually decreasing.

Choose “Accuracy” if a primary goal of monitoring is to estimate the time required for
contaminant concentrations to decrease to a numerical clean-up goal.

17
These variables are described below.

7.2.1 Accuracy

The Optimization Tool will determine how many monitoring events are required to determine the
long-term attenuation rate with a level of accuracy or a confidence specified by the user. The
number of monitoring events is calculated so that 80% of the resulting monitoring records will
meet the level of accuracy or confidence specified by the user. The accuracy is based on the 95%
confidence interval for the attenuation rate.

Two options are available for selection: i) “High” (i.e., the value must be below either 0.05 or
0.25 times the Long-Term Attenuation Rate, whichever is larger); and ii) “Medium” (i.e., it must
be below 0.1 or 0.5 times the Long-Term Attenuation Rate, whichever is larger). Selecting
“Medium” will result a reasonable minimum amount of monitoring data needed to support
decision-making. The user should recognize that, after this minimum is obtained, accuracy will
continue to increase over time if monitoring is continued.

7.2.2 Confidence

The confidence is the statistical confidence that the attenuation rate is not zero, based on the p-
value for the first-order attenuation rate using least square regression.

Two options are available for selection: i) “High” (i.e., the value must be below 0.05); and ii)
“Medium” (i.e., value must be below 0.1). Selecting “Medium” will result a reasonable
minimum amount of monitoring data needed to support decision-making. The user should
recognize that, after this minimum is obtained, confidence will continue to increase over time if
monitoring is continued.

7.3 Advanced Controls

The advanced controls give the user the ability to input a specific value for the level of accuracy
or level of confidence, rather than “High” or “Medium”. The user can also specify the percentage
of monitoring records that will meet the specified level of accuracy or confidence.

 Accuracy: The confidence interval for the attenuation rate. If the user enters 0.1, then the
maximum acceptable confidence interval will be ±0.1. For advanced controls, the user
also specifies the percentage of well monitoring records that must achieve this maximum
confidence interval.

 Confidence: The required maximum p-value for the attenuation rate. For example, a p-
value of 0.1 corresponds to a 90% confidence that the attenuation rate is not equal to
zero. For advanced controls, the user also specifies the percentage of well monitoring
records than must achieve this maximum p-value.

 Percent of Wells: The percentage of monitoring well records that will meet requirements
for Accuracy or Confidence. For regular controls, this value is 80%. For advanced
controls, the user specifies the value.

7.4 Calculate

18
Once the inputs are set, the user should click the “Calculate and Show Results” button to
calculate and populate the “Results” table as described in Section 7.4 below.

7.5 Results

As previously described, the Tool helps answers two main questions as follows:

Question 1: How much monitoring data do I need to determine a site’s long-term source
attenuation rate with a defined level of accuracy or confidence?

As the Answer indicates, different sampling options, from weekly to every 5 years, are presented
in order to understand “how much additional monitoring data you need to accurately characterize
your long term trend”. A sample table is shown below.

Figure 6: Example Sampling Requirements to Characterize the Long-Term Attenuation

Rate with High Confidence

This table presents the following values:

 Sample Frequency - The length of time required to sample at different frequencies, from
weekly (Option 1) to every 5 years (Option 7).

 Total Sampling Events - Number of sampling events that will occur at the specific
sampling frequency for the total time period.

 Cost Per Well - The approximate cost of this sampling program per well, in thousands of
dollars. This value per well is based on a sample cost of $1,500 per event. This assumed
cost includes labor and materials cost for sample collection, laboratory analysis, data
management, and reporting.

Question 2: What are the trade-off between monitoring frequency and time required for trend
identification?

As the Answer indicates, the table also responds to Question 2, where “[a]ll options presented are
equivalent”. These trade-offs are explained below.

As shown in the “Sample Frequency” column in Figure 6, there are trade-offs between the future
sampling frequency and the total time of monitoring required. For instance, if the user decides to
sample once a month, it would take 2.7 years to know a site’s long term source attenuation rate

19
with a defined confidence or accuracy. Comparatively, if the user were to sample once a year, the
total monitoring time period would be 6.5 years to attain the same information.

