W.Warren McHone, Ph.D. Report

W. Warren McHone, Ph.D. Professor of Economics University of Central Florida'
Orlando, Florida 32816

Per my review of March 5, 2003, 1 have reviewed all of the materials that you provided on Centron Products. These materials included:

• A detailed description of your product, its production process and its intended use and benefits
• The United States Patent Office review and patent number assignment for Centron products
• The US Environmental Protection Agency notification of registration of your products
• Product safety information
• A series of fuel economy and pollution emissions reduction tests that were conducted over the period 1999 through 2002.
• I can verify that all of the about materials appear to be present and valid. In addition to reviewing the above listed materials, I performed standard statistical analysis of each of the fuel economy and pollution emission reduction tests to establish the levels of statistical confidence that can be attached to these test results. These statistical analyses are provided in Tables 1 through 6 of the attachment to this letter. My conclusions are as follows:
  • All fuel economy and emission reduction tests were conducted in a sound and appropriate manner.
  • The range of tests covered a wide variety of vehicle types and operating environments.
  • The test results strongly suggest that the application of Centron Products to gasoline and diesel fuels of the test vehicles yielded substantial average improvements in fuel economy with the average of all tests equaling 18% and ranging as high as 58%.
  • The statistical analysis of the results indicates that there is a 95% degree of statistical certainty that the fuel economy improvements were at least 15.4%. Moreover, more strenuous statistical tests at the 97.5% and 99% degrees of certainty are, respectively, associated with 14.9% and 14.3% minimum levels of fuel economy improvement.
  • The pollution emission reduction test results suggest that the application of Centron Products to gasoline and diesel fuels of the test vehicles led to very large reductions in harmful air pollutants. The average reduction in CO was 57%, reductions in HC emissions averaged 67% and NOx reductions averaged 50%.
  • The statistical analysis of the emission reduction test results reveals that there is a 95% level of statistical certainty that emission reductions were at least 43% for CO, 57% for HC and 43% for NOx. Furthermore, these reductions are still substantial even when imposing the more strenuous statistical confidence levels of 97.5% and 99%. More specifically, there is a 97.5% level of confidence that the CO, HC and NOx emission reductions were at least 40%, 55% and 42%, respectively. Finally, the statistical tests indicate that there is a 99% level of confidence that emission reductions averaged at least 37% for CO, 53% for HC and 40% for NOx.
  • The average reduction in Opacity for test vehicles was 36%. The statistical analysis of these test results indicates that there is a 95% degree of certainty that Opacity was reduced by at least 25%; a 97.5% degree of certainty that the reduction was at least 22% and a 99% level of confidence that the reduction was at least 19%.

In summary, in my professional opinion the Centron test results for fuel economy and pollution emission reduction that were provided to me support Centron Energy Corporation, Inc.'s claims that Centron substantially improves fuel economy and reduces the emission of harmful pollutants. Furthermore, the claims are supported at a very high level of statistical confidence.

W. Warren 1VIcHone, Ph.D. Professor of Economics

[This work was conducted under a private consulting agreement between W. Warren McHone and CEC, Inc. The analysis, opinions and conclusions presented in this document are those of the consultant and are neither endorsed nor sanctioned by the University of Central Florida or the UCF Department of Economics.]

Statistical Methods and Results the raw test results data that was provided by CEC listed the changes in miles per gallon (MPG) of fuel consumption and changes in the levels of various air pollution emissions (CO, HC, NOx and Opacity) for individual test vehicles.

There were eight sets of test results for fuel economy with a total of 57 individual observations. The eight test scenarios along with the range of test result outcomes, number of observations in each test, test results average and test results standard deviation are presented in Table 1. The table also provides an "Overall Summary" row that is based on a combined sample of individual observations from all eight tests. While the sample averages of fuel economy improvement in Table 1 provide a rough indication of expected improvement, the question that emerges is: "How much confidence can we have that the calculated average of a set of fuel economy improvement observations really provides a reliable measure of the fuel economy improvement of a vehicle selected at random from the test observations or a similar set of test observations?" To answer this question it is necessary to use standard statistical techniques to establish a minimum level of fuel economy improvement that should be expected with various degrees of confidence.

The key elements in establishing these confidence levels is the sample average and a summary measure of the extent to which the individual observations' deviate from the sample average. The latter is referred to as the "Standard Deviation" of the individual results in the sample. A simple example can illustrate why it is important to consider the Standard Deviation when comparing a sample's average outcome. Consider two hypothetical sets of test result observations from two different sets of test vehicles: Set A consists of the following fuel economy improvements for four individual vehicles: Al-28%; A2-32%; A3-29% and A4-31%. Set B consists of the following fuel economy improvements for three different individual vehicles: B1-10%; B2-70% and B3-10%. The average for Set A equals 30% and the average for Set B also equals 30%. However, in comparing the averages for the two sets of vehicles, we can be much more confident that a vehicle selected at random from Set A will be closer to SET A's sample mean than we can that a vehicle selected at random from Set B will be close to Set B's sample mean. Therefore, the mean for Set A is a much better measure of the expected level of fuel economy improvement than the mean for Set B.

Using the sample average, standard deviation and the appropriate standard probability distribution for our test results sample, it is possible to determine the minimum level of fuel economy improvement that corresponds to various levels of statistical confidence. These minimum levels of fuel economy improvement are presented at three progressively stringent levels of statistical confidence (95%, 97.5% and 99%) for each of the eight individual fuel economy tests and for the overall set of results from all tests combined are shown in the last three columns of Table 1. For an example of an interpretation of these confidence levels consider the entries in the last three columns of row 1 in the table: The entry in the "95% Confidence Level" column is 8.17%. This means that we can be 95% confident that a vehicle selected at random from this test group, or a similar test group, will have a minimum fuel economy improvement of 8.17% as the result of the use of Centron. Furthermore, if a higher level of confidence is required, we can be 97.5% that the improvement will be at least 7.7% and 99% confident that the minimum improvement will be 7.03%.

The "Overall Summary" row in Table 1 reveals that all test observations combined show an average fuel economy improvement of 18.05% with a standard deviation of 12%. The confidence levels for the total set of observations are 15.39% improvement with 95% confidence, 14.87% improvement with 97.5% confidence and 14.25% with a 99% level of confidence.

Pollution Reductions Emissions Tests Tables 2 through 5 provide statistical summaries for four different air pollutants and Table 6 brings together the "Overall Summary" rows for the four pollutants. Referring to Table 6, the average emissions reduction ranges from 35.46% for Opacity to 66.78% for HC (hydro carbons). The confidence levels for pollution reduction suggest that we can be highly confident that the emission reductions are substantial across all four pollutants. For example based on the test results we can be 99% confident that the reduction in CO will not be less than 36.58%, the reduction in HC will not be less than 52.63%, the reduction in NOx will not be less than 40.38% and the reduction in Opacity will not be less than 19.24%.

↑ Back to Top