Why emission factors are usually better than source tests

As we discussed, some disadvantages of testing:

The cost of source testing is obviously dependent on the number of stacks that must be sampled. For two to twelve stacks, the cost of testing would range from $15,000 to $50,000 for one just one day of testing. EPA's rule-of-thumb for the cost of testing is $10,000 per stack, but multiple stacks offer substantial savings due to the economy of scale. In addition to the direct cost of testing, a facility must also pay for test preparation and supervision of material usage. These additional costs can also be significant, especially if temporary scaffolding must be installed around several stacks on a steep roof.

Source testing has significant error. The emission rate is determined from a flow measurement and a concentration measurement that both have inherent errors. These errors could easily combine to at least plus or minus 15% overall. Hence, two consecutive tests of the same stack using the same equipment, personnel, methodology, and laboratory would probably return two different test results. This is why several consecutive tests are needed to determine an accurate average result. Unfortunately, a facility cannot afford to conduct multiple tests. A single source test normally provides only a single data point, so the accuracy of such a test is never known.

Source testing assumes tight process enclosures, but this is seldom the case. Emissions may escape from the process being testing (negative bias) or outside emissions from other processes may be drawn into the test stack (positive bias). I usually recommend that all of the emission points at a plant should be tested at one time. I have the equipment and personnel to accomplish this, but most testing companies prefer to test a facility in a piecemeal fashion.

Source testing to verify an emissions factor also requires accurate measurement of the associated material usage and styrene content. The usage must correspond directly to the emissions. This is often difficult to achieve in practice. Unfortunately, the test results would have little real meaning without accurate associated usage & content data.

Most composites plants emit styrene from lamination and acetone from cleanup operations. Acetone is a non-VOC. Therefore, Method 25A cannot be used, because the acetone would falsely register as styrene emissions and exaggerate the test results. MMA from some gelcoats also presents the same problem although MMA is also a VOC. Method 18 must be used to measure just the styrene (and perhaps MMA). However, sampling for styrene with Method 18 is very tricky. Although I have ten years of styrene test experience, and I still approach each test with caution. Most testing firms are set up for Method 25A, not Method 18.

Let me know if you need more information.

Rob Haberlein, PhD, PE