[27 Pa.B. 4181]
[Continued from previous Web Page] Section 250.707. Statistical tests.
This section specifies the requirements for using and applying statistical tests to demonstrate attainment. The statistical tests may also be used to establish background concentrations at a site, as required by the background standard subchapter. The statistical test used to establish background must correspond with the statistical test used to demonstrate attainment with that standard.
The final-form regulations allow a person to choose between the 75%/10X rule for demonstrating attainment with the Statewide health standard, or a 95% Upper Confidence Level (UCL) of the mean statistical test, or other methods that meet specified performance standards, for demonstrating attainment with the Statewide health and site-specific standards. For the background standard in soil, a person may use a test which compares the highest measurements, or a combination of the Wilcoxon rank-sum test and quantile test, or other methods that compare the population of analytical results of background samples with a population of the medium of concern and meet specified performance standards. For the background standard in groundwater, a person may use the nonparametric Tolerance Intervals, or a retesting strategy using nonparametric Prediction Limits in accordance with the EPA guidance or other statistical methods that meet specified performance standards. A nonparametric statistical test compares distributions rather than parameters and is intended to apply to a large class of distributions rather than a single distribution. A parametric statistical test estimates parameters, such as arithmetic average, and tests hypotheses concerning them. The assumptions generally specify the form of distribution.
The 95% UCL test is documented in Federal guidances. The 95% UCL of the mean test is a parametric statistical procedure for determining whether the mean (average) concentration in the area of concern attains the cleanup standard. If the 95% upper confidence limit of the mean value is below the cleanup standard, the area of concern would be considered clean.
The 75%/10x rule requires that 75% of all samples collected for attainment purposes must be equal to or less than the standard with no individual sample exceeding ten times the standard. This rule requires that a sufficient number of samples be collected in the field to provide an acceptable result in the test. Therefore, a minimum of eight samples must be collected in order to reduce the false positive rate in the test. A false positive conclusion means that the statistical finding of clean is not representative of the overall field conditions at the site. To substantially reduce the false positive rate, the regulations require a minimum of eight samples in groundwater and a minimum of eight samples in soil equal to or less than 125 cubic yards.
On proposed rulemaking, using statistical simulations based on the median, the tests were evaluated using log normal distributions with coefficients of variation (Cv) ranging from 0.5 to 4.0, and with the number of samples ranging from 5 to 40. The conclusions were discussed more specifically in the preamble to the proposed rulemaking at 26 Pa.B. 3985. Since the time of the proposed rulemaking, the Department obtained additional information regarding the 75%/10x rule, based on additional simulation studies to evaluate the true arithmetic mean at a site that meets the cleanup standard. Specific changes to each subsection are as follows.
Subsection (a) has been changed to clarify the demonstration of attainment of the background standard. When the site involves naturally occurring substances, the area to which the site is being compared must be representative of naturally occurring regulated substances on the site. When the substances are not naturally occurring, the background area to which the site is being compared must be an area from which the contamination is migrating. The background areas to which the site is being compared are termed the background ''reference'' areas; this clarification occurs throughout this section.
Subsection (a)(1) provides that attainment of background in soil may be demonstrated by using the highest measurement comparison, a combination test, or other methods meeting general requirements. Subsubsection (a)(1)(i) has been modified to clarify that it is the obligation of the person demonstrating attainment to show that the highest measurement of contamination on- site is not higher than the highest measurement from the background area. In response to commentators, subsubsection (a)(1)(i) also has been changed to state that the Department may accept insignificant variations in the numbers.
Alternatively, a person may demonstrate attainment of background in soil by using a combination of the Wilcoxon rank-sum test and the Quantile test, and showing that background is not exceeded. This is a change from proposed rulemaking, when the nonparametric upper tolerance limit was required in conjunction with the Wilcoxon rank-sum test. This change was in response to a request from a commentator. Additionally, the clause relating to the percentage of nondetect data was deleted because it would be more appropriate to discuss the use of a combination of the Quantile and Wilcoxon tests in the Manual. The final-form regulations in subsubsections (a)(1)(iv) and (vi) provide that the application of this method, along with other methods, must also use a false positive rate not greater than 0.2, with a minimum number of 10 samples from the background area and 10 from each area of contamination, and that it must meet the requirements of subsection (d), which relates to the null hypothesis, the alternative hypothesis, and performance standards for statistical tests.
