![[logo] US EPA](http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Procedures for Detection and Quantitation
WORKING DRAFT FOR DISCUSSION
Policy Issues Identified by Policy Work Group Members
In the Revised Consolidated Description of Uses
Numerous policy issues were identified by the group in the "Revised Draft Statements of Current Uses of Detection and Quantitation, Data Quality Objectives and Policy Issues," (November 3, 2005). In two of the topic areas, the authors suggested priorities for policy discussion. The priority policy issues identified within the uses of permits and compliance/enforcement are highlighted below, followed by the remaining policy issues listed by use.
- Permit Applications
- To what level should we require laboratories and permit applicants to report data? Should there be a censor point and if so, at what level (LQ, LD, LC, etc) or should we report the numerical data along with the LQ, LD, and LC and then perhaps use flags to qualify data?
- At what level should the regulatory agency begin assigning numerical values (non-zero) in calculations to determine reasonable potential?
- Should D/Q limits be whatever an individual laboratory determines or should the agency set acceptable D/Q limits that a lab must meet?
- Should we and when should we be requiring use of methods with the lowest D/Q?
- How large of a body of data (statistical significance) should the regulatory agency require before we choose to impose an effluent limit in a permit when D/Q issues cannot be avoided because the lowest D/Q method is not enough?
- Compliance Assurance and Enforcement
- What is the appropriate level of confidence for both type 1 and type 2 errors that should be used to establish LQ, LD, and LC for determining compliance? For enforcement?
- Where the calculated limit is less than LQ, LD, or LC should the limit be set at the calculated value or at LQ, LD, or LC, respectively. (Really three questions)
- Where the calculated limit is less than LQ, LD, or LC should LQ, LD, or LC, be used to determine compliance? (Again, three questions)
- Should LQ, LD, and LC be determined at the individual laboratory level or based on inter-laboratory studies?
- If LQ, LD, or LC are used to determine compliance, how would that be accomplished where averages (weekly or monthly) would have to be calculated?
- With respect to determining compliance and taking enforcement, should there be any distinction between the types of discharge (permitted process water, permitted stormwater, or un-permitted)?
- Water Quality Criteria and Local Water Quality Standards Setting
- Water quality criteria provide the basis for water quality based effluent limits, which have largely been responsible for D/Q issues. In addition, water quality criteria typically determine the targets for waters requiring total maximum daily loads (TMDLs). Because they are often lower than national criteria (and often below detection or quantitation levels obtained using EPA-approved methods), criteria resulting from the Great Lakes Initiative have helped highlight the importance of measurement and method sensitivity in EPA-approved methods.
- EPA Method Promulgation and Validation
- Numerous stakeholders have criticized the current procedures for calculating the MDL and ML. These criticisms have been captured by the FACDQ's Technical Workgroup.
- Should the method development process define the measurement quality objectives or should the various uses drive measurement quality objectives?
- What is an appropriate level of false positives? False negatives?
- Should quantitation be defined in terms of both precision and accuracy?
- Should precision and accuracy be established without reference to detection and quantitation?
- If the procedures by which EPA calculates the detection and quantitation limits change, should EPA revise detection and quantitation limits for previously promulgated methods? Which ones (there are hundreds)?
- If we change the EPA detection and quantitation limits, how can we ensure the further development of more sensitive methods?
- EPA Development and Approval of New and Alternative Methods
- Many of the CWA reference methods do not have published single or inter-laboratory validation method detection limit performance criteria.
- Lower range of quantitation stated in the method where there is no published single or inter-laboratory validation method detection limit performance criteria, is often lower than what can be achieved through actual method detection limit demonstration.
- Not all determinants methods are amenable to the current MDL and ML model.
- States and EPA Regions don't understand the Tier method approval process, and thus are reluctant to approve Tier I method applications with method detection limit performance criteria.
- Due to Agency resources (monetary and human) and agendas, method approvals are not expeditious, often taking 18 to 36 months. Thus, new or alternative methods that are superior in method detection limit performance are delayed in their use for regulatory purposes.
- Effluent Limitation Development BAT Averaging and Effluent Guideline Development Variability
- None except there is currently a concern that a more sensitive PCB method might detect PCBs that are inherent in a facility's environment.
- Laboratory QA/QC
- Should LQ, LD, and LC all three be used or would LQ and LC suffice?
- What will the transition period be for implementation of the new procedure?
- How will communication, notification and training be accomplished for the new procedure?
- What procedures or allowances will be incorporated for matrix effects that may prohibit achieving promulgated detection limits?
- Pretreatment (enforcement of pretreatment standards)
- Policy issues closely track those for Permit Applications and Compliance Assurance and Enforcement. In addition, how will the possibility for matrix effects unique to indirect dischargers be addressed? Though there are few instances now where categorical pretreatment standards or local limits have been set at or below the level of detection or even quantitation, there are cases where TMDLs may drive pretreatment programs to regulate at detection (mercury limits for some POTWs can also be a driver). Compounded by the possibility for large matrix effects, additional guidance will be needed for determining where expected detection and quantitation limits cannot be met.
- Water Quality Monitoring
- As with the establishment or modification of water quality standards (in particular water quality criteria), detection and quantitation issues may drive the need for development of more sensitive test methods that allow for detection and quantitation at levels low enough in order to make more informed use support decisions for surface waters.
- What detection and quantitation criteria should be used as part of decisions to identify waters not meeting standards [per 305(b)] or water quality limited segments needing more than mandated minimum technology-based controls [per 303(d)]?
- Non-Regulatory Operational Monitoring
- Permittees are required to report all analytical results generated on a permitted outfall in the DMR if a 40 CFR Part 136 procedure is used. This makes all such measurement subject to the same certification requirements as for routine compliance monitoring.
- Other Studies
- Should there be a specific level of confidence in data when scoping problems like fish tissue contamination or biosolid characterization (the two examples provided)? This would be established by the MQOs associated with the detection and quantitation limits used for these studies.