EPA Method 1664: Waiting for Godot

U.S. Epa Method 1664, Oil and Grease Analysis in Wastewater, was proposed in the Federal Register (1996 Jan 23; 61:1730). On August 22, 1996, Ben Honaker of the U.S. EPA Analytical Methods staff (Washington, DC) stated that his office will promulgate 1664 during January 1997. Furthermore, the Federal Register proposes to rescind U.S. EPA Method 413.1 and Standard Method 5520B six months after promulgation of 1664. The U.S. EPA conducted extensive evaluations of possible alternative solvents prior to issuance of this method, ultimately concluding that none duplicated the Freon-113 results of Method 413.1. n-Hexane was selected for the new method, and recovery comparisons were run on numerous wastewater types during the validation process. The data in Table 1 were generated during the U.S. EPA Chlorofluorocarbon (CFC) Replacement Study. There are several areas laboratories must address as this method change approaches.

One of the differences between the new method and the old method is the change in extraction solvents. Because the new solvent, n-hexane, required in 1664 is generally equal to or slightly less aggressive than Freon, expectations hold that most waste streams will yield an equivalent or smaller amount of oil and grease. However, a small percentage of waste streams have shown solvent-related matrix interferences that substantially increase recoveries of the method-defined pollutant. In such cases where the discharge stream has met permit requirements under 413.1 or 5520B, but fails under 1664, side-by-side results of homogeneous samples run by Methods 413.1 and 1664 will prove advantageous to the permit holder in negotiating with the permit issuer.

A prime example of matrix interferences associated with 1664 took place in Tuscaloosa, AL, at TTL, Inc., a commercial environmental laboratory. Under the guidance of Steven C. Martin, the laboratory director, TTL conducted method comparison studies. These analyses were performed in an effort to alert clients to potential violations due to matrix interferences. One stream tested turned into a solid-like gel within the separatory funnel after the solvent was added and the funnel shaken. Recovery of any material from that waste stream by 1664 liquidÐliquid extraction (LLE) proved impossible. Martin then had his laboratory run that same stream by the solid-phase extraction (SPE) method allowed under 1664. TTL was then able to properly analyze this sample stream because SPE eliminates emulsions and other matrix interferences.

As opposed to 413.1, Method 1664 requires four QA/QC samples per analytical batch. And a batch, as defined by 1664, may not exceed 10 samples. QA/QC samples consist of 1) a blank, 2) a precision and recovery sample (or blank spike), 3) a matrix spike, and 4) a matrix spike duplicate. Therefore, if only one real-world sample is run, a total of five extractions need to be performed; if 10 samples are run, 14 extractions will be undertaken.

Another major difference between Freon (CFC-113) extractions and n-hexane extractions is that n-hexane is lighter than water. If a laboratory decides to run oil and grease determinations by n-hexane separatory funnel extractions, as allowed under 1664, a great deal more time will be spent dealing with emulsions, draining off the water, bringing off the solvent, and then repouring the water into the separatory funnel. This cumbersome process must be repeated a total of three times before the sample is extracted according to method requirements. In transferring the sample this many times, the opportunity for the introduction of error greatly increases. And because n-hexane is lighter than water, 1664 LLEs take much longer to run than traditional 413.1 analyses. Fortunately, however, another option is available: Method 1664 allows SPE.

A&L Analytical Laboratories, Inc., a Memphis, TN, corporation, runs a large Environmental Services Div. under the direction of Jeffrey H. Papasan, Ch.E. A&L has been experimenting with SPE since November 1995. In a letter dated August 1, 1996, Papasan reports that A&L found that U.S. EPA 1664 (by SPE) is not as labor intensive as U.S. EPA 413.1 and will allow the technician the freedom to perform other laboratory tests while at the same time running U.S. EPA 1664. A&L's estimated complete run time for seven to eight real-world samples, including blank, matrix spike duplicate, and blank spike, should be 20 min per sample. Of this 20 min, approx. 50% is dedicated solely for U.S. EPA 1664 analysis.

Carlton Gibson, Senior Environmental Technician at A&L, has performed most of the hands-on work in this area. Gibson states that he likes this system because it allows him to run additional tests, which means he can be more productive.

