Review of National Primary Drinking Water Regulations: Analytical Methods - Reassessment of Practical Quantitation Limits

BACKGROUND

The review of regulated contaminants will involve evaluating factors that contribute to the determination of a MCL. More specifically for analytical methods, the re-assessment will involve the re-evaluation of the Practical Quantitation Limits (PQLs) which were previously derived for some of the chemical contaminants at the time of promulgation of the NPDWRs. The purpose of this paper is to brief EPA stakeholders on the analytical methods-related issues for the 6-year review and the approaches EPA is considering for re-assessment of the PQL for these contaminants. With the help of stakeholder input, EPA will also consider developing a screening procedure to decide which PQL re-evaluation approach(es) will be most suitable for each regulated contaminant.

Analytical Methods Issues

The SDWA states that an MCL for a primary drinking water regulation must be set at a level at which "it is economically and technologically feasible to ascertain the level of such contaminant in water in public water systems...." [§1401(1)(C)(ii)], including quality control and test procedures to insure compliance [§1401(1)(D)]. According to section 1412 (b)(4)(B) of SDWA, EPA is to set the MCL as close to the MCLG as is feasible with the best available technologies. For the determination of some MCLs, the measurement capability may have been the limiting factor. This could be especially true for NPDWRs with MCLGs set at zero. Several NPDWRs having non-zero MCLGs may also have MCLs which were set due to the limits of the measurement capabilities.

Since the promulgation of pre-1996 SDWA contaminants, newer analytical methods for measuring SDWA contaminants have been approved. The approval of newer analytical techniques may have provided laboratories with the analytical capability to measure some contaminants at lower levels. The Agency would like to recommend to stakeholders that the chemical contaminants, whose MCLs were set due to the limits of the analytical capabilities, be selected as an initial screen for PQL re-evaluation. Table 1 lists a few of the subject NPDWR chemical contaminants that could be chosen for the initial screen, their PQL, acceptance limits, MCL and MCLG. The Agency is developing a complete listing of the pre-1996 SDWA contaminants with the above parameters that will be used for the initial PQL re-evaluation.. The Agency recognizes that a second set of contaminants for PQL review may result from the health effects screen in cases were the MCLG is lowered.

In considering analytical methods for use in compliance monitoring, EPA evaluates the overall sensitivity of the techniques. In previous regulations, EPA used two measures of analytical capability, the Method Detection Limit (MDL) and the Practical Quantitation Level (PQL).

• As per definition, the MDL is a measure of method sensitivity and it is defined at 40 CFR Part 136 Appendix B as "the minimum concentration of a substance that can be reported with 99% confidence that the analyte concentration is greater than zero." MDLs can be operator, method, laboratory, and matrix specific. Due to normal day-to-day and run-to-run analytical variability, MDLs may not be reproducible within a laboratory or between laboratories. The regulatory significance of the MDL is that EPA uses the MDL to determine when a contaminant is deemed to be detected and it can be used to calculate a PQL for that contaminant.

    

• In the preamble to a November 13, 1985 rulemaking (50 FR 46906), the PQL was defined as "the lowest concentration of an analyte that can be reliably measured within specified limits of precision and accuracy during routine laboratory operating conditions." The Agency has used the PQL to estimate or evaluate the minimum concentration at which most laboratories can be expected to reliably measure a specific chemical contaminant during day-to-day analyses of drinking water samples.

How are PQLs Determined?

A PQL is determined either through the use of inter-laboratory study data or, in absence of sufficient information, through the use of a multiplier of 5 to 10 times the MDL. EPA has conducted water supply (WS) performance evaluation (PE) studies twice a year for the last twenty years. Data from these studies can be used for PQL determinations. Using graphical or linear regression analysis of the WS data, the Agency sets a PQL at a concentration where at least 75% of the laboratories (generally EPA and State laboratories) could perform within an acceptable level of precision and accuracy.

What are the Potential Approaches for the Re-evaluation of PQLs?

There are several approaches that could be taken for the reassessment of the PQLs for SDWA contaminants:

Approach 1
The PQL-by-MDL-Multiplier Approach - Using the MDL in the currently approved method(s) for each contaminant, the 5 to 10 multiplier method could be used to estimate the PQL. This value would then be compared to the PQL that was derived before 1996 SDWA amendments.

Advantages - A relatively easy and clear process.

Disadvantages - (a) The WS PE-studies test laboratory performance near the MCL not near the MDL. A PQL derived from the MDL method may not be representative, because the reproducibility of a result obtained at the MDL is often not as good as that obtained near the MCL. (b) Because several methods may be approved for the same contaminant, it can be difficult to decide which MDL to select for the PQL calculation.

Approach 2
Analysis of WS PE-Data - Use data from these studies to derive a new PQL. Compare this value to the old PQL.

Advantages - (a) Uses inter-laboratory data collected at concentrations near the MCL. (b) More representative of what methods are being used for the analysis of that contaminant. (c) May be the preferred approach for contaminants with MCLGs of zero.

Disadvantages - (a) The PQL derived for each contaminant is affected by the Agency's choice of "an acceptable level of precision". Acceptance units are not the same for all methods or contaminants. These levels have been set at ±10%, ±20%, ±30%, ± 40%, etc. (b) The PQL is also influenced by the PE-data used i.e., all data or only data from EPA State and Regional laboratories? (c) Laboratory performance on PE-data may be skewed, because PE-samples may be treated as special samples that are critical for laboratory certification. (d) The derivation of PQLs from WS data is resource-intensive.

Approach 3 The Minimum Level (ML) MDL-Multiplier Approach. This approach is used in the discharge permit and pretreatment programs for wastewater. The ML is calculated by multiplying a specified MDL for a contaminant by 3.18 and rounding to an integral number.

Advantages - Relatively easy and clear process and provides consistency across water monitoring programs.

Disadvantages - (a) In addition to the disadvantages under Approach 1, many MLs will be less than current PQLs which could lead to lower MCLs on a seemingly arbitrary basis. (b) Unlike the previous two approaches this approach has not gone through peer and public review under the drinking water program. This review would delay the reassessment process.

QUESTIONS FOR STAKEHOLDERS

• Which of the above approaches for calculating PQLs do you favor and why? Are there other approaches that could be considered?

• Do stakeholders agree that the proposed initial screen is appropriate (i.e., if a PQL is equal to the MCL, and both are greater than the MCLG, then that contaminent is to be "caught" by the screen)?

• What other issues or approaches should the Agency consider in re-evaluating the PQLs for SDWA contaminants?

• If the currently approved methods for a contaminant contain different MDLs and the Agency adopts a PQL-by-MDL-multiplier approach, which MDL should be selected? The lowest, an average or highest MDL? Or should the selected MDL represent the method that is most widely used for the analysis of that contaminant?