May 21, 2007

NOTE

SUBJECT:	Potential Approach to Establish Firing Rate Limits on
Emission-Comparable Fuel

FROM:	Bob Holloway, OSW, USEPA

TO:		Docket ID No. EPA-HQ-RCRA-2005-0017

We are proposing to expand the comparable fuel exclusion for fuels that
are produced from hazardous waste but which generate emissions that are
comparable to emissions from burning fuel oil when such fuels are burned
in an industrial boiler.  Such fuel would be called emission-comparable
fuel (ECF).  ECF would be subject to the same specifications that
currently apply to comparable fuels, except that the specifications for
benzene, toluene, and the oxygenates listed in Table 1 to §261.38 would
not apply.  This paper describes a potential approach to establish
firing rate limits on emission-comparable fuel (ECF) that may be
necessary to ensure that ECF emissions are either:  (1) comparable to
emissions from oil-fired industrial boilers; (2) at de minimis levels;
or (3) are protective of human health and the environment.

Under the approach, we would identify a target emission level for each
compound of concern, estimate a destruction and removal efficiency (DRE)
for the compound, and calculate a maximum ECF firing rate as a function
of the concentration of the compound in the ECF. 

Target Emission Levels

The target emission levels would be identified as:

For compounds of concern for which we have emissions data from oil-fired
industrial boilers, the target level would be the highest test condition
average or a de minimis level, whichever is higher;

For compounds of concern for which we have only hazardous waste boiler
emissions data, the target level would be the highest test condition
average or a de minimis level, whichever is higher; and

For compounds of concern for which we have neither oil-fired boiler nor
hazardous waste boiler emissions data, the target level would be a de
minimis level.

See Table 1 which presents the recommended target emission levels,
corrected to 7% oxygen.

We believe that it is reasonable to consider an emission level of 20
ug/dscm to be de minimis.  This emission level is comparable to
approximately 0.01 ppmv propane equivalents for the high molecular
weight compounds of concern, and is 3 orders of magnitude lower than the
10 ppmv total hydrocarbon emission limit the Agency has established for
liquid fuel boilers that burn hazardous waste.  See §63.1217(a)(5)(ii).

We also believe it is reasonable to use the highest test condition
average emission level for oil-fired boilers as the target emission
level rather than the average test condition average or the high end
(e.g., 95th percentile) test condition average.  Using the average test
condition average would be inappropriate because the principal rationale
for the exclusion is to ensure that emissions are comparable to oil
emissions.  Therefore, it is appropriate to consider the range of oil
emissions rather than the average emission.  In addition, we are
recommending using the highest test condition average rather than the
95th percentile test condition average because the data base is not
robust.  If additional data were available, both the 95th percentile and
the highest test condition average would likely increase.  Consequently,
using the highest test condition average is a reasonable approach to
represent the high end of the range of emissions.

We believe it is reasonable to use the highest test condition average
emission level for hazardous waste boilers as the target emission level
(i.e., absent oil emissions data and provided that the test condition
average is not de minimis) because we have determined that emissions
from hazardous waste combustors are generally protective of human health
and the environment.  See 70 FR at 59535.  Ensuring that ECF emissions
are protective is another primary rationale for the proposed exclusion.

Destruction and Removal Efficiency

As a default, we believe it is reasonable to estimate a DRE of 99.99%
for compounds of concern that have a Thermal Stability ranking of Class
I or Class 2 (i.e., benzene, toluene, and methyl methacrylate) and a DRE
of 99.995% for the other compounds.  The Thermal Stability ranking is a
principal tool for selecting difficult to destroy compounds for DRE
testing required to establish operating requirements for hazard waste
combustors.  We note that, while we have several DRE data from hazardous
waste boilers for Class 1 and Class 2 compounds showing DREs below
99.995% at relatively low feedrates of principal organic hazardous
constituents (POHCs), our (albeit limited) DRE data for Class 3-7
compounds are all greater than 99.995%.  See Figure 1.  Of the nine
compounds plotted in the figure, carbon tetrachloride, acetophenone, and
1,2 dichloropropane are ranked in Thermal Stability Class 3 or below. 
All of the other compounds that are plotted are ranked in Class 1 or 2.

