&EPA
                   68-01-6425
 APPROACH TO MIDCOURSE EVALUATION
                    OF
    MECHANICAL INTEGRITY TESTING
                  SUBMITTED TO
                 DR. JENTAI YANG
              OFFICE OF DRINKING WATER
          U.S. ENVIRONMENTAL PROTECTION AGENCY
                  JANUARY 1983
                          BOOZ ALLEN & HAMILTON, INC.
                           UNDER THE DIRECTION OF
                           GERAGHTY & MILLER, INC.

-------
&EPA
                   68-01-6425
 APPROACH TO MIDCOURSE EVALUATION
                    OF
    MECHANICAL INTEGRITY TESTING
                  SUBMITTED TO
                 DR. JENTAI YANG
              OFFICE OF DRINKING WATER
          U.S. ENVIRONMENTAL PROTECTION AGENCY
                  JANUARY 1983
                          BOOZ-ALLEN & HAMILTON, INC.
                           UNDER THE DIRECTION OF
                           GERAGHTY & MILLER, INC.

-------
                          ACKNOWLEDGEMENT
     This report was prepared by Booz, Allen & Hamilton, Inc., under
the direction of Geraghty & Miller, Inc.  The Booz, Allen project
manager was Dr. Joanne Wyman.  She was assisted by Mr. Walter Mardis
of Booz, Allen.  The Geraghty & Miller project manager was Mr.
William Thompson, and Mr. Roger Anzzolin was the EPA task monitor.

-------
               TABLE   OF   CONTENTS
                                                        Page
                                                       Number
1.   OVERVIEW OF MECHANICAL INTEGRITY                     1
     DEMONSTRATION REQUIREMENT

2.   OBJECTIVES OF THE MIT MID-COURSE EVALUATION          2

3.   APPROACH TO THE CONDUCT OF THE MIT                   3
     MID-COURSE REVIEW

-------
            APPROACH TO MIDCOURSE EVALUATION  OF
                MECHANICAL INTEGRITY TESTING
     On June 24, 1980 the  U.S.  Environmental Protection
Agency  (EPA)  issued final  regulations for  the Underground
Injection  Control  (UIC) program.  These rules require a
more extensive midcourse program evaluation than EPA had
set forth  in the proposed  rules.   Accordingly,  EPA's
Office of  Drinking Water  (ODW)  asked Geraghty and Miller,Inc.
(G&M) and  Booz, Allen & Hamilton, Inc.  (BA&H)  to expand
our initial analysis on the conduct of the midcourse eval-
uation to  include the mechanical integrity testing  (MIT)
requirement for Class II wells.^  In response to ODW1s re-
quest, we  developed baseline information  for the MIT re-
view similar to that we prepared for the AOR midcourse
evaluation.3  This information is summarized in Exhibit  1
and the accompanying narrative.

1.   OVERVIEW OF THE MECHANICAL INTEGRITY  DEMONSTRATION
     REQUIREMENT

     The principal purpose of requiring well operators to
demonstrate mechanical integrity is to ensure:

           The absence of  significant leaks in the well
           casing, tubing,  or packer

           The absence of  significant fluid movement into
           an underground  source of drinking water through
           vertical channels adjacent to  the injection well
           bore.
      The rules EPA proposed in April 1979 required a midcourse evaluation only
      of the costs and benefits of the Area of Review requirement
      for Class II wells.  The final rules, however, extend the midcourse
      evaluation to the AOR for Class I and III and to the mechanical
      integrity testing program for Class II.

      Prior to promulgating final rules, EPA asked us to assist in de-
      signing the midcourse evaluation of AOR.  After issuance of the
      final rules, EPA asked us to extend our initial analysis of
      alternative study methodologies to mechanical integrity testing
      in lieu of preparation of an AOR midcourse evaluation study manual.

      See Geraghty & Miller, Inc. and Booz, Allen & Hamilton, Inc.,
      Development of Midcourse Evaluation Methodology  ( Task I Report:
      Evaluation of Alternatives), June 17, 1980 pp. 1-5.

-------
In its April 1979 version of the proposed rules,  EPA called
for the use of eight  tests to determine the likelihood of
leaks and the use of either well records or tests to determine
the existence of fluid movement.  The requirements covered
Class Ir II, and III well operations.

     In the preamble to the final regulations,  EPA indicates
that it now has serious doubts about the reliability of well
records as indicators of the absence of fluid movement.
Accordingly, for Class I and III operations the Agency now
requires the use of formal tests.  For Class II operations,
however, because of the large number of wells and the Safe
Drinking Water Act  (SDWA) requirement that regulations not
interfere with oil and gas production, EPA continues to allow
the use of well records to demonstrate mechanical integrity.
At the same time, under a modified Section 146.08, EPA requires
a sample of Class II wells to conduct mechanical integrity
tests.  Wells in the sample are to be ones which initially use .
well records in place of temperature or noise logs to
determine the likelihood of fluid movement.