Also, in regard to the false-positive (Type I) error rate in subsection (a)(1)(iv), a commentator requested that the Department consider attainment based on a controlling constituent, rather than each constituent, to control the effect of multiple constituents on the decision error of the process. The Department believes that the overall error rate should be controlled, and the Type I error for each individual constituent can be adjusted to meet that goal as long as the power to detect contamination is not compromised and the overall error rate does not exceed 0.20.
Several commentators indicated that ten background samples and ten impacted area samples were excessive. The Department retained this requirement, however, because it was needed to provide sufficient power in the statistical method to detect contamination.
Other statistical methods besides the highest measurement comparison test and the combined Wilcoxon-Quantile test can be used, but both parametric and nonparametric methods must meet the conditions to which the combination Wilcoxon-Quantile test is subject regarding the false positive rate, number of samples, the null and alternative hypotheses, and the performance standards for statistical tests. In addition, if a parametric method is chosen, subsubsection (a)(1)(v) has been changed to provide that the censoring level for each nondetect shall be an assigned value that is randomly generated that is between zero and the limit related to the PQL.
Minor changes were made to subsections (a)(1)(iv)--(vi) to clarify what standards apply to various statistical methods.
Subsection (a)(2) governs the demonstration of attainment of background groundwater quality when there is a known upgradient release of a regulated substance. Subsection (a)(2)(ii) has been modified to provide that the upgradient concentration must be representative of concentrations in groundwater that are migrating onto the site. Attainment of background where groundwater is affected by substances migrating from an upgradient release must be demonstrated by sampling taken over eight quarters in both the background reference wells and the onsite wells, as clarified in the change to subsection (a)(2)(iii). The attainment method is applied to each monitoring well over eight quarters. In lieu of the eight quarters of sampling, the Department may accept eight samples over four quarters or, under the change in subsection (a)(2)(x), fewer than four quarters with written Department approval, if certain conditions, such as knowledge of the plume and parameters affecting fate and transport, are satisfied.
In addition to a minor change, subsection (a)(2)(ix) has been modified to say that the censoring level for nondetects is randomly generated between zero and the limit related to the PQL. Subsection (a)(2)(x)(C) has been modified to provide that the coefficient of variation may not exceed 1.0 for metals and 2.0 for organics, rather than using 1.0 for certain organic compounds as was done on proposed rulemaking. This change was made for the sake of simplicity. Another minor change was made to subsection (a)(2)(ix).
Subsection (a)(3) relates to the determination of background conditions due to naturally occurring or areawide contamination. The person using this subsection needs to obtain the Department's approval that the site is appropriate for the areawide contamination process. At least twelve samples are required, but sampling may be accelerated as long as serial correlation in the data does not result. Subsection (a)(3)(iv) was modified to clarify that all samples do not have to be taken simultaneously. Subsection (a)(3)(v) was modified for clarity; the substance remains unchanged. Subsection (a)(3)(vi) has been clarified to say that the values may be used with appropriate methods to compare the populations. Subsection (a)(3)(vii) has been changed. The proposed language provided that the sampling could not have ''extreme values.'' It has been changed to provide that the sampling results in the plume onsite may not exceed the sum of the background arithmetic average and three times the standard deviation calculated for the background area. This is in response to a comment asking the Department to explain what is meant by ''extreme values.'' Subsection (a)(3)(ix) has been modified to reflect the change of the censoring level to a random value, consistent with other sections.
Subsection (a)(2) of the proposed regulations has been deleted. This subsection related to the use of the standard related to the PQL as a default background standard for substances that are coming from a release somewhere off-property. This provision is not necessary since persons can sample reference areas to determine the background levels.
Subsection (b) sets out the statistical tests which may be accepted by the Department to demonstrate attainment with the Statewide health standard. The statistical tests apply to each area of soil contamination and to each groundwater well. Tests are performed individually for each regulated substance identified as present at the site, for which a person wants relief from liability under the act.
