innovative environmental solutions, inc.

MEMORANDUM

Date:	July 27, 2011

To:	Lisa Beal, INGAA	

From:	Jim McCarthy and Jeff Panek, IES Inc.

Re:	Concentration versus distance profiles from reciprocating internal
combustion engines – Dispersion modeling results for two example
engines from previous INGAA modeling.

Emission impacts from reciprocating internal combustion engines (RICE)
have been discussed in conjunction with ongoing negotiations regarding
the RICE NESHAP, 40 CFR 63, Subpart ZZZZ.  In a recent conference call,
it was noted that RICE dispersion modeling could be used to show
concentration degradation versus distance plots for example RICE.  In
the call, it was noted that previously completed modeling may be
available to demonstrate how modeled ground-level impacts decrease with
distance from the stack.  The concentration versus distance plots could
thus support approaches being discussed for establishing a distance
criterion for RICE located in sparsely populated areas.  

IES completed dispersion modeling for INGAA in 2002 and early 2003
during development of the original RICE NESHAP.  That rule was proposed
in December 2002 and promulgated in June 2004.  Although the rule was a
technology-based standard, modeling was conducted to understand the
range of modeled impacts from example engines to support comments on the
proposed rule and discussion associated with the Industrial Combustion
Coordinated Rulemaking process.  Modeling was completed for two example
reciprocating engines, and the effect of stack height on modeled impacts
was assessed.  Dispersion modeling was completed using ISC3, the
preferred EPA model at that time.  Assumptions associated with the two
examples are: 

Example engine #1 – Natural gas-fired lean burn engine

800 hp capacity (modeling based on full load operation)

30% efficiency (i.e., 8,470 Btu/hp-hr heat rate based on lower heating
value)

8% oxygen in the exhaust (i.e., commensurate with 4-stroke lean burn
engine)

700 oF exhaust temperature

15 inch stack diameter

18 ft stack height (“baseline” – a range of stack heights were
investigated)

Exhaust velocity based on the above parameters of 51.1 ft/sec

Example engine #2 – Natural gas-fired rich burn engine

1,000 hp capacity (modeling based on full load operation)

27% efficiency (i.e., 9,400 Btu/hp-hr heat rate based on lower heating
value)

0.5% oxygen in the exhaust (i.e., commensurate with 4-stroke rich burn
engine)

1,000 oF exhaust temperature

12 inch stack diameter

12 ft stack height (“baseline” – a range of stack heights were
investigated)

Associated exhaust velocity based on the above parameters of 91.5 ft/sec

The concentration versus distance plots for the two example engines are
presented in two figures below.  The concentration plotted is the
maximum modeled 1-hour average concentration associated with RICE
emissions.  These graphs were developed in the 2002 – 2003 timeframe
and were not created for this memo (e.g., not updated using AERMOD, the
currently preferred model).  The range of stack heights shown were
modeled to assess the effect of stack height, and the taller heights
modeled are not indicative of actual installations or stacks that could
actually be installed.  Flat terrain was assumed for the modeling runs. 

Modeling was conducted using a “unitized” emission rate of one gram
per second.  Pollutant-specific impacts can easily be scaled based on
proportionality to the unitized emission rate. Thus, the modeled impact
(shown on the y-axis) is not based on any particular exhaust pollutant. 
This is a standard approach for dispersion modeling when pollutant
in-plume reactions are not assumed – i.e., the exhaust species does
not change from the stack to the point of impact – because it
eliminates the need for multiple, pollutant specific model runs.  In
contrast, this approach would not be appropriate for modeling NOx
impacts when NO to NO2 conversion reactions are assumed as the plume
disperses downwind.  When applying a unitized emission rates for
modeling, impacts for an array of pollutants can be assessed by scaling
the modeled impact based on the actual emission rate for the
pollutant(s) of interest, as compared to the unitized rate.  Actual RICE
hazardous air pollutant emission rates are much lower than 1 g/s.  Thus,
the scale on the y-axis is not shown in the figures below because it is
not relevant – i.e., the plots are intended to demonstrate how
emission impacts degrade with distance rather than assess the impact for
comparison to a threshold level.  

The first figure shows concentration versus distance plots for the
example lean burn engine.  The second figure presents the rich burn
engine results. The RICE NESHAP discussions have considered different
distances. When interpreting downwind impacts with distance on the
x-axis in the figures below, 402 meters is equivalent to a quarter mile
(i.e., the impact at 400 meters is indicative of a ¼ mile radius).

As expected, both figures show that increasing the stack height can
significantly reduce the maximum modeled impact by minimizing or
eliminating the effects of structure-related downwash.  More
importantly, for the recent RICE NESHAP discussions, the distance plots
show that higher modeled impacts (associated with shorter stacks)
decrease rapidly with distance and the results converge over a
relatively short distance.  

For the example lean burn engine, maximum model predicted impacts from
all stack heights investigated occurred at a distance of less than 50
meters (equivalent to 164 feet, 55 yards and 1/32nd of a mile).   At a
distance of 100 meters (equivalent to 328 feet, 109 yards and 1/16th of
a mile), higher impacts (from shorter stacks) are greatly diminished and
at a distance of 150 meters (492 feet, 164 yards, just under 1/10th of a
mile) the impacts are minimal and indistinguishable for the broad range
of stack heights modeled.  

For the example rich burn engine, the maximum impacts from all stack
heights investigated are predicted to occur at a distance less than 130
meters (equivalent to 427 feet, 142 yards and 1/12th of a mile)
primarily attributable to building “downwash”.  At 200 meters (656
feet, 2190 yards and 1/8th of a mile), the higher impacts from shorter
stacks are greatly diminished and modeled impacts for various stack
heights converge.  At 400 meters (1312 feet, 437 yards, ¼ mile) the
impacts are minimal and indistinguishable for the broad range of stack
heights modeled.

The figures demonstrate the dispersion and rapid decrease in
concentration with distance from “typical” engines, and support the
use of a distance criterion on the order of ¼ mile or less.  

  PAGE  3 

 

