2010 EPA Protocol Gas Audit Results

% dif = 100 x (Tag Value - NIST Value) / NIST Value

Very Low = 0 - 10 ppm for NO and SO₂; and 0 - 1.0% for CO₂.

Low = 11 - 100 ppm for NO and SO₂; and 1.1 - 8.0% for CO₂.

Mid = 101 - 600 ppm for NO and SO₂; and 8.1 - 15.0% for CO₂.

High = Above 600 ppm for NO and SO₂; and above 15.0% for CO₂.

Without informing the production site that a blind sample of their cylinders was being audited, three cylinders per production site were purchased by a company that normally purchases these types of cylinders. The cylinders in the sample were analyzed by the National Institute of Standards and Technology (NIST). Within each parameter (NO, SO₂, CO₂), all low concentrations were essentially identical, all mid concentrations were essentially identical, and all high concentrations were essentially identical.

The above audit results are for the cylinder original concentration for each component and do not include any re-analyses that may have been performed by a production site. The rationale for including summary results for only original tag values is that gas cylinder purchasers will receive cylinders with tag values based on original analysis, not re-analysis, so it is important for a gas production site to produce an accurately labeled cylinder the first time. Please see the full audit report for more complete information and for any re-analyses and corrective actions performed by a production site.

A gaseous component is said to fail when the absolute value of the difference between the audit and vendor concentration values is greater than 2.20%. The 2.20% value is determined by using the "paired t test" at 95% confidence, with an uncertainty of plus or minus 2.0% (fixed by Part 75, Appendix A, section 5.1.4(b)) for the gas vendor and an expanded uncertainty (coverage factor k=2) of plus or minus 1.0% (maximum acceptable) for the audit. If on future audits, e.g., for very low concentration gases, the plus or minus 1.0% audit expanded uncertainty value changes, the 2.20% value may change. If the absolute value of the difference between the audit value and the vendor value is 2.20% or less, then (because of the uncertainties in the total measurement system) statistically there is no difference between the two values. Thus, a difference of 2.10% would be interpreted as being equal to one of, for example, 0.40%.

Nothing can be said regarding the performance of any EPA Protocol gas production site not included in the audit. Any accuracy assessment is an instantaneous snapshot of the process being measured. These results should not be regarded as a final statement on the accuracy of EPA Protocol gases. They can be used as a general indicator of the current status of the accuracy of EPA Protocol gases as a whole. However, individual results should not be taken as definitive indicators of the analytical capabilities of individual producers. EPA presents this information without assigning a rating to the gas vendors, for example, who is the best or who is approved or not approved and specifically does not endorse any particular vendor.