|
Obtaining meaningful, quality data is not an easy task. Rather,
good quality data results from hard work and often from considerable
expenditure of time and money. Many times, it costs as much or more
to produce poor quality data, especially in the long run, when data
must be recollected or extra time is necessary to interpret the
results. Poor data collection can also result in misguided reclamation
efforts. This can lead to an under- or over-designed reclamation
solution for the problem at hand. In either case, time and money
are wasted not to mention the environmental impacts that may be
caused from a poor design solution.
Collection of good quality data takes considerable planning to
ensure accurate, consistent results. Planning before conducting
experiments is now generally recognized as an emerging scientific
discipline. Out of this emerging discipline, the Data Quality Objectives
(DQO) process was developed as a planning tool. The DQO process
helps determine when enough data of sufficient quality has been
collected to enable accurate decision-making. In the case of a mine
reclamation project, the DQO process helps reclamationists determine
which reclamation alternatives would be most effective with the
least cost. An analysis of reclamation alternatives is often conducted
using the following seven steps of the DQO process:
- State the problem
- Identify the decision
- Identify inputs
- Define boundaries
- Develop a decision rule
- Specify limits on decision errors
- Optimize the design
This process is often used for developing a sampling and analysis
plan (SAP) for each site. For an example of how the DQO process
is used in a SAP, click here.
Step 1: State the Problem
The identification and delineation of the problem to be investigated
is a critical first step. Too often this step is overlooked or taken
too lightly. The desire to jump in and solve the problem can result
in the initiation of a project without really understanding the
problem(s). The first step includes developing and refining a map
of the site that shows the locations of sources of contamination;
the types and expected concentrations of contaminants; possible
movement of the contaminants in the soil, water or air; and the
location of humans, animals or sensitive ecological environments.
This initial investigation should not be an intensive sampling project;
it should include only enough sampling to establish the problem.
For an abandoned mine site investigation, the initial sampling may
include sampling of all areas suspected to be producing acid mine
drainage.
Budgets for sampling and analysis over the course of the project
should be determined during this step as well as the various specialists
that will be involved with the project (i.e. environmental scientists,
hydrogeologists, toxicologists, lab technicians, etc).
Step 2: Identify the decision
After data collection, the problem can be stated and alternative
actions can be proposed. Often, the problem will be stated in terms
of a number of questions which must be answered. Different solutions
to the problem are presented depending on the answers to the questions.
This step allows investigators to organize their decision making
process which can save time and money in the long run.
Example:
An abandoned mine site is proposed for reclamation. From step 1
it was determined that numerous tailings piles on the site are producing
acid mine drainage. Questions that may be asked include:
- Is the acid mine drainage leaving the site?
- Is it entering downstream waterbodies?
- What are the concentrations of heavy metals and the pH of the
acid mine drainage leaving the site?
- What are the concentrations of heavy metals and the pH of the
acid mine drainage downstream from the site?
- At any time are the concentrations above the action levels for
protection of humans, animals and the environment?
Depending on the answers to these questions, different action alternatives
will be taken. The level of remediation at the site will be based
on the severity of its environmental impact.
Step 3: Identify Inputs
The next step is to determine how the questions from step 2 will
be answered. In other words, what sampling and analysis program
must be conducted in order to determine which course of action will
be taken to solve the problem. The sampling and analysis program
will define what type of samples will be taken (i.e. tailings, waste
rock, soil, water), how many samples will be taken, where the samples
will be taken from, and the appropriate methods and techniques for
collecting and analyzing the samples. The type, number, and location
of the samples will depend on the nature of the problem under investigation.
The number of samples necessary will also depend on the capability
of the measurement program to provide data of adequate quality.
Step 4: Define Boundaries
In this step, time and geographic boundaries are defined. How much
time is available for reclamation of the given site? What is the
area of influence that will be included in the reclamation solution?
These boundaries are a reality of every project. Time and money
is not of unlimited availability and usually has a considerable
influence on the ultimate decision for the final design solution.
Step 5: Develop a Decision Rule
This step uses all the information gathered from the previous four
steps in the DQO process in order to make good decisions. This step
defines what environmental impacts may be caused by the problem
and defines whether people, animals, or plants, etc. may be influenced.
