EPA 530-SW-87-014A
PB87-]73829
BACKGROUND INFORMATION DOCUMENT FOR THE
DEVELOPMENT OF REGULATIONS TO CONTROL THE
BURNING OF HAZARDOUS WASTES IN BOILERS
AND INDUSTRIAL FURNACES - VOLUME I, INDUSTRIAL
BOILERS
Enginoering-Science
Fairfax, VA
Jan 87
-------
BACKGROUND INFORMATION DOCUMENT FOR THE DEVELOPMENT OF
REGULATIONS TO CONTROL THE BURNING OF HAZARDOUS WASTES IN
BOILERS AND INDUSTRIAL FURNACES
VOLUME I
INDUSTRIAL BOILERS
Submitted to
U.S. Environmental Protection Agency
Office of Solid Waste
401 M Street, S.W.
Washington, D.C. 20460
January 1987
FA035/50A-E
Submitted by
Engineering-Science
Two Flint Hill
10521 Rosehaven Street
Fairfax, Virginia 22030
-------
DISCLAIMER
Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
-------
TAB^E OF CONTENTS
SECTION 1 SUMMARY AND CONCLUSIONS '"-1
SECTION 2 INTRODUCTION 2-1
Objective 2-1
SECTION 3 CHARACTERIZATION OF INDUSTRIAL BOILERS FOR BURNING .
HAZARDOUS WASTES 3-1
Description of Existing Boiler Population 3-1
Existing Emissions from Industrial Boilers 3-9
Currently Installed Emission Controls 3-12
Potentially Applicable Control Equipment 3-26
References 3-34
SECTION 4 DESTRUCTION EFFICIENCY AND CONTROL TECHNIQUES 4-1
Evaluation of Test Methods 4-1
Test Burn Results 4-11
Modifying Boiler Systems to Burn Hazardous Waste 4-83
Boiler Operating Conditions Providing Acceptable DREs 4-98
References 4-102
SECTION 5 CHARACTERIZATION OF WASTE STREAMS 5-1
Sources of Data 5-1
Quantity of Hazardous Wastes Generated 5-7
Characterization of Waste Streams 5-7
Quantity of Hazardous waste Burned in Industrial
Boilers 5-20
References 5-23
SECTION 6 COST ELEMENTS FOR REGULATORY ANALYSIS 6-1
Conventional Fuel Prices 6-1
Boiler System Modification Costs 6-2
References 6-47
11
-------
SECTION 1
SUMMARY AND CONCLUSIONS
The initial hazardous waste management facility standards promulgated
on May 19, 1980 under the Resource Conservation and Recovery Act (RCRA),
controlled the burning of hazardous waste i. incinerators but exempted the
burning of hazardous waste for the purpose of energy recovery. This exemp-
tion for boilers and other cououstion devices was allowed by EPA because the
Agency had not investigated the exr.ent of the practice, the risks that may
be posed to human health, the environment, or regulatory alternatives.
Since the promulgation of the incinerator rule, EPA has undertaken a re-
search program to obtain the information needed r.o develop and evaluate al-
ternatives for regulating the burning of hazardous waste in boilers and indus-
trial furnaces. The results of this research are presented in this dov-jumenr
and two companion volumes, which together comprises the Background Informa-
tion Document (BID) for use by E?A in its Regulatory Impact Analysis (RIA) of
the practice of burning hazardous waste in industrial boilers and industrial
furnaces. This volume contains background information on industrial boilers
as related to hazardous waste, burning. The practice of burning hazardous
waste in industrial furnaces is addressed in Volume II. An assessment of the
risk associated with burnirg hazardous waste in industrial boilers and fur-
naces is made in Volume III of the BID.
One of the major tasks of EFA's regulatory development efforts was to
characterize industrial boilers in terms of parameters believed to affect
their hazardous waste burning potential and/or which impact the selection
of regulatory alternatives. This was dcr.e to provide an understanding of
industrial boilers needed to develop practical regulatory approaches. The
boilers were characterized in terms of their design, fuel usage, air emis-
sions, and control techniques as well as their population distribution.
Another major task completed by EPA in its regulatory development ef-
forts was to determine the hazardous waste incineration performance capa-
bility of boile.'s. EPA Conducted field testing on a total of 15 industrial
boilers. The test program was designed to: (1) determine if boilers oper-
ated under steady-state conditions to achieve maximum combustion efficiency
could achieve 99.99% destruction and removal efficiency (DRE) of principal
organic hazardous constituents (POHCs) in the waste; and (2) determine how
changes in operating conditions (e.g., waste firing rates, boiler load, ex-
cess flue g;»s oxygen levels) would affect the boiler's ability to achieve
99.99% DRE Df POHCs — so-called parametric testing. The boilers tested
1-1
-------
ranged in size from a small 8 million Btu/hr fire-tube ooiler to a 2?0 million
3tu/hr water-cube boiler. .'he hazardous waste burned ranged from toluene and
nethancl wastes with a heating value of 18,500 Btu Ib ( si.T..lar to heating val-
ue 3f No. o fuel oil) to a methyl acetate waste with a heating vaJ.ue of less
than half that of No. 6 fuel oil. Both these fuels were spiked vr. th chlori-
nated organics for test purposes. Auxiliary fue..s burned included natural
gas, No. 6 fuel oil, coal, and waste woqd.
In total, over 100 individual stack tests were conducted. The results
indicated tnat the 99.99% destruction and removal efficiency (ORE) standard
for the tested POHCs is achievable. These POHCs included some of the more
difficult compounds to destroy such as carbon tecrachloride, chlorobenzene,
trichloroethylene, and tetrachloroethylene. The; 99.99% ORE of POHCs was
found to be achievable under various operating conditions including load
changes, waste feed rate changes, and excess air rate changes for boilers
cc-firing hazardous waste fuels with fossil fuels. There appears to be
no direct correlation between combustion efficiency (as evidenced by smoke
emissions) and POHCs destruction. Boilers operated under poor combustion
efficiency still achieved 99.99% CRE of POHCs. Wher the boilers were oper-
ated at maximun oomoustion efficiencies, the DREs exceeded 99.99%.
These results should not be inteipreted to indicate that any boiler
burning any hazardous waste will achieve 99.99% ORE. Not all parameters
could be tested at the boilers where the operating conditions were varied.