2.   OBJECTIVES OF THE MIT MIDCOURSE 'EVALUATION

     The purpose and objectives of the AOR and MIT midcourse
reviews differ markedly.  The purpose of the AOR review is
to ascertain whether EPA regions and the states have imple-
mented the requirements effectively.  By contrast, the pur-
pose of the MIT review is to establish a mechanical integrity
demonstration procedure which is more economical and less
time consuming than universal formal testing, while still
guaranteeing a high degree of accuracy.

     Establishment of a screening procedure would be one
possible approach.  EPA believes that it may be possible
to identify certain well characteristics (such as depth, age
and type of cementing) which are reliable indicators of
well integrity.  For each well, well records will indicate
whether the well has the characteristics indicative of mechani-
cal integrity.  Wells not having those characteristics will
have to perform formal tests.

     In order to determine the feasibility of this approach,
EPA intends to compare the mechanical integrity demonstration
results from the well record reviews with those from the
formal tests.  Key questions the Agency needs to consider
in analyzing the data are:

-------
          How accurate  are  Class II well records in identi-
          fying mechanical  integrity problems?

          What savings  to well owners/operators result from
          the use of well records instead of noise and tem-
          perature  logs?

          What are  the  administrative impacts on state
          regulatory agencies  of substituting well records
          for log data?

In addition, the Agency is  interested in examining the
applicability of any Class  II  screening procedure to Classes
I and III.  For each of these  key study questions, Exhibit 1
displays five specific  study objectives for answering the key
questions.  For each-objective,  it discusses the study
rationale, the data needed  to  respond to each question, and
methodological or procedural problems in conducting the
analyses.

3.   APPROACH TO THE CONDUCT OF THE MIT MIDCOURSE REVIEW

     EPA's planning activities for the MIT evaluation must
differ from those for the AOR  review in order to be consistent
with differences in:

          Measurement of objectives
          Level of  regulatory  detail.

Unlike the AOR review,  in which measuring administrative
effectiveness must  serve as a  surrogate for measuring
achievement of substantive  program objectives,  the MIT
review directly measures substantive and administrative
objectives.   Thus, the MIT review will require a more de-
tailed, but less extensive  in  scope, collection and analysis
of program organization, activity, and resource data.  In
addition, unlike the proposed  rules for .the AOR midcourse
review, the final regulations  for the MIT evaluation contain
specific guidance on the midcourse review design.  Whereas
EPA had the opportunity for the AOR review to consider three
alternatives  (two of which  involved sampling), EPA must
evaluate sampling data  for  the MIT.  Furthermore, the
regulations require the states to comply with the following
in drawing the sample:
     For an explanation of the difficulties, see Geraghty & Miller, Inc.
     and Booz, Allen & Hamilton, Ibid, pp.  5-6.
                               •«
     A substantive objective is to determine well record reliability;
     an administrative one is to determine  changes in the administrative
     burden resulting from use of well record data instead of log data.

-------
                                                         EXHIBIT  1
                      Overview  of  Study Objectives  For  MIT Midcourse  Review
        STUDY OBJECTIVES
                                       BASIS OF ANALYSIS
                                                                 NECESSARY DATA FOR THE ANALYSIS
                                                                                                     ADDITIONAL QUESTIONS
1.   Determine if reviews of
    well  records are suitable
    substitutes for tempera-
    ture  or noise logs in
    identifying mechanical
    integrity problems in
    outer casings or well
    bores.
A comparison will  be made of
the results of well record
analyses for a sample of Class
II wells compared  to results
of noise and temperature logs
conducted for the  same  sample.
The key concern is the  precent
of variation between the
findings of the two analyses
and the characteristics of
wells for which variation
occurs.
Relevant data on  a  random
sample of Class II  wells as
required in Section 146.25
of the UIC regulations
including:
    results of well record
    reviews
    characteristics of
    wells
    characteristics of well
    record data

Type of state regulations and
requirements for  injection
wells in effect prior to the
implementation of the UIC
program.
What constitutes an
acceptable level of error?

Do the variations in accuracy
differ based on certain
characteristics of the sub-
ject well such as:
   age?
   location?
   past state regulatory
   practices?
   condition and detail of
   well records?
   type of cementing
   techniques employed?
   other factors?

