v>EPA
U n ited States
Environmental Protection
Agency
Office of
Solid Waste and
Emergency Response
DIRECTIVE NUMBER: 9481.04(83)
TITLE: SCA Request for Alternative Ground-Water Monitoring
System and "Corrected" T-Test
APPROVAL DATE: 7-1-83
EFFECTIVE DATE: 7-i-83
ORIGINATING OFFICE: office of solid waste
Q FINAL
D DRAFT
STATUS:
( ]
(
[
[
REFERENCE (other documents):
A- Pending OMB approval
B- Pending AA-OSWER approval
C- For review &/or comment
D- In development or circulating
headquarters
OSWER OSWER OSWER
VE DIRECTIVE DIRECTIVE Dl
-------
PARTS 264 AND 265 SUBPART F - GROUND-WATER PROTECTION
DOC: 9481.04(83)
Key Words:
Regulations:
Subject:
Addressee:
Originator:
Source Doc:
Date:
Summary:"
Ground-Water Monitoring, Student's T-Test
40 CFR 264.97 and 265.91
SCA Request for Alternative Ground-Water Monitoring
System and "Corrected" T-Test
Richard Kotz, Statistician, Air and Ground Water Protection
Program
Mary Ann 0'Gorman, Statistician, Air and Ground Water
Protection Program
#9481.04(83)
7-1-83
Subpart F of Part 265 requires ground-water monitoring upgradient of a
facility to provide an indication of background ground-water quality. Back-
ground samples cannot be obtained from individual wells for the purpose of
comparison with subsequent samples from the same wells.
There are no provisions for establishing alternatives to the statistical
procedure under Part 265. However, under Part 264, alternatives to the student's
t-test can be approved when the specific situation indicates a need for them.
Any alternative statistical procedure must meet two important requirements.
First, the procedure must be appropriate for the distribution of the data used
to establish background values or concentration limits. The second requirement
is that the procedure provide a reasonable balance between the risk of "false
positives" and "false negatives."
-------
9481.04 (83)
l/
SCA Requeirt for Waiving interim Status Requirements
Mary Ann O 'Gorman, Statistician
Air and Ground water Protection Prog ran (WH-565E)
Richard Kotc, Statistician
Air and Ground Water Protection Program (WH-565E)
The following is a response to a request by SCA Chenical
Services Inc. to use a modified ground-water monitoring system
evaluation. The request was made in a letter of May 6, 1983
from Kevin Grant of SCA to James Reidy of Region II. A follow-
up letter and report were also sent on May 20, 1983 to Mr.
Jonathan Josephs in the Regional Office by Mr. Grant.
In the May 6 report a request is made to use background
ground-water data at each individual well for use in detecting
contamination rather than using background ground-water data at
upgradient wells to determine contamination at downgradient
wells. Subpart F, section 265.91(a) ( 1) ( i) states that the
number, locations and depths of the upgradient wells must be
sufficient to yield ground-water samples that are representative
of background ground-water quality in the uppermost aquifier
near the facility. If the upgradient well does not give an ac-
curate representation of background ground-water quality the
appropriate procedure is to drill additional wells at various
locations upgradient from the site. Thus the proposed procedure
is inappropriate and should not be considered.
-*
The report also states that all other requirements of Part
265 Subpart P would reraain the same. However, in the report
of May 20 a "corrected student's t-test" is proposed and used
to show no significant differences in the ground-water quality
at downgradient^ wells.
It should be noted that under Interim status regulations,
Part 2C5 Stltopart T, which govern this case, there are no provisions
for establishing alternatives to the statistical procedure. However,
the permitting regulations, under Part 264, do provide for
alternative statistical procedures where the specific situation
indicates a need for them. In that caae, any alternative procedure
must meet two requirements deemed to be important to EPA.
First, the procedures must be appropriate for the distribution
of the data used to establish background values or concentration
limits. The second requirement is that the procedure provide
a reasonable balance between the risk of false positives and
WH-565E:MO'Gorman:SWalker:nnM2 102: 382-4658 :WSM:6/29/83
-------
false negatives. SCA did not even address these two requirements
in their>request that we waive the interim status requirements.
SCA gives no indication of the risk of false positives or
false negatives. Thus, no comparison is Bade between the
•corrected student's t-test* proposed by SCA and the student's
t-test required under 265.93(b).
In addition, no justification whatever is given to explain
why this corrected t-test is appropriate for this situation.
Before this proposed test can be considered for use in this case
some justification is needed from SCA as to its appropriateness.
In the analysis done in the report of May 20 it is unclear as
to the comparisons being made. Since all of the data used are
not included it is impossible to determine exactly what compar-
isons are being made or even to verify the calculations of
aeans and variances used.
In the description of the procedure it is stated that
if normality is assumed then the background groundwater variance-
will be equal to the monitoring ground-water variance. This
statement is not true; normality does not imply equality of
these variances* The variances may or may not be equal regard-
less of the normality assumption.
Therefore, unless SCA can address these problems and give
a justification for the corrected t-test it should not be con-
sidered an acceptable alternative. The original analysis given
in the report of May 20 using the procedure in 265.93(b) should
be considered the appropriate analysis.
------- |