For example, the maximum waste firing rate tested was 56% of the total heat
input, and the boilers we-e not operating at loads below 25%, the heating
values of the wastes were all greater than 11,000 Btu/lb, and excess oxygen
levels did not exceed 10%. Therefore, we do not fully know how narrow the
envelope of operating conditions may be to ensure both peak combustion ef-
ficiency and 99.99% DRE for a boiler operating at the extremes of "steady
state" operation.
Typical chlorinated products of incomplete combustion (PICs) found
during the boiler testing included chloroform, trichloroethane, tetrachlor-
oethylene, dichloronethane, chlorcrcethane, and carbon tetrachlondo . Two
non-chlorinated PICs always found were benzene and toluene. The measured
chlorinated PIC emissions were generally of the same order of magnitude as
measured during the conventional fuel tiring. They were also generally of
the same order of magnitude as the POHC emissions. Exceptions to these
generalizations occurred during sootblowing and waste atomizer upset por-
tions of the parametric testing.
Several potential problems should be noted related to the testing pro-
gram. There is a substantial degree of uncertainty when trying to quantify
emissions of unburned organics. The test results may have over- or under-
estimated the unburned organic emissions attributed to the burning of haz-
ardous wastes. The potential for overestimating these emissions exists be-
cause the hazardous waste was co-fired with fossil fuels. The burning of
fossil fuels produces similar PICs as the burning of hazardous wastes. Some
organics found in the -sst samples could have been the result of contamina-
tion of the sampling train absorbent or the use of laboratory solvents. It
is also possible that the PIC emissions could have been underestimated, since
only Appendix VIII of the Code of Federal Regulations pollutant PICs were
1-2
-------
identified ithere are in fact mere PlCs than the 100 pollutants the GC/MS was
used to quantify). £ven thougn additional research is needed in order to
fully understand tr.-. comcustion reactions, the available data suggests that
health risks vpresented 1:1 Volume 3) posed by PIC emissions are probably not
significant when a D'*E of 99.99% of POHCs is achieved and the combustion
efficiency is good.
Several surveys, -were completed by EPA and other organizations to help
quantify and characterize the hazardous waste being generated and that which
may be burned in industrial boilers. A review of these surveys indicate that
approximately 264 netnc tons of hazardous wastes regulated by RCRA were gen-
erat-?d in 1981. Total burnable nazardous waste is estimated at 160 million
metric tons per year. Of this total, it is estimated that Itss than 4 mil-
lion metric tons were burned in industrial boilers in 1982.
Cost elements needed for an economic impact analysis of regulatory al-
ternatives were developed. Capital and operating and maintenance costs were
provided for retrofitting industrial boilers for burning hazardous waste.
This BID includes cost elements for:
o Equipment to pretreat the waste by blending, straining, and thermal
treatment for viscosity acliustment.
o New or modified burner gons to fire the waste.
o Equipment for combustion controls.
o Equipment for monitoring waste feed rates, oxygen, and carbon mon-
oxide levels.
o Equipment for controlling particulate and gaseous stack emissions.
In addition to these cost elements, fuel costs were compiled and projected
through 2010 for natural gas, residual fuel oil, distillate fuel oil and
coal for the r.en EPA regions of the country.
1-3
-------
50A
SECTION 2
INTRODUCTION
The Resource Conservation and Recovery Act (RCRA) requires EPA to develop
regulations for the storage, handling, aid disposal of hazardous wastes. It
also requires that EPA encourage in its regulations the recycling of wastes.
On May 19, 1980, EPA issued requirements for hazardous waste incinerators.
At that time, the burning of hazardous wastes in boilers was not regulated
because SPA lacked sufficient information co determine the extent of the risk
to public health posed by such burning, as well as the extent of regulatory
controls that would be necessary to address 'che risks. Since 1980, EPA has
researched the nature and extent of the practice of burning hazardous waste
as fuel T.n boilers. A boiler is defined in 40 CFR 260.10 as an enclosed de-
vice using controlled flame comtustion with specified design and operating
char tctenstics related to the recovery of energy. This document presents
the results of research efforts for industrial boilers.
OBJECTIVE
The objective of this project was to provide sufficient research and
investigation into the concept and practice of burning hazardous wastes in
industrial boilers such that EPA could develop and evaluate regulatory alter-
natives. The information and data collected by this investiqation constitutes
a Background Information Document (BID) for use by EPA in praparing its Regu-
latory Impact Analysis (RIA).
The investigation undertaken by EPA to obtain the information needed for
its regulatory development effort covering hazardous waste burning in indus-
trial boilers entailed tne following tasks:
o Establishing an understanding of boilers by characterizing them in
terms of those parameters thought to influence regulatory decisions.
o Field testing a variety of industrial boilers to establish their
capacity for destroying hazardous waste.
o Assessing the risk to human health and the environment associated
with burning hazardous waste in industrial boilers.
o Defining the extent of the practice of burning hazardous waste in
industrial boilers.
o Compiling the cost of items to be included in an economic impact
analysis of regulatory options.
2-1
-------
These tasks were completed by a number of EPA groups and tnej.r contractors.
A summary of the results of the field tes'.ing and hazardous waste practice
surveys are documented in tnis volume of the BID. More detailed descriptions
of these efforts may be found in the reports referenced throughout tnis docu-
ment. A summary of risk assessment results are presented in Volume 3 of this
BID. The other task results are presented along with the summaries of the
test program ana waste usage surveys in the following sections of tnis docu-
ment. The ir.formation is organized as described in the following paragraphs.
Section 3 characterizes industrial boilers in terms of their design,
fuel usage, population, air pollution emissions, and control techniques.
This characterization is mads to provide a basic understanding of boilers
needed to develop a practical regulatory approach.
Section 4 describes the results of tests conducted in order to measure
the performance of industrial boilers in burning hazardous wastes. This
performance is described in terms of acnieved ORE of PCHCs, and the emis-
sions of PICs, particula-.e matter, HC1, metals, and combustion gases.
Section 5 characterizes the various waste streams used as supplementary
fuel in industrial boilers.
Presented in Section 6 are cost data needed by EPA to conduct an Econo-
mic Impact Analysis (EIA) of regulatory alternatives. Costs are presented
for those irems used in analyzing the economic impact in terms of how the
fuel and waste disposal savings of hazardous waste burners are altered by the
vanour regulatory alternatives. The items being provided in this document
fall into three major categories: (1) conventional fuel prices, (2) costs to
modify the boiler system to fire the waste, and (3) the major operating and
maintenance costs associated with burning hazardous waste.