For how many wells were well
records incomplete or in-
adequate, thus necessitating
use of temperature or noise logs?
2.   Determine if there are any
    technical factors which
    might preclude reliance on
    well records as an indica-
    tor of well integrity.
The analysis  will  identify any
technical problems which appear
to cause variations between the
results of record  reviews and
the results of  formal mechani-
cal integrity testing.  The
significance  of these factors
will be assessed and an
attempt will  be made to iden-
tify means for  avoiding or
overcoming them.
Examples of well  integrity
problems found  in formal
mechanical integrity  tests
which were not  identified
in well record  reviews.

Evidence on how well  records
could be improved so  as to
provide necessary informa-
tion.
Is there sufficient capacity
on the part of well testing
firms to manage all of  the
potential need for  mechanical
integrity testing?

-------
                                                    EXHIBIT  1  (Continued)
      STUDY OBJECTIVES
    Determine if major  cost or
    time differences  exist for
    well operators  between well
    integrity analyses  based on
    record reviews  and  those
    based on mechanical integ-
    rity tests.
4.  Determine if  the  adminis-
    trative burden for  state
    regulatory agencies varies
    depending upon the  type of
    review conducted  by Class
    II well operator/owners.
                                      BASIS OF ANALYSIS
The analysis will attempt
to determine if reliance on
record reviews rather than
on formal mechanical inte-
grity testing will result in
any time or cost savings to
well operators/owners and
lence any reductions in the
overall cost/production
Impact of the DIG regulations.
                                                                 NECESSARY  DATA  FOR  THE ANALYSIS
                                                                                                     ADDITIONAL QUESTIONS
The analysis will assess state
agency review procedures for
owner/operator submittals
Qased on well record reviews
and for submittals based on
temperature or noise logs to
determine:
-  The length of time
   associated with the state
   evaluation
-  The types of additional
   supporting information
   which the state agency
   must obtain
   The types of personnel
   required to conduct the
   evaluation
   Numbers of on-site in-
   spections of wells.	
Empirical data on costs of
reviewing well records and
costs of performing tempera-
ture and/or noise logs for
the sample of wells covered in
the survey.

Analyses of time involved in
both types of reviews in-
cluding:
   Overall time from start
   to finish of each type of
   review
   Actual well down-time
   attributable to each type
   of review
-  Time spent in administra-
   tive analyses of data
   supplied to state regula-
   tions for each type of
   review.
Can downtime attributed to mechani
cal integrity listing delays be
distinguished from delays associ-
ated with other factors?

Do time delays or cost increases
related to mechanical integrity
testing appear to vary according
to:
   The state in which they are
   performed
   The conditions of existing well
   records
-  The age of the well
-  Other factors.
Review calendar time frames

Number and skills of individ-
uals involved in reviews

Manhours spent on collecting
supporting information

Number and types of site
inspections.
Does not having log data on
well integrity adversely effect
any other aspect of the state's
UIC program?

Do states typically have or can
they acquire the skills necessary
to interpret mechanical integrity
testing data?

-------
                                           EXHIBIT  1  (Continued)
   STUDY OBJECTIVES
                                   BASIS OF ANALYSIS
                                                            NECESSARY DATA FOR THE ANALYSIS
                                                                                                  ADDITIONAL QUESTIONS
Determine if the  use of
well records as a sub-
stitute for log data
could be acceptable for
other classes of  wells.
If there is a strong  correla-
tion between the  conclusions
reached by reviewing  well
records and those resulting
from temperature  and  noise
logs. EPA may wish to allow
Class I and III well  owner/
operators to also use well
records.  This analysis will
assess the feasability of
doing this.  In part, it will
entail a comparison of any
technical differences be-
tween Class II wells  and I
and III wells.
Number of Class I arid  III
wells which would be affected

Potential cost and time
savings of expanding use of
well record reviews

Percentage of Class I  and  II
wells which have adequate  and
complete well records.
Are there any factors which
would increase the likelihood
that well records would be
less accurate in determining
mechanical integrity for
Class I or III wells than for
Class II wells.

Would there be a greater risk
from having misjudged Class I
or III wells than for mid-
judging Class II wells.

-------
          Sampling must be formal,  random selection.

          It must be  done on a field or pool basis and  be
          statistically representative of the fields or pools,

          At least 50%  of the wells tested must be existing
          ones.

Consequently, EPA is  in a position to begin detailed planning
Some of the principal issues it must address during this
planning phase  include  the format of data submittals; system
of data storage, retrieval and analysis; frequency of data
analysis and aggregation;^  and lead office/work group  for
conducting the  analysis.
     For example, EPA must determine whether to conduct the analysis
     after each six month reporting period or after 12 months, 18 months,
     or 24 months.

-------