Mercury Work Group
Phase II Reports >> Hg Management Guidebook

 
APPENDIX D

SAMPLING AND ANALYTICAL TEST TECHNIQUES

MWRA Sampling and Reporting Requirements

MWRA sewer use permits often require that a sampling port be installed at the regulated discharge point for regular collection and analytical testing of representative samples of the discharged wastewater. The regulated discharge point usually means a location within the facility just prior to discharge to the sewer system after all appropriate pretreatment processes and operations but before any mixing of the industrial wastewater with other waste streams. Such a location is called the permitted sampling location.

Permit holders should be aware of MWRA sampling and reporting requirements, such as:

• Analyses submitted to comply with an MWRA sewer use permit or enforcement order must be based on samples that are both collected and analyzed by an independent analytical laboratory that is fully MA-DEP certified to analyze wastewater samples for the pollutant being reported. As an alternative to a MA-DEP certified laboratory collecting the samples at the permitted sampling location, a facility may request approval from the MWRA to use a specific sampling contractor or to sample its own waste streams.

• EPA requirements and MWRA permits also require that if a facility samples more frequently than required by its permit for any pollutant being discharged, the analytical test results from such sampling must be submitted to the MWRA following the procedures for submitting required sampling reports.

• Any analytical report being submitted must be for samples representative of the waste stream being discharged. For example, the sample must be:
  
  1. Collected to measure a parameter of a discharge at a permitted sampling location.
  2.  Representative of the discharge over the entire process day (for most parameters).
  3. Taken as a grab sample¹ or as a flow-proportioned (preferred) or time-interval composite sample as required by the MWRA permit.
  4. Correctly preserved.
  5. Analyzed using test procedures and within allowable holding times according to 40 CFR Part 136 or other EPA approved methods.

• If a sample is analyzed using an unapproved method, the analysis should not be submitted.

• If sampling shows a violation of the provisions of 360 CMR 10.022, 10.023, or 10.024, or a violation of a limit or prohibition contained within the sewer use permit, the facility shall notify the MWRA within 24 hours of becoming aware of the violation. (The notice is required no matter who took the sample). An additional sample must be taken and the analytical results must be reported within thirty (30) days from the date the violation is discovered.

The MWRA encourages permit holders to investigate any permit compliance problems by taking additional samples (at permitted and non-permitted locations) in order to identify contaminant sources. To account for possible sampling errors and precision and accuracy limits of analytical testing, it is suggested that a facility also acquire discharge concentration data over time to establish baseline discharge concentrations for applicable pollutants at each permitted sampling location.

As indicated above, it is required that all analytical data derived from permitted sampling locations be submitted to the MWRA.² It is recommended that analytical data for samples taken from non-permitted locations be submitted as evidence of good faith efforts to regain compliance. Such analytical data should be clearly identified so the data is not interpreted as representing wastewater from the permitted sampling location. Ideally, the data should be submitted as part of a summary report. Even if not submitted to the MWRA, testing results from non-permitted sites should be saved as documentary evidence of investigative activities.

For the study of environmental samples, ultra trace level analysis (UTLA) (i.e., at nanogram per liter or parts per trillion levels) is sometimes used. The EPA has supported the development of a new test method (Method 1631) for ULTA of mercury.³ Because of the very high sensitivity of ULTA, a "Clean Hands" sampling technique has been developed to reduce positive bias from mercury contamination that can occur during the sampling process. This sampling technique is recommended for the collection of water samples (e.g., surface water samples) before being analyzed for mercury because the samples are more prone to contamination with mercury than with other metals. The "Clean Hands" sampling technique, originally developed by the University of Wisconsin Trout Lake Station, employs two people, one designated as the "clean-hands person" and the other designated as the "dirty-hands person."

While Clean Hands sampling techniques are generally too rigorous and not warranted for conventional analytical testing of wastewater (including compliance sampling for mercury), an awareness of these techniques could be useful to identify possibly interfering sources of mercury that should be considered when conducting any sampling for mercury.⁴

For example, important Clean Hands or ULTA sampling considerations include:

• Sample containers should be cleaned according to EPA protocols and should be obtained from a respected commercial vendor. Field and equipment blanks using ultra-high purity water in the field can be used to identify sources of contamination during sampling activities. (Refer also to the section below on quality assurance.)

• Gloves used as personal protective equipment may have powders that contain mercury. If powders are used, it is important that they be free of mercury. Mercury free or powderless vinyl gloves are recommended.

• Airborne dust and atmospheric deposition are two examples of ambient sources of mercury. Facilities with on-site incinerators are environments prone to this type of mercury contamination.

• All sampling equipment wetted parts, including sampler pump tubing and containers, should be thoroughly decontaminated or replaced between sampling sessions.

Using the conventional EPA testing technique for mercury in wastewater by cold vapor atomic absorption (EPA Method 245.1), a typical commercial environmental testing laboratory will produce test results with a method detection limit (MDL) of 0.2 µg/L (ppb). The results are reported with a 95 percent degree of certainty (plus or minus approximately two standard deviations) with a relative standard deviation of about 20 percent.