2-2
-------
50A
SECTION 3
CHARACTERIZATION OF INDUSTRIAL BOILERS
FOR BURNING HAZARDOUS WASTES
An essential element for defining the impact of regulating the disposal
of hazardous wastes in industrial boilers is a characterization of the Doil-
ers for this use. Not all types of boilers are suitable for burning every
type of waste, e.g., gas-fired and oio.-fired units are not suited for burn-
ing solid hazardous waste materials. An understanding of the different boil-
er types is therefore necessary to develop a practical regulatory approach.
Also, since the regulatory impact will depend on the number of boilers cap-
able of burning hazardous waste, it is necessary to describe the boiler pop-
ulation and how this population is distributed by size, type, fuel, and ca-
pacity. Finally, the current and future extents of the practice of burning
hazardous waste in boilers must be estimated. This section presents the ex-
isting boiler population, the estimated air emissions, and the types of air
pollution control devices presently used to abate these emissions. Est_;nates
of the quantity of hazardous waste burned in these boilers are discussed in
Section 5 of this document.
DESCRIPTION OF EXISTING BOILER POPULATION
Industrial boilers are generally described in terms of the heat trans-
fer configuration and the fuel burned. The first term defines the physical
structure of the boiler while the latter indicates the fuel type and firing
mechanism.
Heat Transfer Configurations
There are three basic types of heat transfer configurations: water-tube,
fire-tube, and sectional which are also referred to as "cast iron" from the
material of construction used. A brief description of these three types fol-
lows.
Water-tube boilers are designed to transfer heat from the combustion
gases flowing over the outside of the tubes to water, steam, or other fluid
contained inside the tubes. Because the tubes in these units are of relative
small diameter, they provide rapid heat transfer. They are available in many
sizes, generally in the range of 15 x 106 to 1500 x 106 Btu/hr. All boilers
greater, than 50 x 106 Btu/hr are of this type. These boilers generate high-
pressure, high-temperature steam up to 12,000 KPA (1740 psig) and 810°K
(1000°F) (Paference 1 ).
3-1
-------
In fire-tube boilers, the hot combustion gases flow through the ir.b
of tubes with water, steam, or other fluid contained outside the tubes.
Fire-tube units are not available m capacities as large as water-tube unit;
with most less than 20 x 10° BwU/hr (Reference 1). They constitute Lhe lar-
gest share of small and medium-size industrial boilers. Because they are
susceptible to structural failure when subject to large variations in steam
demand, they are generally used where loads are relatively constant.
Sectional or cast iron boilers employ irregularly shaped heat excnangers
and hence cannot be classed as either water-tube or fire-tube. Hot combus-
tion gases are directsd through some of these passages, transferring heat
through metal walls to water or steam i^; other passages. These units are
manufactured in identical sections which can be joined together according to
the needs of the operator. Cast iron boilers are the smallest of the three
boiler types, with a maximum size of only 10 x 10^ Br.u/hr thermal input (De-
ference 2) . They are generally used for producing low pressure steam or hot
water for commercial or institutional establishments. Generally, cast iron
units cost more than firetube units for comparable size, but require less
ilntenance and can handle overloading in demand surges.
Fuel Usage
Boilers are also described by the type of fuel they are designed to burn.
The vast majority of industrial boilers are designed to burn one or more of
the fossil fuels: gas, oil, or coal. Nuclear powerc-d boilers are currently
used only by the utility industry and military. They are inherently unsuit-
able for hazardous waste disposal. Wood, bagasse, municipal solid waste,
industrial solid waste, and refuse derived fuel are also used as fuels but
comprise less than 1% of the boiler population and less than 0.01% of the
heat input capacity. Because they comprise the bulk of the boilers capable
of burning ha/.ardous waste, this analysis will focus on 'jasfired, oil-fired,
and coal-f.red types.
Oil-fired boilers are often distinguished by the type of oil used, i.e.,
whether they use distillate or residual oil.
Coal-fired boilers are further categorized by firing mechanises which
can be divided into three major groups: stokers firing, suspension firing,
and fluidized bed combustion. A stoker is a conveying system that feeds coal
into a furnace while providing a moving grate upon which the coal is burned.
In suspension faring the fuel is blown into the boiler and burned as a sus-
pension of particles in combustion air. Fluidized bed combustion boilers
burn the coal on a bed of inert particles through which air is blown so that
the bed behaves as a fluid. There are very few of these units in use as
this is an emerging technology.
Stoker finr.; systems can be further divided intrv tr.xee groups: under-
feed stoker, overfeed stoker, and spreader stoker. In an underfeed stoker,
coal is fed to the bottom of a fuel bed, where moisture and volatiles are
driven off and the coal is coked. The volatiles rise through the bed and
undergo combustion above the bed. The coked coal is forced to the top of
the bed by newly fed coal and spills out of the bed onto side gates, where
combustion is completed. Combustion air is supplied at the side grates;
also overfire air is often supplied to the flame zone abovr^ the bed. In
3-2
-------
an overfeed stoker, coal is fed onto a continuous conveyer called a traveling
grate. The grate carries the coal under
-------
TABLE 3.1
BOILER POPULATION DISTRIBUTION BY HEAT -
TRANSFER CONFIGURATION
Boilfer Population
Heat-
Transfer
Configuration
Number of
Boilers
Percent
of Total
Total Boiler Capacity
HW Thermal
Input Percent
( 1 O6 Btu/hr ) of
Water-Tube
37,696
7.5
638,665
(2.2 x 106;
70.0
Fire-Tube
173,936
34.3 219,360
(0.76 x 106)
24.2
Cast Iron
295,298
58.2 £2,570
(0.18 x 106)
5.8
3-4
-------
I
u>
300,000
(1,000,000)
o
n.
