40 C.F.R. §§ 1090.1360–1090.1375

The fuel must meet the parameter specifications in table 1 to paragraph (b)(4) of this section. This may require that you modify the fuel you typically produce to be within the specified range. Absent a specification (maximum or minimum), select a fuel representing values that are typical for your testing. Store and mix the fuel to maintain a homogenous mixture throughout the measurement period to ensure that each fuel sample drawn from the batch has the same properties.

Measure the fuel parameter from a homogeneous fuel batch at least 20 times. Record each result in sequence. Do not omit any valid results unless you use good engineering judgment to determine that the omission is necessary and you document those results and the reason for excluding them. Perform this analysis over a 20-day period. You may make up to 4 separate measurements in a 24-hour period, as long as the interval between measurements is at least 4 hours. Do not measure RVP more than once from a single sample.

An alternative procedure for measuring oxygenate in gasoline must account for every type of oxygenate covered by the referee method.

Obtain gravimetric sulfur standards to serve as representative reference samples. The samples must have known sulfur content within the ranges specified in paragraph (c)(3) of this section. The known sulfur content is the accepted reference value (ARV) for the fuel sample.

Measure the sulfur content of the fuel sample at your laboratory at least 10 times, without interruption. Use good laboratory practice to compensate for any known chemical interferences; however, you must apply that same compensation for all tests to measure the sulfur content of a test fuel. Calculate the arithmetic average of all the measured values, including any compensation.

You may use the alternative procedure only if you follow all the statistical protocols and meet all the criteria specified in Section 6 of ASTM D6708 (incorporated by reference in § 1090.95) when comparing your measurements using the alternative procedure to measurements at a reference installation using the appropriate referee procedure identified in § 1090.1360(d).

For qualifying alternative procedures, determine whether the alternative procedure needs a correlation equation to correct bias relative to the reference test method. Create such a correlation equation as specified in Section 7 of ASTM D6708. For all testing, apply the correlation equation to adjust measured values to be statistically consistent to measuring with the reference test method.

If an alternative VCSB procedure states that the procedure has a successful assessment relative to the referee procedures in this section under ASTM D6708, that finding applies for all laboratories using that procedure.

Demonstrate whether the procedure meets statistical criteria and whether it needs a correlation equation as specified in paragraphs (d)(1) and (2) of this section. Apply the correlation equation for all testing with the alternative procedure.

Demonstrate at your laboratory that the alternative procedure meets the accuracy criteria specified in paragraph (c) of this section.

Send EPA a written request to use the alternative procedure. In your request, fully describe the procedure to show how it functions for achieving accurate measurements and include detailed information related to your assessment under paragraph (e)(1) and (2) of this section.

Demonstrate whether the procedure meets statistical criteria and whether it needs a correlation equation as specified in paragraphs (e)(1) and (2) of this section. Apply the correlation equation for all testing with the alternative procedure.

Test with a range of fuels that are typical of those you will analyze at your laboratory. Use either consensus-named fuels or locally named reference materials. Consensus-named fuels are homogeneous fuel quantities sent around to different laboratories for analysis, which results in a “consensus name” representing the average value of the parameter for all participating laboratories. Locally named reference materials are fuel samples analyzed using the reference test method, either at your laboratory or at a reference installation, to establish an estimated value for the fuel parameter; locally named reference materials usually come from the fuel you produce.

You may qualify your procedure as meeting the requirements of paragraph (f)(1) of this section only for a narrower, defined range of fuels. If this is the case, identify the appropriate range of fuels in your request for approval and describe how you will screen fuel samples accordingly.

Qualify the precision of the alternative procedure by comparing results to testing with the referee procedure based on “between methods reproducibility,” Rxy, as specified in ASTM D6708. The Rxy must be at or below 75 percent of the reproducibility of the referee procedure in § 1090.1360(d).

Perform testing at your laboratory as specified in paragraph (b) of this section to establish the repeatability of the alternative procedure. The repeatability must be as good as or better than that specified in paragraph (b)(4) of this section.

Fully describe the procedure to show how it functions for achieving accurate measurements. Describe the technology, test instruments, and testing method so a competent person lacking experience with the procedure and test instruments would be able to replicate the results.

Send EPA a written request to use the alternative procedure. Include the specified information and any additional information EPA needs to evaluate your request.

Use the precision testing procedure specified in § 1090.1365(b) to show that your standard deviation for tests using the reference test method is at or below 0.3 times the reproducibility for a given fuel parameter.

You must correlate your test results for a given fuel parameter against the accepted reference values from a monthly crosscheck program based on Section 6.2.2.1 and Note 7 of ASTM D6299 (incorporated by reference in § 1090.95) as follows:

You must demonstrate that the reference installation is in statistical quality control for at least 5 months with the designated procedure as specified in ASTM D6299. If at any point the reference installation is not in statistical quality control, you must make any necessary changes and restart testing toward meeting the requirement to achieve statistical quality control for at least 5 months, except as follows:

The testing requirement applies for the referee procedures in § 1090.1360(d) and for alternate procedures that are qualified or approved under § 1090.1365. The testing requirements apply separately for each test instrument at each laboratory.

If you fail to conduct specified testing, your test instrument is not qualified for measuring fuel parameters to demonstrate compliance with the standards and other specifications of this part until you perform this testing. Similarly, if your test instrument fails to meet the specified criteria, it is not qualified for measuring fuel parameters to demonstrate compliance with the standards and other specifications of this part until you make the necessary changes to your test instrument and perform testing to show that the test instrument again meets the specified criteria.

If you perform major maintenance such as overhauling an instrument, confirm that the instrument still meets precision and accuracy criteria before you start testing again based on the procedures specified in ASTM D6299 (incorporated by reference in § 1090.95).

Keep records to document any test results excluded for being out of control under Section 8.5 and A1.5.4.1 of ASTM D6299. Identify the assignable cause and include any appropriate additional supporting justification.

Meeting the precision criteria of this paragraph (b) qualifies your test instrument for performing up to 20 tests or 7 days, whichever is less. Include all tests except for testing to meet precision or accuracy requirements.

Perform precision testing using the control-chart procedures in ASTM D6299. If you opt to use procedure 2A (Q-Procedure) or 2B (dynamically updated exponentially weighted moving average), validate the first run on the new QC batch by either an overlap in-control result of the old batch, or by a single execution of an accompanying standard reference material. The new QC material result would be considered validated if the single result of the standard reference material is within the established site precision (R') of the ARV of the standard reference material.

Use I charts and MR charts as specified in ASTM D6299 to show that the standard deviation for the test instrument meets the precision criteria specified in § 1090.1365(b).

Except as specified in paragraph (c)(3) of this section, test every instrument using a check standard meeting the specifications of ASTM D6299. Select a fuel sample with an ARV representing fuel that is typical for your testing.

The following provisions apply for method-defined non-VCSB alternative procedures with high sensitivity to sample-specific bias:

You meet accuracy requirements under this section if the difference between your measured value for the check standard and the ARV is less than the value from the following equation:

Your results from the crosscheck program are not valid for demonstrating compliance with accuracy requirements for a test instrument under this section if any of the following apply:

If your results from the crosscheck program are not valid under paragraph (a)(1) of this section, perform a root cause analysis and document your findings and the steps you take to correct the problem. You continue to meet accuracy requirements under this section for the affected parameter only if you correct the problem and demonstrate compliance with the accuracy requirements of this section within 45 days after learning of a failure under paragraph (d)(1) of this section. The compliance demonstration may be based on in-house testing using a check standard qualified by a third party, on a non-VCSB correlation program administered by a third party, or on testing in the next crosscheck program.