Estimating Emissions from Open Molding of Composites

AP-42 Emission Factors

In March 1998, EPA withdrew the AP-42 emission factors for the open molding of composites, citing the availability of more accurate emission information being developed by industry. The AP-42 factors for composites have long served as the primary emission factor reference for state regulators and composites manufacturers. Emission factors are especially important for open molders, since most emissions from open molding are difficult to measure, and therefore stack testing is expensive, difficult to reproduce, and seldom used. EPA’s withdrawal of the AP-42 factors for open molding has left facilities and regulators unsure of how to estimate emissions for permitting, compliance demonstrations, TRI reporting, and other purposes.

(The AP-42 factors for non-open molding operations remain in effect.)

ACMA’s Emission Factors

When EPA withdrew the AP-42 factors for open molding, it posted on its web site a set of emission factors prepared by ACMA. In April of 1999, ACMA released a revised and enhanced set of factors, called the "Unified Emission Factors" (UEF). (Click here to go to the Unified Emission Factors and supporting technical documents.) ACMA is now working with EPA to gain formal adoption of the UEF as the new official AP-42 factors for the open molding of composites.

Facilities and regulators can readily use the UEF to estimate emissions for permits, TRI reports, and other purposes. Permits based on the UEF are readily enforceable, and several states have now granted permits to facilities that use the UEF. These facilities use the UEF to estimate emissions and to show compliance with the emission or operational limits contained in their permits. Click here for list of permitting authorities that have accepted the UEF.

The UEF are based on several independent studies of emissions from open molding processes. The UEF account for the impact on emissions of certain process variables: monomer content in resins and gel coats, resin application technology, vapor suppressants, worker training, and open versus covered cure.

While the UEF do not directly account for the impact of other process variables that may in theory impact emissions (notably gel time, temperature, amount of overspray, laminate thickness, air flow, and gun-to-mold distance), the UEF still result in accurate emission estimates. This is because these other process variables have little actual impact on emissions, or because they commonly vary around some typical value, which is assumed in the UEF. Also, these other process variables can not be reliably or accurately measured or estimated by facilities or regulators.

ACMA has compared recent data from stack tests on open molding facilities with emission estimates prepared using the UEF for these same facilities. This comparison shows that the UEF do accurately predict emissions from open molding operations. Click here for a comparison of stack test data and emission estimates derived using the UEF.

ORD’s "FRP" Emission Model

At the same time that ACMA has developed its emission factors, EPA’s Office of Research and Development (ORD) used much of the same data to develop a separate emission model for open molding. While ORD’s model (commonly called the "FRP Model") usually gives similar emission estimates to the UEF, the FRP Model requires many more inputs. Use of the FRP Model requires users to measure and input gel time, gun-to-mold distances, temperature, air flow, laminate thickness, and amount of overspray – precisely the process variables that are purposefully not included in the UEF.

The FRP Model may be a useful research tool, primarily because it accounts for all of these process variables. (ACMA believes that ORD has relied on inadequate data when defining the relationships between some of the process variables and emissions. The inadequate nature of the data may limit the usefulness of the FRP Model even for research purposes. Click here for ACMA’s correspondence with ORD on this topic.) However, ORD has recently suggested that its model serve as a permitting and real-world emission estimation method. ACMA has argued strongly that the very inputs that may make the FRP Model useful in a research setting make it unworkable for estimating emissions in actual manufacturing settings. This is because many of the inputs required for the FRP Model – again, laminate thickness, gel time, gun-to-mold distance, temperature, air flow, and amount of overspray – are impossible to precisely or reliably measure or control in actual manufacturing operations. Facilities with permits built around the FRP Model may find that they are out of compliance if any of the inputs vary from the values used for the model. Permits that require sources to comply with limits on these variables (for example certain gel times or certain temperatures) will not be practically enforceable.

For example, a facility might assume that its spray-gun-to-mold distance was 24 inches, use the FRP Model to obtain an emission estimate based on this assumption, and then use this emission estimate to apply for a permit. Later, a state inspector might determine that the distance was actually 30 inches, and find that the facility was not in compliance with its permit. This might happen, not because the facility had actually changed its operation, but because procedures for measuring the distance are not well defined (and, on a practical basis, can not be well defined) and are open to interpretation. There are similar measurement, interpretation, and compliance problems associated with all of the FRP Model inputs mentioned above.

How to estimate emissions from open molding

ACMA is working with ORD and the EPA office responsible for the AP-42 emission factors to demonstrate the technical and permitting difficulties of the FRP Model, and to further encourage EPA’s adoption of the UEF as the new AP-42 factors for open molding. In the meantime, facilities and regulators will have to determine how to estimate emissions. ACMA recommends that open molding facilities and regulators consider use of the Unified Emission Factors.

For TRI reports, facilities are required to estimate emissions "as accurately as possible," using the "best readily available data." While each facility is to make its own good faith evaluation of what source of emission information best fits its own operations, EPA has brought enforcement actions against companies that have used outdated EPA emission factors. Of course, if a facility has stack test data, or laboratory tests or engineering calculations tailored for its own operations, then this will often be the "best readily available data." However, ACMA believes that emission factors will be used for estimating emissions from most open molding operations.

Some facilities have older emission factors written into their permits. These companies may wish to contact their permitting authorities and ask about revising their permits to reflect the newer information. ACMA has explained to the TRI office at EPA that these companies may be caught between the conflicting requirements of the TRI program and their state permits, and has asked for some flexibility in the case of TRI enforcement actions. The TRI office has responded that they understand our concerns, but then cautioned that the law does require use of "best readily available data" for TRI.