c
HI
.C
I I
«c
a
73.3
(O-IO) (10-50) (bO-100) (100-250) (>2iiO)
SIZE MANGE, Mil thermal i'M>"L (1«6 Htu/hr)
n
II
u>
R|ir.ATIVIi DfliTHIHUTION UY CAPACITY Of THIS TIINEB TYPKS Of INDUSTHIAI,
-------
TABLE 3.2
DISTRIBUTION OF U.S. WATER-TUBE INDUSTRIAL BOILERS BY UNIT SIZE AND FUEL TYPE
(MW Thermal Input do6 Btu/hr))
bJ
I
Fuel
Pulverized Coal
Number of Units
Total Capacity, MW
Thermal (106 Btu/hr)
Spreader-Stoker Coal
Number of Units
Total Capacity, MW
Thermal (106 Btu/hr)
Under feed -Stoker Coal
Number of Units
Total Capacity, MW
Thermal (106 Btu/hr)
Over feed -Stoker Coal
Number of Units
Total Capacity, MW
Thermal (106 Btu/hr)
Residual Oil
Number of Units
Total Capacity, MW
Thermal (10*> Btu/hr)
Distillate Oil
Number of Units
Total Capacity, MW
Thermal (106 Btu/hr)
Natural Gas
Number of Units
Total Capacity, MW
Thermal ( 106 Btu/hr)
Total All Fuels
Number of Units
Total Capacity, MW
Thermal (10^ Btu/hr)
0 to 2.9
(0 to 10)
0
0
(0)
46
70
(240)
578
680
(2,300)
59
85
(290)
3,217
3,960
(13,500)
3,151
2,560
(8,700)
4,414
4,475
(15,300)
11 ,465
11,830
(40,330)
2.9 to 14.7
'10 to 50)
0
0
(0)
464
4,650
(15,900)
1,500
14,105
(48,000)
345
3,470
(11,800)
5,637
48,190
(164,000)
1,067
8,280
(28,200)
6,533
57,900
(197,500)
15,546
136,595
(465,400)
Capacity by
14.7 to 29.3
(50 to 100)
0
0
(0)
285
6,175
(21,060)
788
17,265
(58,900)
207
4,455
(lb,200)
1,654
35,640
(122,000)
191
4,295
(14,600)
2,515
53,585
(182,800)
5,640
121,415
(414,560)
Unit Size
29.3 to 73.3
(100 to 250)
453
19,895
(67,800)
474
20,295
(69,000)
169
7,080
(24,200)
85
3,555
(12, -00)
1,OJ9
44,790
(153,000)
170
6,370
(21,700)
1,443
63,320
(216,000)
3,833
165,305
(563,800)
>73.3
(>250)
266
40, 180
(137,000)
88
1 1 ,010
(37,600)
41
5,230
(17,t*00)
2
-------
TABLE 3.3
DISTRIBUTION OF INDUSTRIAL FIRE-TUBE BOILERS
BY SIZE AND FUEL TYPE
(MW Thernal Input (106 Btu'hr) )
Capacity by Unit Size
Fuel
0 to 2.9
(0 to 10)
2.9 to 14.7
(10 to 50)
Total
Coal
Number of Units
Total Capacity, MW
Thermal (1O6 B^u/hr)
Residual Oil
Number of Units
Total Capacity, MW
Thermal (10s Btu/hr)
Distillate Oil
Number of Units
Total Capacity, MW
Thermal (106 Btu/hr)
Natural Gas
Number of Units
Total Capacity, MW
Thermal (1C6 Btu/hr)
8,112
5,650
(19,270)
46,884
35,280
(120,330)
22,643
17,770
(60,610)
79,456
59,120
(201,630)
1 ,224
7,780
(26,530)
4,353
25,860
(88,200)
2,653
15,770
(53,790)
8,611
52,130
(177,.90)
9,336
13,430
(45,800)
51,237
61,140
(208,530)
25,296
33,540
(114,400)
88,067
111,250
(379,420)
Number of Units 157,095 16,841 173,936
Total Capacity, MW 117,820 101,540 219,360
Thermal (106 Btu/hr) (401,840) (346,310) (748,150)
3-7
-------
TABLE 3.4
DISTRIBUTION OF INDUSTRIAL
CAST IRON BOI^ERs 3Y FUEL TYPE
(MW Thermal Iv.pot (1 O6 Btu/hr ) )
Fuel Boiler Capacity'
Coal
Number of Units 35,965
Total Capacity, MW 6,330
Thermal 'JO6 Btu/hr) (21,590)
Residual Oil
Number of Units 59,Pd*
Total Capacity, MW 10.780
Thermal (106 Btu/hr) (36,770)
Distillate Oil
Number of Units 37,612
Total Capacity, MW 6,740
Thermal (106 Btu/hr) (22,990)
Natural Gas
Number of Units 161,827
Total Capacity, MW 28,720
Thermal (106 Btu/hr) (97,950)
Total All Fuels
Number of Units 295,298
Total Capacity, MW 52,570
Thermal (106 Btu/hr) (179,400)
All cast iron boilers have a capacity
less than 4.0 MW thermal input (14 x
106 Btu/hr).
3-8
-------
EXISTING EMISSIONS FROM INDUSTRIAL BOILERS
Criteria Pollutants
The estimated (by Kemp and Dykema (Reference 3) and Devitt, et al. (Re-
ference 1 ) ) emissions of criteria pollutants fron, industrial boilers are suai-
.n-rized in Table 3.5 (Reference 2). There is no information indicating that
control devices or techniques for criteria pollutants other than particles
have been adopted to a significant degree in the industrial ooiler industry.
The estimated net control of particle emissions, presented in the next sub-
section have been applied to the uncontrolled emission rate to yield the
estimated existing emission rate of particles.
None of the test data, accumulated during test burns of industrial
boilers co-firing hazardous wastes (collected in conjunction with develop-
ment of this background information document) indicate that the rates of
emission of carbon monoxide or hydrocarbon are affected by co-firing of
hazardous waste. The rates of particles, sulfur dioxide, and nitrogen ox-
idas emissions will be affected only insofar as the waste stream contains
ash, or sulfur or nitrogen compounds.
Metals Emissions
Trace elements are found in fossil fuels. The largest amounts are
found in coals and residua1- oil. No aata were found on the trace element
concentrations in either gas or distillate oil. Nor were any emissions fac-
tors for trace elements from gas or oil-fired combustion equipment found.
The concentration of trace metals in distillate oils is generally believed
to be quite low; the concentration in gas fuels is thought to be nearly zero.
Table 3.6 (Reference 4) lists some toxic metals and their approximate, un-
controlled rates of emission from oil- and coal-fired boilers. The emission
rate of a trace metal depends upon:
o Its concentration in the fuel
o The vapor pressure of the element and its products of combustion
o The combustion zone temperature
o The temperature history of the combustion gases in cne boiler
It is not necessary that the fla.ne temperature exceed the boiling point of
an element for it to evaporate, the temperature must only oe high enough to
create a significant vapor pressure. There is a correspondence between en-
richment of metals in small particles and their occurrence as mineral sul-
fides in the earth's crust. This implies that ease of reduction to base
metals (which are usually, but not always, more volatile than metal oxides)
or metal hydrides during the initial phases of combustion of a fuel particle
may facilitate evaporation of the metals (Reference 6). The more volatile
elements (and those that form volatile oxides) appear to partition favorably
into the fly ash. The less volatile elements partition evenly between the
fly.ash and slag. Apparently, the more volatile elements become vaporized
in the high temperature flame and condense into very small particles (fumes)
as the flame temperature is quenched by radiant and convective cooling. The
distribution of particle sizes and the elements affected depend on both the
maximum flame temperature and the rate of cooling.