7.6 How Do I Decide between Sampling Options?

Each of the different sampling options yields a monitoring dataset that can be used to characterize
the long-term attenuation rate with the same level of confidence and accuracy. The final choice
of monitoring frequency should be based on considerations of time and cost, answering the
question:

What is the most cost effective monitoring option that will allow me to make decisions within the
required timeframe?

Take, for example, a site where MNA has been selected as the remedy with the performance
requirement that contaminant concentrations are demonstrated to be, in fact, decreasing at the
time of the first five-year review. For this site, if the Monitoring Optimization Tool provided the
monitoring shown in Figure 6, then the user would choose to monitor semi-annually because this
five years of monitoring at this frequency would serve to document the long-term attenuation rate
with high confidence. In contrast, 6.5 years of annual monitoring would be required to determine
the long-term trend with high confidence, a time period that exceeds the requirement for
demonstrating decreasing concentrations within five years.

7.7 Navigation

The navigation options on this page are as follows:

 View Well Trends - Takes the user back to the Well Trends page, if the user went through
the Monitoring Variability Tool
 Calculate and Show Results - Calculates the monitoring frequency results and fills out the
results table
 Start New Project - Asks the user for confirmation, then clears all current data and brings
the user to the Main Menu page
 Save & Exit - Allows the user to name and save the file, then exits the Toolkit

20
8.0 MISCELLANEOUS
8.1 Unprotecting Individual Sheets and Viewing Hidden Interface

Individual spreadsheets and cells within the tool have been protected, except for any input cells.
In order to generally unprotect the sheet, go to Review on the toolbar, and click Unprotect Sheet.

In order to view any of the hidden interface, go to File>Options>Advanced or go to View on the

toolbar and select what you would like to view (eg. sheet tabs, formula bar, or headings). In order
to view hidden cells, select cells near the hidden cells and go to Home on the toolbar and click
Format>Hide & Unhide and select what you would like to unhide.

8.2 Calculations for Advanced Users

Several tabs with background calculations have been hidden for simplicity. These can be seen by
going to Home on the toolbar and clicking Format>Hide & Unhide>Unhide Sheet… and
selecting which sheet you would like to unhide.. The hidden sheets and a brief description of their
contents are as follows:

1. Well Calculations - Calculates the Individual Well Results shown on the Results page.

2. Residuals Calculations - Used to correct the Short-Term Variability to match a log-

normal distribution.

3. Random Data Calculations - Calculates the simulated well data that are checked against
the accuracy or confidence parameters.

4. Trend Calculations – Calculates the trend information for the Well Trends page.

5. Referenced Values - Contains values that are referenced in the Toolkit (e.g. in
dropdowns).

21
9.0 LIMITATIONS
This Excel-based tool has been developed by GSI Environmental under Department of Defense
contract W912HQ-12-C-0055. Neither GSI nor any of GSI’s employees, subcontractors,
consultants, or other assigns make any warranty or representation, either express or implied, with
respect to the accuracy, completeness, or utility of the information contained herein, or assume
any liability or responsibility for any use, or the results of such use, of any information or process
disclosed in this publication, or represent that its use would not infringe upon privately owned
rights.

Users of this tool should not rely exclusively on the information contained in this document.
Sound business, scientific, engineering, and safety judgment should be used in employing the
information contained herein.

Where applicable, authorities having jurisdiction should be consulted.

GSI is not undertaking to meet the duties of employers, manufacturers, or suppliers to warn and
properly train and equip their employees, and others exposed, concerning health and safety risks
and precautions, nor undertaking their obligations to comply with authorities having jurisdiction.

22
10.0 REFERENCES
Newell, C.J., Rifai, H.S., Wilson, J.T., Connor, J.A., Aziz, J.A., Suarez, M.P. 2002. Calculation
and Use of First-Order Rate Constants for Monitored Natural Attenuation Studies. USEPA
Ground Water Issue. http://nepis.epa.gov/Adobe/PDF/10004674.pdf

McHugh, T., Kulkarni, P., Beckley, L., Newell, C., Strasters, B. Draft Task 2 Report: Methods
for Minimization and Management of Variability in Long-Term Groundwater Monitoring
Results, ESTCP Project ER-201209. GSI Environmental Inc. July 2014.

23