Subsection (b) has been modified to delete the application of the 75%/10X rule to attainment demonstrations of standards relating to PQLs and to the site-specific standard. This was done because the Department obtained additional information, since the time of proposed rulemaking, involving simulation studies of the tests based on the arithmetic mean. These simulations indicated that a site could pass the 75%/10X rule, although the true site arithmetic average concentrations could exceed the cleanup standard by two to three times, if the coefficient of variation is small. Because the Department believes that the Statewide health standards are based on conservative assumptions this concern is not as significant for that standard. However, under the site-specific standard, exposure assumptions are based on site-specific conditions. Therefore, although commentators supported the use of the 75%/10X rule for site-specific remediations, it was determined that the 75%/10X rule was appropriate only for the Statewide health standard. Subsection (b) was also modified to make it clear that the statistical test applies to each well used for monitoring compliance, as opposed to wells that are irrelevant because they are not within the area of contamination.
Subsection (b)(1) governs attainment of the standard in soil. Subsection (b)(1)(i) has been changed in the description of the requirements for the 75%/10X rule. The changes make it clear that the samples must be randomly collected, during a single event, and 75% of these samples must be equal to or less than the Statewide health standard or the limit related to PQLs, with no individual sample exceeding 10 times the Statewide health standard.
Subsection (b)(1)(ii) has been changed in regard to the use of the 95% UCL test for demonstrating attainment with the Statewide health standard for soil. The change allows the 95% UCL of the arithmetic mean to be at or below the Statewide health standard. This was made specifically in response to a request from a commentator to clarify whether the arithmetic mean or the geometric mean is of concern, and to a comment asking that the arithmetic mean be used. It is the arithmetic average concentration that represents the exposure to a receptor. Geometric average concentration is not representative of exposure. The Department considers a site to be adequately remediated when the true site arithmetic average concentration, not geometric average concentration, of a site is at or below the numerical standard.
Subsection (b)(1)(iii) includes a new procedure for demonstrating attainment with the Statewide health standard for soils. For sites that are localized storage tank contamination sites, if samples are taken in accordance with closure guidance for underground storage tanks, then that process may be used if the remediator takes fewer samples than would be required under the other tests in this section. Attainment can be demonstrated under Act 2 if no samples exceed the standard. This allows the use of fewer samples for storage tank sites. The concern that the statistical method would not have a large enough base is resolved by the requirement that none of the samples exceed the standard.
Attainment of the Statewide health standard in groundwater is discussed in subsection (b)(2). The first change to paragraph (2) makes it clear that attainment is demonstrated at each monitoring well used for compliance. Subsection (b)(2)(i) describes the functioning of the 75%/10X rule. Changes here clarify that 75% of the samples collected within each monitoring well over time shall be equal to or less than the Statewide health standard or the limit related to PQLs. No individual sample may exceed ten times the Statewide health standard on the property, or two times the Statewide health standard beyond the property boundary. For clarification, ''at the point of compliance'' was deleted.
In subsection (b)(2)(ii), the use of the 95% UCL for demonstration that the Statewide health standard has been achieved in groundwater has also been changed to clarify that the 95% UCL of the arithmetic mean must be at or below the Statewide health standard.
A new subsection, that allows the use of other statistical tests that satisfy the requirements of subsection (d), is provided for in subsection (b)(3).
Demonstration of attainment methods for the site-specific standard is discussed in subsection (c). This section allows the use of the 95% UCL for soil or groundwater, or a statistical test that meets the general requirements for other methods in subsection (d). The 75%/10X rule is not available for demonstrating attainment with the site-specific standard. This subsection also sets forth the requirement that the method used to demonstrate attainment must be the same as the method used in the risk assessment.