Shown in Figure 1 is A&L's quality control chart of ongoing precision and recovery samples. The mean recovery is 36.08 mg/L, or 90.2%. Only two of 41 data points fall outside of one standard deviation. Papasan states that the methods and systems employed yield data that prove the method will perform within established QC/QA requirements.

A&L performs SPE with a three-position, 47-mm StepSaverª SPE system (Environmental Express, Ltd., Mt. Pleasant, SC). The system is used in conjunction with Emporeª disks (3M, St. Paul, MN). The patented StepSaver system allows A&L to collect the dried eluant directly into the pretared boiling flask, eliminates emulsions, requires no cleaning between samples, and is easy to operate. Of all brands tested, UltimAr 95% n-hexane (Mallinckrodt, St. Louis, MO) provided the best blank numbers.

Another area of concern with U.S. EPA Method 1664 is the proposed method detection limit of 1.4 mg/L. The 17th, 18th, and 19th editions of Standard Methods, no. 5520, Oil and Grease, state in paragraph 2, ÒSelection of Method,Ó that for low levels of oil and grease (<10 mg/L) Method C is the method of choice because gravimetric methods do not provide the needed precision. Note that Method C is the infrared method. It is therefore not surprising that, of the two companies that contracted with the U.S. EPA to develop 1664's MDL, only one met current requirements. A&L's method blank from November 15, 1995, to July 17, 1996, averaged 1.30 mg/L. A variance of +10% in A&L's blank would exceed the MDL.

The Lockheed Martin waste water chemistry laboratory in Oak Ridge, TN, led by Gary Emmerson, has performed an initial evaluation of 1664. In May 1996, Emmerson appointed Rife Chambers, a respected senior chemist, to begin work on an MDL study. Chambers spiked his blanks at 8 mg/L for his initial MDL study. That initial study by Martin Marietta yielded an MDL of 2.1 mg/L. Chambers is confident, however, that he can reduce his current MDL to an acceptable level by spiking in a range that is less than or equal to the reporting limits of the method (i.e., 5 mg/L).

Laboratories that have achieved lower-range MDLs have run the study with a spike that is at or below the method reporting limit, generally in the 4-5 mg/L range. The spiking solution required is a hexadecane/stearic acid, acetone matrix solution. A 40-mg/L spike of the same solution design is required by the U.S. EPA for ongoing precision and recovery samples (blank spikes or LCS) and matrix spikes. Both of these solutions are supplied by Environmental Express.

The U.S. EPA currently lends support for the use of this method. Although Method 1664 has not yet been promulgated, Bill Telliard of the U.S. EPA Analytical Methods Staff (Washington, DC), in a letter addressed to Pretreatment Coordinators, Regional National Pollutant Discharge Elimination System (NPDES) Contacts, dated July 9, 1996, writes in part that, until Method 1664 is promulgated by publication of the final rule in the Federal Register, U.S. EPA supports the use of Method 1664 as an interim limited-use alternate test procedure in place of U.S. EPA Method 413.1 and Standard Method 5520B. . Telliard continues his directive by stating that several states have already taken this step by allowing interim, limited use of Method 1664, although a few of them have limited the use of performance-based options that are a part of Method 1664. Telliard addresses those states' position by emphasizing that the U.S. EPA believes that all analytical options within Method 1664 should be allowed.

CFC replacement study

a CFC-113 and n-hexane LLE data were provided by U.S. EPA. SPE results were generated by 3M Laboratories; results shown in the table are the average of three separate 1-L sample analyses. Samples were collected by and provided to participating laboratories by the U.S. EPA.

b These samples required 90-mm disks to obtain acceptable extraction times. The remainder were extracted on 47-mm disks.

c Average percent recoveries and RSDs were determined using CFC-113 LLE results as the true value for oil and grease content.

Figure 1 Quality control chart of oil and grease by SPE Method 1664 (true value = 40.00 ppm).

Mr. Stone is Vice-President of Environmental Express, Ltd., 490 Wando Park Blvd., Mt. Pleasant, SC 29464, U.S.A.; tel.: 803-881-6560; fax: 803-881-3964.