3-9
-------
TABLE 3.5
SUMMARY OF EXISTING EMISSIONS OF
CRITERIA PDLLiJT.-n'.'TS ^OM INDUSTRIAL BOILERS
(.Reference 2)
Particles
Emission Fa ctor (lb/10° Btu)
SO?
NO
CO
HC
Gas-Fired Uncontrolled
Boilers Net Control
Controlled
0.005 to 0.015
0
0.005 to 0.015
0.067 to 0.442C 0.001 0.017 0.003
0 000
0.067 to 0.442 0.001 0.017 0.003
Distillate Uncontrolled 0.015
Oil-Fired Net Control 0
Boilers Controlled 0.015
Residual Uncontrolled (0.068S3
Oil-Fired + 0.020}
Boilers Net Control 0
Controlled (0.068S
+ 0.020)
Pulverized Uncontrolled 0.679Ab
Bituminous Net Control 81%
Coal-Fired Controlled 0.129Ab
Boilers
(Dry
Bottom)
Bituminous Uncontrolled 0.551Ab
Coal-Fired Net Control 75%
Spreader Controlled 0.138Ab
Stoker
Boilers
0.102 to 0.249C 1.040Sa 0.037 0.007
0 000
0.102 to 0.249 1.040S 0.037 0.007
0.156 to 0.842C 1.
0.034 0.007
0 000
0.156 to 0.842 1.060Sa 0.034 0.007
0.637 1.612Sa 0.042 0.013
0 000
0.637 1.612Sa 0.042 0.013
0.637 i.612Sa 0.085 0.042
0 000
0.637 1.612Sa 0.085 0.042
a S is the fuel sulfur content.
b A is the fuel ash content.
c NO emissions are strongly dependent on firing type.
3-10
-------
E 3.6
EMISSION FACTORS (UNCONTROLLED) OF
SELECTED TOXIC TRACE ELEMENTS
FROM INDUSTRIAL BOILERS
(Reference 4)
Element
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
T ^ AO
XjCSOU
Mercury
Selenium
Thorium
Residual
Oil-Fired
Boilers
0.000023
0.000042
0.000067
0.000005
0.000121
0.000070
0.000119
0.000002
0.000037
<0. 000002
Emission Factor (Ib
Spreader Stoker
Bituminous Coal-
FLred Boilers
0.00047
0.00279
0.00744
0.00018
0.00014
0.00465
0.00072
Onm AT
• WVJ 1 O 1
0.00002
0.00054
0.00012
/10b Btu)
Pulverized
Bituminous Coal-
Fired Boilers
0.0003"
0.00214
0.00977
0.00023
0.00019
0.00605
0.00093
n nnooQ
U . *J\J &\J3
O.OC002
0.00070
0.00014
3-11
-------
Health risk considerations imply that the most critical emissions from
boilers co-firing hazardous wastes will be toxic metals. Cowherd, et al.,
in the course of an investigation of the hazardous emissions from coal-fired
utility boilers, ranked various trace element constituents of coal according
to their reported toxicities (Reference 5). They then sampled a boiler ex-
naust gas with cascade impacters and analyzed the individual stages for the
various elements. Table 3.7 lists the mass fraction of the total amount of
the eight most toxic metals that was found to be in the particles of less
than 0.87 micrometers iiameter. They did not report enrichment for Hg, Se,
and As. Lyon, following similar tests, classified elements according to
their propensity to be incorporated into the boiler slag (Class 1); be con-
centrated in the fly ash and exhausted from the electrostatic precipitator
(Class 2); or remain completely in the gas phase iClass 3) (Reference bj,
see Table 3.7.
Where che measurements of these two investigations overlap the conclu-
sions agree except for thallium. There appears to be an enrichment of the
most toxic metal elements in the smallest particles emitted during coal com-
bustion. These small particles are difficult to remove; mechanical collec-
tors and scrubbers are relatively ineffective. Only 15% of industrial coal-
fired boilers (3% of all industrial boilers) are presently equipped with
either electrostatic precipitators or fabric filters, which are needed to
control small particle einissions.
CURRENTLY INSTALLED EMISSION CONTROLS
Data gathered from the National Emissions Data System (NEDS) reveal
that uncontrolled oil- and gas-fired industrial boilers generally comply
with emissions regulations encoded in State Implementation Plans (Reference
1,7). Therefore, there are few air pollution control devices installed on
boilers that burn either of these fuels. Approximately two-thirds of coal-
fired industrial boilers have some particle control device installed. Sul-
fur dioxide emissions are generally not controlled. A survey of 2,533 boil-
ers (Reference 3), summarized in Table 3.8, revealed that cyclones are the
most commonly installed control devices.
Table 3.9 summarizes the percent collection efficiency and percent ap-
plication of control devices by boiler firing type. By combining these data
with those in Table 3.2 it can be estimated that approximately 70% of the
installed capacity of coal-fired industrial boilers is equipped with some
type of particle control device. However, coal-fired equipment comprises
only about one-fourth r>* the total installed capacity, so only about 17%
of the installed industrial boiler capacity has any type of control device.
Control Techniques for Particulate Matter
The four types of particle emission control devices that are presently
in use to any significant extent are electrostatic precipitators, fabric
filters, wet scrubbers, and mechanical collectors (cyclones). The present-
ly installed equipment is in place exclusively f^r the purpose of attaining
compliance with standards for emission of particles. These four devices are
discussed in this section. Other, advanced design equipment is in use or
undergoing evaluation. The more promising of these devices are discussed
in the next section.
3-12
-------
Element
TABLE 3.7
ENRICHMENT 0? TOXIC ELEMENTS IN .SMALL PARTICLES
% of Element
in Particles Less
Than 0.87 Microns
Diameter (Ref. 5}
% of Element
in Particles Greater
Than 0.87 Microns
Diameter (Ref. 5)
Element Class3
(Lyon, Ref. 6)
Be
Cd
Pb
Hg
£e
Te
Th
As
s^r*
82
59
32
—
—
72
34
—
— —
18
41
63
—
—
28
16
—
4 —
..
2
2
3
2 or
—
1
2
1 or
3
2
Class 1 elements partition evenly between boiler slag and fly ash.