The MWRA currently prohibits mercury in industrial sewer discharges, requires EPA Method 245.1 be used for discharge compliance monitoring, and applies an enforcement action threshold of 1.0 µg/L (ppb) for the mercury prohibition. The mercury enforcement threshold is five (5) times the typical MDL of EPA Method 245.1.

As mentioned above, the EPA has supported the development of Method 1631 for UTLA of natural waters to achieve an MDL for mercury in the low to sub-part per trillion (ppt) range. The analytical methods developed for UTLA, however, are often not applicable to wastewater samples because the actual detection limits that can be achieved using these methods are not much lower than those obtained by conventional methods. The higher detection limits are typically caused by sampling and matrix interference problems resulting from the highly variable qualities of industrial and institutional wastewater. However, good sample handling techniques derived from the ULTA studies can be important to reduce contamination problems in wastewater samples.

Beyond the potential for contamination during sampling, the measurement of mercury concentrations in wastewater samples may be subject to the following testing limitations:

Matrix interferences: Inherent sample qualities may preclude obtaining optimum test method performance by elevating the usual MDLs. For example, common laboratory reagents that may be present in the wastewater under investigation such as organic solvents, acids, bases, and surfactants may require considerable dilution of the samples to avoid damage to sensitive testing instrumentation. The dilution ratio that applies to the sample would also apply to the MDL.

Laboratory contamination: Mercury at trace levels may be present in the laboratory in the following areas:

• Inside ambient air, particularly in older facilities where spills from mercury thermometers and other sources may have occurred.
• Air supply - outside air or dust entering the laboratory through supplied-air fume hoods.
• Laboratory reagents, possibly including high-purity supply water.
• Laboratory glassware and other equipment such as disposable pipette tips, chem-wipes, and gloves.

Because the mercury content of wastewater samples may be close to or less than that detectable by EPA Method 245.1, some variations in the results from multiple analytical tests of the same sample may be considered normal. However, contamination during collection and processing of the samples can totally compromise the integrity of the analysis and subsequent interpretation of results. In fact, contamination and analytical problems in analysis of trace concentrations of mercury are well known and provide significant challenges for many laboratories. Therefore, analytical test results of the same samples by different laboratories can, in some cases, significantly vary for low concentration samples. Test result variability can sometimes even affect whether or not regulatory compliance is realized.

The following are recommendations to reduce these potential problems:

• Use extreme care in the collection and processing of samples. Typically, the collection of several hundred milliliters (ml) for analysis is required, more if replicate analysis is to be conducted or sample splits prepared. Contamination from dust and residual vapor phase mercury, improper handling (mercury-free gloves should be used at all times), improperly cleaned labware and cross-contamination from samples with very high mercury concentrations are especially prevalent sources of error.
   
• Labware in contact with samples should be cleaned by use of a common laboratory detergent and then acid-cleaned either in hot concentrated nitric acid or by soaking in 6M nitric acid (HNO₃) overnight followed by thorough rinsing with high purity deionized water. There must be strict adherence to safety procedures in handling of the acid. Store sample bottles filled with 10% HNO₃ prepared using high purity deionized water.
   
• The use of Teflon or glass labware is preferred to reduce mercury sorption losses. Sample containers (glass-stoppered 500 ml Erlenmeyer flasks or 500 ml Teflon bottles are useful) should be stored in plastic bags after cleaning and after filling with sample to reduce the possibility of contamination further. If possible, do not use sample bottles that previously contained samples with high mercury concentrations to collect samples that may have mercury at much lower levels (e.g., after pretreatment). After cleaning and between uses, store sample bottles filled with 10% HNO₃ prepared using high purity deionized water.
   
• Because, in some cases, environmental testing laboratories may have difficulty in producing consistent high quality results at 1 µg/l (ppb) or lower, a pre-evaluation process can be used to assess the capabilities of the laboratory (and your sampling personnel's ability to prepare clean sample bottles and collect uncontaminated samples). Using your current sampling procedures, send the candidate testing laboratory approximately five samples of high purity water. If the test results for mercury concentration of the samples are erratic, you will need to evaluate whether the sampling technique or the testing laboratory is the source of error.
   
• If test results for mercury concentration of high purity deionized water samples are erratic, a study of split samples sent to different environmental testing laboratories can be performed to find out whether one of the laboratories generates more erratic or higher results than the other on a consistent basis. The study can also determine whether good sampling techniques were used by performing the same analytical tests on high purity deionized water samples (or even wastewater samples) collected simultaneously by different sampling personnel and equipment.
   
• If a testing laboratory is unable to report consistent mercury concentrations for high purity deionized water samples, and if sampling tests suggest that sampling technique was not the source of error, the testing laboratory's quality assurance and quality control procedures and data (such as matrix spikes and matrix duplicates) should be reviewed. If the review is unsatisfactory, another environmental testing laboratory should be considered for the pre-evaluation process.

RETURN TO HG MANAGEMENT GUIDEBOOK
TABLE OF CONTENTS