We also believe that it is reasonable to conclude that DRE increases
with an increase in feedrate of the target compound.  It is common
knowledge that feedrates of POHCs must be high enough to avoid DRE
failures attributable to stack method or analytical method imprecision
and the baseline level of products of incomplete combustion.  A recent
paper by Brukh, et al, appears to support this view.  Moreover, a plot
of hazardous waste boiler DRE run data versus feedrate MTEC indicates a
general trend toward higher DREs as feedrates increase for those
compounds for which we have DRE data over a range of feedrates.  See
Figure 2.

It appears from these figures that, when MTECs exceed 1.0E+07 ug/dscm,
DRE exceeds 99.999% for all compounds.  And, it appears that, for MTECs
in the range of 5.0E+06 to 1.0E+07, DRE exceeds 99.995% for all
compounds, except for two DRE runs for hydrogen cyanide.  We believe our
conclusion is nonetheless valid because hydrogen cyanide has atypical
thermal stability, even for a Class 1 compound, and is much more
difficult to destroy than benzene, which has the highest thermal
stability of the compounds of concern.  Consequently, it may be
appropriate to consider this feedrate/DRE relationship to identify
potential ECF firing rate restrictions.

Potential ECF Firing Rate Limits

We have reviewed the ACC survey to determine the concentrations of the
compounds of concern that may be present in ECF.  Toluene may be present
in ECF at concentrations ranging from 5% to nearly 100%, and benzene may
be present at concentrations up to 25%.  Other compounds appear be
present at lower concentrations; some appear to be present only at very
low concentrations (but levels that nonetheless exceed the
specifications in Table 1 to §261.38).  See Table 2.  Although ECF may
often be blended (after meeting the specifications) with other fuels
before burning, we assumed a worst-case scenario where these waste fuels
were burned without blending.  We considered these potential
concentrations to calculate potential ECF firing rate limits considering
the estimated DREs and target emission levels discussed above.

As expected, at low compound concentrations in the ECF, the ECF firing
rate would not be restricted (i.e., other than the restrictions that
would apply as a basic condition of the exclusion—25% maximum firing
rate if the benzene or acrolein concentration exceeds 2%, and 50%
maximum firing rate for all other ECF).  See Table 3.  

We noted an anomalous situation for most compounds, however, where the
firing rate limit first decreased as feedrate increased, and then at
higher feedrates, the firing rate began to increase.  This was caused by
our assumption that DRE increases in a step-wise function rather than,
as likely, in a smooth progression as feedrate increases.  For example,
we estimated DRE at 99.995% when the MTEC is 9.9E+06, and at 99.999%
when the MTEC is 10E+06 (1.0E+07).  

Clearly, this is not a realistic representation of how DRE relates to
MTEC.  To address this concern, we could, for example, consider whether
it is appropriate to use a best-fit curve of the benzene data to develop
a relationship between DRE and MTEC.  Benzene may be an appropriate
compound to select to define the relationship because it ranks the
highest on the thermal stability index of the compounds for which we
have DREs over a range of feedrates, it has the highest ranking for the
compounds of concern, and it is the third highest ranking compound in
the Thermal Stability index, ranking higher than 341 other compounds.