Class 2 elements concentrate in small particles.
Class 3 elements remain in vapor phase through ESP.
3-13
-------
TABLE 3.8
INSTALLED AIR POLLUTION CONTROL DEVICES
COAL-FIRED BOILER
(Reference 8)
Percent
Control Device by Nuuiber
None 33
Cyclone 47
Scrubber 4
Electrostatic
Precipitator 14
Fabric Filter i
TABLE 3.9
ESTIMATED APPLICATION OF CONTROL EQUIPMENT
TO COAL-FIRED INDUSTRIAL BOILERS, 1978
System Type
Pulverized
Cyclone
Stoker
Average
Collector
Efficiency
(%)
as
82
85
Control
Application
(%)
95
91
62
Net
Control
(%)
81
75
53
Source: Reference 1 and 2.
3-14
-------
Electrostatic Precipitators
Process Description. Electrostatic precipitators (ESPs) remove parti-
cles from a gas stream by impressing an electrostatic charge upon them and
tnen causing them to migrate to oppositely charyed collection plates in a
high potential d.c. field. In addition, gravitational settling can account
for significant fractional removal of large diameter (>40 ••) -particles. In-
coming, particle laden gas flows by a discharge electrode, whicn, because
its electromotive potential is high relative to ground, is surrounded by an
electrostatic corona. As particles flow through the corona they accumulate
charged ions and thereby become charged. Thus charged, the particles mi-
grate toward the collection electrodes (or plates 1 whose potential is at
ground, where they are collected. Removal of the particles from the col-
lection electrode is effected by rapping (vibrating) the electrode.
Rapping the collection electrodes may be done by pneumatic or electric
vibrators or by mechanical dropping hammers. Reentrainment of particles,
released from the collection electrode during rapping, is a significant
cause of inefficiency, whatever the rapper design. Precipitotors normally
consist of 2 to 4 parallel sections, each consisting of 2 to 4 sequential
fields. Thus, if rapping cycles are well conceived, reentrainment from
only the final field is problematical.
Air flow inside the precipi tat ~>r must be evenly distributed to avoid
overloading some portions while underloading others. Velocities are kept
low (4 to 8 ft/sec) to avoid scouring collected particles from the collec-
tion electrode.
Applicability to Industrial Boilers. The first successful application
of an electrostatic precipitator was in 1907, when Cottrell installed a unit
to collect acid mist from a sulfuric acid plant. Since then, many thousands
of units have been installed. ESPs are available iit sizes to handle from
5000 acfm (0.5 x 1 O6 Btu/hr boiler) to the largest electric utility boilers.
Electrostatic precipitation is a well established technology that has been
studied and researched for nearly 30 years. Adequate solutions have been
developed for most technical problems so that ESPs are applicable to nearly
any boiler fly ash control situation. The largest power utility boiler can
be equipped within 2 to 4 years, smaller boilers even more quickly.
Factors Affecting Performance. The boiler operating parameter that has
the greatest effect on the performance of an ESP is the resistivity of tne
ash particles. If the resistivity is too high (greater than 1011 ohm-cm)
the charges or. individual particles will flow through the dust layer at a
rate slower than the rate of particle (and charge) accumulation. Ultimately
an electric breakdown will occur in the dust layer that wilJ initiate exces-
sive sparking between the precipitation electrodes (Reference 9). Sparking
wastes charges and thereby attenuates performance. At higher resistivity
(>1012 ohm-cm) back corona effects will introduce positive ions into the
electrode interstices and reduce the charges on the particles. If the ash
resistivity is too low (OO7 ohm-cm) particles will loose their charges im-
mediately upon contact with the collection electrode. Then, lacking elec-
trostatic adhesive, they will be easily reentrained.
3-15
-------
Ash resistivity is rr.ore problematical for ESPs in service on coal-firea
industrial boilers than those in service on electric utility boilers because,
unlike utilities, industrial users ao not normally purchase long-term coal
commitments. As a result induatria- boilers burn a more variable fuel, hav-
ing a more variable ash resistivity. Ash with acceptable resistivity results
from combustion of coal with high alkali (Na20 and L^O), F^C^ Content ana
low concentrations of Ca, Mg, Si, and P oxides. Resistivity drops as sulfur
content increases. Injections of SOj (and some proprietary compounds) ap--
pears to be an effa;tive means of reducing excessive ash resistivity and
thereby improving ESP performance.
The size of the ash particles is also, an important variable. ESPs ap-
pear to be least efficient for collection of particles in the size range of
0.1 to 1.0 microns diameter (Reference 10). There are indications that pen-
etration through hot side ESPs (those located upstream of the air preheater)
are primarily the result of rapping losses of particles greater than 2 mi-
crons diameter.
Most oil-fired boilers now employing ESP control were converted from
their original coal jruel (Reference 11). The ESP on such a unit, if not
modified, may be only 50% efficient. Oil ash tends to have a high carbon
content. The carbonaceous nature of the ash causes two types of problems.
First, the ash can be sticky and tar-like in nature and therefore difficult
to remove from the collection plates, insulators, frame members, and other
surfaces. Its accumulation on the collection plates eventually effects
sparking which creates a fire hazard. This difficulty can be overcome some-
what by moving the ESP to the upstream side of the air preheater, or by prac-
ticing fly ash reinjection. Both techniques yield a dryer, less sticky ash
but neither alleviates the second problem - the resistivity of the ash is
low because of the high concentration of relatively conductive carbon. Low
resistivity ash is subject to reentrainment because it is not electrostati-
cally bound to the collection electrode.
There is no reason why an ES? could not be installed to control the par-
tic.'.e emissions from a gas-fired boiler. However, particle emissions from
gas-firing are so low that control is unnecessary.
Impact of Burning Hazardous Waste. Corrosion is the major anticipated
impact of combustion of hazardous waste in boilers equipped with ESPs. Com-
bustion of S, P, or Cl containing compounds will result in emission of the
corresponding inorganic acids. ESPs ..an be designed to remove particles
from highly corrosive gases (the first installation was at a sulfuric acid
plant) by incorporating special materials of construction. It is unlikely
that this will have been done for an ESP that was originally designed to
control particle emissions from a coal-fired boiler.
A second potential .mpact could be alteration of the ash resistivity by
these acids. This could be beneficial or detrimental to the ESP performance,
depending upon the resistivity of the coal ash.