Subsection (d) sets forth the general requirements that apply to attainment tests other than the 75%/10X rule or the highest measurement comparison for demonstrating attainment in soil and groundwater. Subsection (d)(1) relates to methods for demonstrating attainment of the Statewide health standard or the site-specific standard. The application of these methods to background standards after remediation and to standards relating to PQLs is deleted. For the Statewide health and site-specific standards, the statistical method must employ a null hypothesis (Ho) that the true site arithmetic average concentration is at or above the cleanup standard, rather than that the null hypothesis is that the cleanup is not achieved. The final rule states that the alternative hypothesis (Ha) is that the true site arithmetic average concentration is below the cleanup standard, rather than that the alternative hypothesis is that the cleanup standard is achieved. The new language regarding the true site arithmetic average does not change the hypothesis, but clarifies that the intent of the section is that the true site arithmetic average is the statistical parameter of interest. The regulation incorporates a commentator's request to reverse the hypothesis statements for attainment of the postremediation background standard. This is because the proposed rulemaking would have required the person to do a remediation that could have resulted in actual site conditions being cleaner than the background area. Reversing the hypothesis for background using current methods allows some on-property measurements to exceed background concentrations and still demonstrate attainment. This is appropriate because it is accepted by the EPA and because it is more reasonable on a consideration of cost, to allow a small number of exceedances of the background standard, than to require a degree of cleanup that would result in all on-property sample concentrations to be at or below the background conditions. As a result of this reversal, both the initial determination and the postremediation determination of the background standard have the same hypothesis statements. Therefore, the word ''initially'' has been deleted from the phrase ''when statistical methods are to be used to initially determine that the background standard other than the limit related to the PQL is exceeded'' in subsection (d) (1).
Another general requirement for statistical methods is that they must be recommended for the particular use in a Department approved guidance or regulation rather than a relevant Federal guidance. This requirement is found in subsection (d)(2)(ii).
In regard to subsection (d)(2)(iii), commentators felt that compositing of the surface soil should be allowed. The Department agreed that composite sampling may be used under these regulations, but for situations other than those involving VOCs or the use of nonparametric methods. A change was made to subsection (d)(2)(iii) that compositing cannot be used for VOCs.
Subsection (d)(2)(iv) was changed to delete the requirement that statistical parameters shall be protective of human health and the environment. Additionally, the language regarding the censoring levels for nondetects was changed to be consistent with other subsections governing this issue.
Subsection (d)(2)(v) allows the Department to approve tests that do not need to account for seasonal and spatial variability and temporal correlation, where appropriate. The word ''account'' was substituted for ''control'' for purposes of readability.
Subsection (d)(2)(vi) has been changed to say that all tests, not just initial tests, used to demonstrate that background is exceeded shall maintain adequate power to detect contamination in accordance with current EPA guidance, regulations, or protocols. ''1-Beta'' was deleted since the power is described in the supporting guidance documents.
Subsection (d)(2)(viii) has been modified to clarify that statistical testing is done for each regulated substance.
In response to a comment, subsection (d)(3)(ii) requires new information to be documented in a final report when a statistical method is applied: it must include a clear statement of the applicable decision rule in the form of statistical hypotheses for each spatial unit and temporal boundary, including the applicable statistical parameter of interest and the specific cleanup standard.
A requirement is added to subsection (d)(3)(v) to specify false negative rates for demonstrating attainment with the background standard. This requirement is added to the general requirements for specification of false positive rates in methods used to demonstrate attainment with background and other standards.
Other minor changes were made to subsection (d).
Section 250.708. Postremediation care attainment.
Section 250.708 applies to remediations that require the use of engineering or institutional controls to attain and maintain a cleanup standard beyond the time that a final report is reviewed and approved by the Department. Implementation of a postremediation care plan is required if engineering or institutional controls are utilized to demonstrate attainment with a cleanup standard.
Commentators recommended that postremediation care requirements should not apply to passive engineering controls or to institutional controls other than those that require security measures. The wording was not changed because passive engineering controls, such as a slurry wall or cap, still require maintenance and care after the final report is approved. The regulation was not changed to limit postremediation care to institutional controls only if they involve security measures because a wide range of institutional measures are possible, and the postremediation care plan need only address applicable care requirements. Deed restrictions need no maintenance and thus would not have to be addressed in the postremediation care plan.
In response to a comment, subsection (a) was revised to clarify that the statistical test (to show that substances do not exceed the standard at the point of compliance) is done after engineering controls are in place and concentration levels have stabilized following any effects from the remediation. This revision is a change from the less precise language on proposed rulemaking that the groundwater has reached a ''consistent concentration level.'' Subsection (e) was clarified to allow persons to terminate postremediation care when they document that the standard is not expected to be exceeded in the future, through use of a fate and transport analysis.