This step also defines the maximum contaminant concentrations that
will be allowed for a given solution. If concentration levels exceed
the given limit for a certain design solution, then a more intense
design solution will be implemented. Detection limits for the analysis
of different contaminants will also be defined in this step. Detection
limits will always have to be lower than the maximum contaminant
concentration levels defined in this step for the different design
solutions
Step 6: Specify Limits on Decision Errors
Errors are inevitable when conducting a sampling and analysis project.
Human error, to some degree, occurs throughout the process. Error
in automated techniques for analyzing samples is also inevitable.
It is extremely important to develop sampling and analysis protocols
that minimize errors as much as possible in order to obtain quality
data that accurately defines the problem at hand. The use of clean
sampling gloves, sampling devices, and sampling containers are important
in order to minimize contamination of the samples. Ensuring that
the sampling plan is unbiased also minimizes error. Randomized sampling
is used to minimize bias. Proper calibration of analytical instruments,
proper handling of samples in the lab, and carefully following laboratory
procedures are of vital importance. Duplicate samples should also
be analyzed to verify that the analysis methods are accurate.
The steps necessary to minimize errors and produce good quality
data have evolved into quality assurance/quality control (QA/QC)
programs that give guidelines for minimizing error in sampling and
analysis projects. Each sampling and analysis project should implement
a QA/QC program that includes the following guidelines. Quality
assurance (QA) is a set of operating principles that are designed
to produce data of know and defendable quality. A QA program includes
the organization and procedures such as staff organization and responsibility,
sample control and documentation procedures, training requirements,
equipment maintenance procedures, calibration procedures, quality
control activities of core staff (internal), validating and reporting.
Quality control (QC) may be either internal and/or external (i.e.
laboratory personnel, subcontractors). QC programs should include:
certification of operator competence; analysis of externally supplied
standards to ensure concentrations are what they should be; analysis
of reagent blanks to determine if interferences are present because
of glassware, reagents, or equipment; calibration with standards;
and analysis of duplicates (commonly 10% replication of the total
sample number). QA/QC is a vital component to any sampling and analysis
program but can add 15% to 20% to lab analysis costs (MEND,
2001).
Once all effort has been made to minimize error, all error has
not eliminated, it has only been reduced as much as possible. From
here, it is important to understand the error is still associated
with the sampling and analysis plan. The science of statistics is
often used as an important and necessary mathematical tool for the
interpretation of measurement results. Statistics can be a powerful
tool if the interpreter of the results thoroughly understands the
basic principles upon which the science and practice of statistics
are based. Many times, statistics are inappropriately applied and
therefore, results are of little use.
Based on the variability of the measurement process and of the
samples to be investigated, statistics provide guidance on the number
of measurements that should be made to obtain a desired level of
confidence in the data. Statistics can be helpful in determining
over what geographic area samples should be taken. They are also
helpful in determining how many duplicate analytical tests should
be conducted to ensure that test results are accurate enough for
decision making. Statistics are also very helpful in determining
the quality of the data which, in turn, allows the data interpreter
to make correct decisions concerning reclamation plans.
It should be remembered that statistical techniques are only tools
and should be used for enlightened guidance and certainly not for
blind direction when making decisions. When statistical results
conflict with results that the interpreter would intuitively expect,
one should stop and take careful consideration. Was one's intuition
wrong or were wrong statistical tools used? For more information
on basic statistical concepts and statistical techniques used for
data analysis, see Taylor
(1990).
Step 7: Optimize the Design
This steps combines all the information gathered in the previous
six steps and uses this information to decide what design solution
would be most effective while making the best use of time and money.
The design solution should describe the methods that will be used
for sampling and analysis, the type of samples that will be collected,
the sample size, and the number of laboratory tests that will be
for each sample. A description of the pros and cons of each different
alternative action will be discussed and along with reasons for
selection of the best possible design solution. By using the DQO
process, specific steps help the reclamationist make streamlined
decisions in a timely and cost effective manor.
For more information on the DQO process, see Boguski
and Data
Quality Objectives Workshop. Also check out the following websites:
Problem | Compliance
| Health & Safety | Sampling
| Analytical | Data
Quality
Site Assessment | Prediction
| Construction | GIS
| Monitoring & Assessment
|