ESPs can be expected to efficiently (>99%) remove particles from the
boiler flue gas. Their minimum efficiency is for particles of approximately
0.5 microns diameter. However, this minimum efficiency is in excess cf. 90%.
As was mentioned previously, toxic metals appear to be concentrated in siiall
3-16
-------
particles; those less than 0.9 microns in diameter. However, even though 10%
of the particles most heavily laden with the toxic metals may penetrate the
ESP, tne net control of tne toxic metals will be over 90%.
A dry ESP will not remove any vapor phase suostances. These include tr.e
hazardous organic compounds, hydrochloric acid and vaporous metals (sucn as
mercury). It would be necessary to cool the exhaust gas to condense or ab-
sorb these substances.
Fabric Filtration
Process Description. The microscopic mechanisms by which fabric filters
(baghousei) remove particles from gas streams are less understood than those
operating in other air pollution control devices (Reference 12). The typical
pore size of fabric is on the order of 10 to 100 microns, which is large re-
lative to the diameters of the particles to be removed. When, at the begin-
ning of a gas cleaning cycle, the fabric is clean, the removal mechanisms
for large particles appear to be direct interception and impaction. Large
?ai -.ides (>2 microns diameter) are those having sufficient inertia to be
separated from gas slip streams as the gas flows around individual fibers.
Small particles (those less than 0.1 microns uiameter) are not subject to
inertial removal mechanisms but are sufficiently small to be subject to the
molecular mechanisms of diffusion and Brownian motion. Penetration of par-
ticles through the fabric is relatively high during the initial moments of
a gas cleaning cycle. Electrostatic forces may have an effect on particle
removal, but neither the magnitude nor the mechanism of the effect is well
understood.
After the first few minutes, accumulated particles bridge the pores and
form a filter cake on the fabric. The filter cake hac, a smaller average pore
size than the fabric and, apparently, sieving becomes the predominant removal
mechanism, borne installations create an artificial filter cake by injecting
lime or recycled ash into the gas stream entering the clean bag chamber. The
artificial cake hastens recovery to full efficiency of the cleaned bags.
Bag cleaning is generally done by one of three met! ods: mechanical shak-
ing, reverse air flow, or pulse jet flow. Mechanical shaking is accomplished
by rapidly moving the frame, to which the tops of the bags are attached, back
and forth in a rapij horizontal motion. The resultant flexing of the bag wall
causes the filter cake to crack and fall off in chunks. Reverse air cleaning
is accomplished by closing off a section of the baghouse from the boiler ex-
haust gas flow and forcing clean air (either filtered stack gas or ambient
air) to flow backwards through the bags. The reversal of the gas flow causes
the bag walls to flex (either expand or contract, depending on whether the
normal flow tc the inside or to the outside of the bag). The flexing cricks
the cake which is then forced off of the bag by the reversed air flow. Pulse
jet cleaning is accomplished by introducing a sudden blast of compressed air
into the top of the bag. The pulse sends a traveling mechanical wave down
the bag, cracking and flaking oft portions of the filter cake. Some back flow
through the bags is created aiding cake removal. Pulse jet cleaning can be
accomplished without cordoning off the bags being cleaned.
Factors Affecting Performance. Once the filter cake is established on
tne bag surfaces, fabric filters remove 99% or more of all size of particles.
3-17
-------
Penetration of particles \s almost exclusively through rips in the fabric,
leaks through poorly sealed connections, or through abnormally large (>200
T.icrons) pores in the fabric. Particle size distributions downstream of
fabric filters ars nearly identical to those upstream. Maintenance is the
most important aspect of operation of a baghouse. Other factors can br im-
portant. Sticky particl3s are poorly removed and can result in excessive
pressure drops that can, ultimately, o-.use bag failures. Condensation of
water can cause bag blinding, condensation of inorganic acids can cause
corrosion and weaken some fabric materials. The temperature of the filter
fabric must be kept above the acid dew point to avoid acid damage to the
filters. The mini.-num acceptable temperature is dictated by the amount of
sulfur trioxide in the flue gas which, in turn, is a function of fuel sul-
fur content. The acid dew point of fossil fuel combustion exhaust gases
generally falls between 325° and 400°F.
Applicability to Industrial Boilers. The first fabric filters were
installed on industrial processes approximately 100 years ago. The first
installations on coal-fired boilers, at the impetus of tightening pollution
abatement regulations, were made in the late 1960's. Performance has been
encouraging and new installations are increasing. Existing applications
range in size from small industrial boilers '.5,000 actual ft3/min) to large
electric utility boilers (3 x 106 actual ft3/min;.
The ash from oil-fired boilers tends to be hygroscopic and sticky and
tends to be irreversibly trapped in the interstices of the bag filter fabric
(Reference 11). The few installations of baghouses on oil-fired boilers
that have been attempted have met with limited success.
There is no known installation of a fabric filter on a gas-fired boiler.
This is because gas-fired boilers, properly operated, emit minute quantities
of particles.
Impact of Burning Hazardous Waste. The primary, expected effect of the
combustion of hazardous waste in a boiler already equipped with a fabric fil-
ter is the potential for increased corrosion.
If the waste burned creates an oily or damp ash, blinding of the fabric
filter could become a problem. The efficiency of fabric filters is relative-
ly insensitive to particle size. They are as efficient for collection of
fine particles as for larger particles. (Excessive amounts of fine parti-
cles will ultimately clog the fabric since they are not effectively removed
during cleaning.) Thus, the efficiency of removal of metals by fabric fil-
ters is expected to be excellent.
However, unless the exhaust gas is cooled so as to condense vapor phase
metals (e.g., Hg), these will not be collected. Vapor phase organic compounds
will not be collected. Cooling the gases sufficiently to cause their conden-
sation probably would result in blinding of the filter.
wet Scrubbers
Process Description. A wet scrubber is a device for providing contact
between a liquid and a gas in order to effect the capture of some constituent
3-18
-------
of the gas. The captured constituent may be solid, liquid, or gaseous. There
are hundreds of designs in use most of which fall into one of four categories.
o Spray towers - high pressure liquid is sprayed into the gas stream
in a relatively open chamber.
o Packed bed scrubbers - liquid flows downward through a porous bed
of inert >naterial, countercurrent to the gas flow.
o Flooded plate towers - similar to packed bed scrubbers except that
the gas bubbles through holes in a discrete number of plates on
which there is a layer of liquid.
o Venturi scrubbers - relatively low pressure water is admitted into
the highly turbulent gas flow at the throat of a venturi.