Appendix A, Tables 1--8
On proposed, Table 1 included MSCs for groundwater in aquifers used or currently planned to be used, considering residential and nonresidential land use. On final rulemaking, the table also includes MSCs for groundwater in aquifers not used or currently planned to be used, considering residential and nonresidential land use. Table 1 is limited to organic regulated substances. In addition, columns have been added to distinguish between aquifers with less than or greater than 2,500 mg/L total dissolved solids, instead of addressing the distinction with a footnote.
A new Table 2 has been added which includes the MSCs for inorganic substances in groundwater. The table follows the same format as Table 1. In addition, a subtable containing substances with secondary MCLs has been added.
Table 3, formerly Table 2 in the proposed regulations, presents the MSCs for organic regulated substances in soil. This table has been divided into two categories: A) direct contact numeric values; and B) soil-to-groundwater numeric values. In addition, soil buffer distances have been added to Table 3, B. for each substance, in accordance with § 250.308(b) and (c). The MSCs for inorganic regulated substances in soil have been moved to Table 4 for the convenience of the user.
A new Table 4 has been added which identifies the MSCs for inorganic substances in soil and follows the format of Table 3. This table has been divided into two categories: A) direct contact values; and B) soil-to-groundwater numeric values.
Table 5, formerly Table 3 in the proposed regulations, has been modified to indicate the aqueous solubility for each regulated substance, which is used by § 250.304(b) as a cap to the MSCs for groundwater. A column has been added to indicate those substances that are liquids at standard temperature and pressure for the purpose of determining the appropriate saturation cap to be applied under § 250.305(b). A column has been added to indicate the boiling point of each substance to determine which are volatile organic substances for the purpose of implementing §§ 250.304(d) and 250.307.
Table 6, formerly Table 4 in the proposed regulations, has been modified to provide threshold of regulation soil-to-groundwater numeric values based solely on 100X the groundwater MSC.
Table 5 in the proposed regulations, relating to MSCs for radionuclides, was deleted because of the deletion of the MSCs for radionuclides in the final rulemaking.
Table 6 in the proposed regulations, relating to exposure assumptions for lead, was renumbered Table 7.
Table 7 in the proposed regulations, relating to the identification of constituents of potential ecological concern, was renumbered Table 8.
F. Benefits, Costs and Compliance
Executive Order 1996-1 requires a cost/benefit analysis of the final-form regulation.
Benefits The final-form regulations provide significant benefits to the public, local government and the private sector. The public and local government are notified of plans to remediate sites, by the person who intends to perform the remediation, prior to the initiation of the cleanup. In the past, this notice was not required. In addition, for cleanups that involve the site-specific standard or special industrial areas, a person who is remediating a site must publish the availability of the opportunity for a municipality to become involved in the remediation and reuse plans for the site.
These final-form regulations will encourage the voluntary cleanup and reuse of contaminated sites, restoring these sites to safe and productive uses, while promoting additional employment and tax revenues to distressed communities. The reuse of these sites will also reduce industrial development of greenfields sites.
Compliance Costs The Department does not anticipate any new compliance costs associated with the final-form regulations. Costs to remediate contaminated sites should be reduced based on the availability of a release of liability for compliance with the cleanup standards. Act 2, however, does impose fees for the submission of plans and reports that are reviewed by the Department. These fees will be collected by the Department and will be used to implement the provisions of Act 2, including implementation of these final-form regulations.
Compliance Assistance Plan Act 2 establishes an Industrial Sites Cleanup Fund, which is administered by the Department of Community and Economic Development. The fund provides financial assistance to persons who did not cause or contribute to contamination on a property used for industrial activity and who propose to undertake a voluntary cleanup of the property.
The Department of Environmental Protection has developed a Manual for the land recycling program. The Manual provides detailed, technical information on how to comply with Act 2 and the final-form regulations.
