There are two methods a facility can use to do this:
* risk-based look-up tables provided by EPA, or
* site-specific dispersion modelling and risk assessment.
The risk-based look-up tables correlate stack heights, source distances from the property line and health risks to acceptable maximum emission rates. A facility is required to estimate emissions and compare emission rates to those listed in the look-up tables to show that it qualifies for delisting. Site-specific modelling, however, is more complicated and requires additional effort; most facilities choose to use the look-up table method as a starting point and then, if the facility fails that analysis, move on to the site-specific modelling. The general perception of many facilities is that the look-up tables are applicable in most, if not all, situations. In reality, very few facilities, if any, will be able to use those tables as a low-risk delisting option.
EPA states, in both the PCWP rule preamble and rule text, that the look-up tables may not be applicable to all facilities. EPA specifically highlights capped sources (sources with rain hats or other obstructions on the emission point), complex terrain and building downwash as factors not accounted for in the development of the tables. The limits placed on the look-up table applicability by EPA are more qualitative than quantitative and are based on how the tables were originally compiled. They were developed by EPA using a simple atmospheric dispersion model (SCREEN3). Only a single stack can be modelled, and EPA used approximate ‘mid-point’ source parameters (height, velocity, temperature and diameter) across facilities in the PCWP source group to model a single representative stack. The meteorology used in the SCREEN3 model is worst-case because winds are assumed to blow from the emission source toward the receptor location at various speeds but the wind direction assumed by the model is not affected by local terrain.
Additionally, the approach to developing the look-up tables assumes that the single modelled stack exhausts vertically, building downwash (the effect on exhaust plumes by adjacent buildings) is not present and the surrounding terrain is flat. These three assumptions limit the tables’ applicability for most facilities attempting low-risk delisting. The assumption that all sources exhaust vertically affects both capped sources and horizontally exhausting ones. If a facility has either of these, the look-up tables may not apply, since the dispersion of a capped or horizontally exhausting source would be poorer than the source representation assumed in the development of the tables. Buildings close to sources can affect the exhaust plume. Their downwash effect can result in ‘pulling’ the plume downward. This can lead to higher modelled ground-level concentrations. Since the look-up tables do not account for any building downwash, a facility wishing to use them cannot have any major sources that are affected by it.
The general rule of thumb (not accurate in all situations) is that a source exhaust point must be two and half times higher than the adjacent buildings in order to be free of downwash. Press vents which are not significantly higher than the press building, ground level bag houses, process water in ponds or open impoundments and a host of other emission points at a PCWP facility are typically caught in the building downwash of the manufacturing buildings. Local terrain may also prevent the use of the look-up tables. The flat terrain assumption in the tables is not a conservative one. Many facilities will have some terrain features close to the facility, including complex terrain (ie terrain elevations higher than the facility emission points). Elevated terrain will generally cause higher modelled concentrations than if flat terrain is assumed. If a facility has capped or horizontal sources, sources which are not two and a half times the height of the nearest buildings, or has elevated terrain close by, the look-up tables may not be applicable.
If a facility determines that the tables are not appropriate, or if a local agency determines that the facility may not use them, the remaining option for low-risk delisting is site-specific modelling. In site-specific dispersion modelling, a more refined model is used which has the ability to address all the shortcomings of the much simpler look-up table model. A dispersion model allows a facility to represent horizontal exhausts, building downwash and complex terrain. After a close examination of the methods used to construct the look-up tables, it is apparent that they are not applicable to most PCWP facilities. In practice, they may only be applicable to a very small number of facilities attempting to qualify for delisting from the PCWP MACT rule. Where the tables are applicable, a very narrow set of criteria must be met and they must be approved by the local agency. Note that the separate EPA MACT rules applicable to boilers and process heaters contain similar risk assessment look-up tables that were similarly developed and have similar limitations to those discussed.
Site-specific modelling can be applied to almost all facilities and there are no rigid parameters a facility must meet in order to conduct that modelling. Those wishing to use the look-up tables as a delisting option should seek local agency approval well before submitting the delisting demonstration. Alternatively, facilities may wish to consider using site-specific dispersion modelling and risk assessment. Approval of the delisting demonstrations is required by October 1, 2008. Geoff Scott is a project engineer with SECOR International. SECOR is a multimedia environmental consulting firm with offices nationwide and membership in the National Council for Air and Stream Improvements.