Table 1

TARGET EMISSION LEVELS FOR FEEDRATE RESTRICTIONS

Table 2

ACC SURVEY:  CONCENTRATIONS OF BENZENE, TOLUENE, AND OXYGENATES IN WASTE
FUELS THAT MAY BE CANDIDATES FOR EXCLUSION BUT THAT EXCEED THE
SPECIFICATIONS IN TABLE 1 TO 261.38

Table 3

EXAMPLE RESTRICTIONS ON COMPOUND CONCENTRATIONS AND ECF FIRING RATES TO
ENSURE THAT TARGET EMISSION LEVELS ARE ACHIEVED

Figure 1

DRE VERSUS POHC FEEDRATE FOR HAZARDOUS WASTE BOILER TEST RUNS

Figure 2

DRE VERSUS FEEDRATE FOR HAZARDOUS WASTE BOILER TEST RUNS:  BEST FIT
LINES FOR COMPOUNDS WITH SUBSTANTIAL RANGE IN FEEDRATE

--DRE TENDS TO INCREASE AS POHC FEEDRATE INCREASES--

 	The compounds of concern are benzene, toluene, and the oxygenates
listed in Table 1 to §261.38.

 	For background support for these analyses, see the memorandum from
Lucky Benedict, EERGC, to Bob Holloway, USEPA, entitled “Background
Information and Sample Calculations for Potential Approach to Establish
DRE based Firing Rate Restrictions for ECF,” dated May 21, 2007.

 	Please note, however, that high statistical outliers were not
considered.  See See USEPA, “Draft Technical Support Document for
Expansion of the Comparable Fuel Exclusion,” May 2007, Appendix C.

 	See USEPA, “Draft Technical Support Document for Expansion of the
Comparable Fuel Exclusion,” May 2007, Section 5.1.

 	The Thermal Stability ranking classifies (generally) hazardous
compounds according to their gas-phase thermal stability under
oxygen-starved conditions.  Compounds are ranked according to the
temperature required to destroy 99% of the compound in 2 seconds under
oxygen-starved conditions.  `See USEPA, “Guidance on Setting Permit
Conditions and Reporting Trial Burn Results, Volume II of the Hazardous
Waste Incineration Guidance Series,” January 1989, Table D-1.

 	Note that methanol has been used as a POHC for DRE testing and is
plotted in the figure even though its thermal stability has not been
classified.

 	See USEPA, Operational Parameters for Hazardous Waste Combustion
Devices,” October 1993. Section 4.3.2.1.

 	See R. Brukh, R. Baret, and S. Mitra, New Jersey Institute of
Technology, “The Effect of Waste Concentration on Destruction
Efficiency During Incineration,” Environmental Engineering Science,
Vol. 23, No. 2, 2006.  The authors conducted experiments in a small,
well-stirred reactor involving the combustion of methylene chloride
(CH2Cl2) with ethylene (C2H4) as the primary fuel at residence times of
5-12 ms and temperatures of 1400- 1750 K (2050 -2700 °F).  Experiments
were done at both fuel rich and fuel lean conditions.  CH2Cl2
concentrations were low (2- 1350 ppm by volume in the main feed.).  The
authors modeled the combustion of methylene chloride, methyl chloride
(CH3Cl), and benzene.  They show limited experimental data for CH3Cl and
C6H6 from previous work.  The authors’ hypothesis is that higher
concentrations of POHC contribute additional radical fractions and the
overall result is a higher destruction efficiency.  This work would
support higher DREs at higher feedrates if the results can be
extrapolated to the higher POHC concentrations of concern to us and the
higher residence times for hazardous waste combustors.

 	MTEC means maximum theoretical emission concentration (ug/dscm) and is
an approach to normalize feedrate for various size boilers.  It is
calculated as the mass feedrate divided by the stack gas flowrate.

 	The temperature required to destroy 99% of hydrogen cyanide at
substoichemetric conditions in 2 seconds  is more than 1500 C, while
the temperature required to destroy 99% of benzene is 1150 C.  The
compound of concern with the next highest thermal stability, toluene,
needs a temperature of 895 C for 99% destruction.  

 	Letter from American Chemistry Council (Carter Lee Kelly, Leader,
Waste Issues Team, and Robert A. Elam, Director, Regulatory Affairs,
Waste Issues Team) to Robert Springer and Matt Hale, USEPA, dated
November 24, 2003

 PAGE   

 PAGE   8 