Paperwork Requirements The paperwork required by these final-form regulations is based on statutory requirements. Act 2 requires notices of intent to remediate and final reports for remediations. In addition, Act 2 requires the preparation of remedial investigation reports, risk assessment reports and cleanup plans for remediations that will attain the site-specific standard. For the remediation of special industrial areas, Act 2 requires the preparation of a work plan and a baseline remedial investigation report. Also, a person undertaking the reuse of a special industrial site is required to enter into an agreement with the Department based on the baseline remedial investigation report. The reports are an important aspect of the cleanup program because releases of liability will be based on the Department-approved reports that identify contamination and demonstrate compliance with a cleanup standard. The final-form regulations do not require additional paperwork beyond what is established by statute.
G. Pollution Prevention
Pollution prevention approaches to environmental management often provide environmentally sound and longer-term solutions to environmental protection because pollution is prevented at the source. Generally speaking, pollution prevention refers to measures taken to avoid or reduce the generation of all types of pollution at their points of origin. These final-form regulations will be applied after the pollution has been generated and a person is remediating the property. It should be noted, however, that these final-form regulations are intended to encourage the reuse of contaminated sites and prevent the generation of pollution at a site that is no longer contaminated.
H. Sunset Review
These final-form regulations will be reviewed in accordance with the sunset review schedule published by the Department to determine whether the regulations effectively fulfill the goals for which is was intended.
I. Regulatory Review
Under section 5(a) of the Regulatory Review Act (71 P. S. § 745.5(a)), on June 18, 1997, the Department submitted a copy of the proposed rulemaking to the Independent Regulatory Review Commission (IRRC), and the Chairpersons of the Senate and House Environmental Resources and Energy Committees. In compliance with section 5(b.1) of the Regulatory Review Act, the Department also provided IRRC and the Committees with copies of the comments, as well as other documentation.
In preparing these final-form regulations, the Department has considered the comments received from IRRC and the public. These comments are addressed in the comment and response document and Section E of this preamble. The Committees did not provide comments on the proposed rulemaking.
These final-form regulations were deemed approved by the House and Senate Environmental Resources and Energy Committees on July 8, 1997. IRRC met on July 17, 1997, and deemed approved the final-form regulations in accordance with section 5(c) of the Regulatory Review Act.
J. Findings of the Board
The Board finds that:
(1) Public notice of proposed rulemaking was given under sections 201 and 202 of the act of July 31, 1968 (P. L. 769, No. 240) (45 P. S. §§ 1201 and 1202) and regulations promulgated thereunder at 1 Pa. Code §§ 7.1 and 7.2.
(2) A public comment period was provided as required by law, and all comments were considered.
(3) These final-form regulations do not enlarge the purpose of the proposed regulations published at 26 Pa.B. 3985.
(4) These final-form regulations are necessary and appropriate for administration and enforcement of the authorizing acts identified in Section C of this Preamble.
K. Order of the Board
The Board, acting under the authorizing statutes, orders that:
(a) The regulations of the Department, 25 Pa. Code are amended by adding §§ 250.1--250.10, 250.201--250.204, 250.301--250.312, 250.401--250.411, 250.501--250.503, 250.601--250.606 and 250.701--250.708 to read as set forth in Annex A.
(Editor's Note: The addition of §§ 250.11--250.13, included in the proposal at 26 Pa.B. 3985, has been withdrawn by the Board. The addition of §§ 250.312, 250.411, 250.607 and 250.708 was not included in the proposal at 26 Pa.B. 3985.)
(b) The Chairperson of the Board shall submit this order and Annex A to the Office of General Counsel and the Office of Attorney General for review and approval as to legality and form, as required by law.
(c) The Chairperson of the Board shall submit this order and Annex A to IRRC and the Senate and House Environmental Resources and Energy Committees as required by the Regulatory Review Act.
(d) The Chairperson of the Board shall certify this order and Annex A and deposit them with the Legislative Reference Bureau, as required by law.
(e) This order shall take effect immediately upon publication.
JAMES M. SEIF
Chairperson(Editor's Note: For the text of the order of the Independent Regulatory Review Commission relating to this document, see 27 Pa.B. 4052 (August 9, 1997).)
Fiscal Note: Fiscal Note 7-300 remains valid for the final adoption of the subject regulations.
[Continued on next Web Page]
No part of the information on this site may be reproduced for profit or sold for profit.This material has been drawn directly from the official Pennsylvania Bulletin full text database. Due to the limitations of HTML or differences in display capabilities of different browsers, this version may differ slightly from the official printed version.