Particle removal is effected by each of several processes. Larger par-
ticles are removed by impaction upon liquid surfaces. These may be film sur-
faces (as in a flooded plate scrubber) or droplet surfaces (as in a venturi
scrubber). Smaller particles are deposited on the same liquid surfaces by
diffusion or Brownian motion. If the gas being scrubbed is warm and moist
then condensation occurs as it is cooled in the scrubber. The condensation,
to some extent, occurs on the surface of the particles, thereby increasing
their size (Reference 12). Increasing the size of the particles, by includ-
ing them in the liquid phase, is the necessary operation. As dry particles,
few have sufficient size to be easily separated from the gas stream inertial-
ly. Their inclusion in a water droplet yields a liquid/solid mixture that
does have sufficient mass to be separated from the gas stream by relatively
simple filters (de-mister pads) or cyclones.
Several studies have shown that the efficiency of particle removal is
proportional to the total power expended by the scrubber. The power expen-
diture includes the pressure drop in the gas and liquid phases. It does
appear, however, that the efficiency is also inversely proportional to the
mass concentration of particles in the inlet gas stream. This implies that
removal of the last few particles is at a great energy expense.
Scrubbers, no matter their design, do not remove small particles effec-
tively at any reasonable power consumption. At practical energy expenditures,
scrubbers are ineffective for removal of particles less than 0.5 to 1 micron
in diameter (Reference 11).
Applicability to Industrial Boilers. Scrubbers, certainly, are appli-
cable to coal-fired industrial boilers. They are small, relative to fabric
filters and electrostatic precipitators, so, space allocation problems are
less. However, scrubbers consume more energy than other control devices.
The energy loss occurs both through the mechanisms of pressure drop in the
gas stream and loss of stack draft that results from cooling of the gases.
Especially high pressure drops are required to remove sub-micron particles.
Bciler de-ratings, on the order of 5-10% have been required on large utility
boilers to provide the power consumed by scrubbers. The usefulness of scrub-
bers for control of particles from oil- and gas-fired boilers is limited be-
cause of the small size, low concentration, and low density of the particles
emitted. There are few existing installations.
3-19
-------
Factors Affecting Performance. The sine qua non of scrubber performance
is intimate contact between the gas and liquid phases. This contact may be
effected by creating turbulence in the gas stream (i.e., venturi) with con-
sequent loss of kinetic energy in the gas, by high pressure sprays to dis-
perse the liquid in fine droplets, or by drawing the gas through a bed with
a small average pore size with consequent high pressure drop.
Particle size is also ;.mportant. Collection of particles of less than
1 micron diameter is at a uracceptably high energy cost (Reference 11). Par-
ticle density and tendency to absorb water, although these do not vary widely
among coal-fired boilers, aru important. Dense, hygroscopic particles are
more easily scrubbed.
Impact of Burning Hazardous Waste. Of r.he four most commonly employed
air pollution control devices on industrial boilers, only scrubbers have the
capacity to remove the vapors of inorganic acids and volatile toxic metals
that may be emitted. Corrosion, that might become a problem with the other
types of control devices, can be overcome in a scrubber system by addition
of alkali to the scrubber water. Should this become necessary it will exa-
cerbate an already existing, spent scrubber liquor treatment and disposal,
problem. Also, even though installed scrubbers will remove, at least some
of, the vaporous metals, they operate at too low a pressure drop to remove
the metal fume that consists of particles less than 0.5 micron in diameter.
Centrifugal Separation (Multiple Cyclones)
Process Description. Multi-tube cyclones (cyclones) consist of banks
of tubes in parallel through which the exhaust gas flows. The tubes vary
from 2 to 12 inches in diameter. Gas enters through an annular opening near
the top of the tube, passes over spin blades (curved blades between the out-
er tube wall and the inner tuoe wall) which impart a tangential t.-ment to
the gas velocity, flows into the chamber, and then back up through the cen-
tral exhaust tube. The vortex created, causes particles, whose density is
greater than that of the gas, to migrate outward to the tube wall. Colli-
sion with the wall absorbs their momentum and they fall into the hopper
below. The literature on cyclones is extensive, techniques for prediction
of collection efficiency based on; particle size and density; tube size;
gas flow rate; and gas density are readily available (Reference '• 1 ).
Applicability to Industrial Boilers. As was noted earlier, nearly half
of all coal-fired boilers are equipped with cyclone separators. They are
relatively small, consume little power (pressure drops range from 3 to 6
inches w.c.) and have no moving parts that require maintenance. However,
their collection efficiency is poor for particles of less than 10 microns
in diameter. Thus, their effectiveness for control of oil- and pulverized
coal-fired boilers is marginal. Even for stoker coal-fired boilers their
prime utility is as a pre-cleaner for more efficient collection equipment.
Factors Affecting Performance. Particle size and density are the most
critical factors affecting the performance of cyclones. Gas density and
flow rate are also important: the latter more so, because cyclone perfor-
mance falls off rapidly when the gas flow falls below its design value.
3-20
-------
Impact of Burning Hazardous Maste. Corrosion of metal parts in general
and cyclones in particular may bt increased by combustion of hazardous corn-
pour, is that contain Cl, P, and S.
The efficiency of cyclones for removal of small diameter particles and
vaporoas substances is nil. Thus, a large portion of the most toxic metals
and all. of the vaporous substances will penetrate the cyclones and be emitted
into the environment.
Control Techniques for Nitrogen Oxides
Nitrogen oxides (NOX) emitted from the combustion of fuel are formed by
oxidation of organic nitrogen compounds in the fuel and by oxidation of at-
mospheric nitrogen (N2) (Reference 2). Approximately 95% of the NOX formed
is nitric oxide (NO), the remainder is nitrogen dioxide (NO2). The ratio
of products varies somewhat. Coal and oil-fired furnaces apparently emit
a smaller fraction of NO2» The rate of formation of NOX by both mechanisms
(fuel and thermal) is a function of combustion conditions, primarily flame
temperature and excess oxygen availability. Both fuel bound and atmospher-
ic nitrogen are more resistant to oxidation than the organic components of
fuels, so their oxidation is effected only at relatively harsh conditions.
Therefore, it is possible to mitigate the combustion conditions in a boiler
firebox somewhat to lessen formation of nitrogen oxides without reducing the
efficiency of fuel combustion. In fact, reducing the excess air (one NOX
reduction technique) may increase the efficiency for the boiler.
Five techniques for reducing NOX emissions are currently practiced (Re-
ference 2):
o Low excess air (LEA)
o Staged combustion (SO
o Flue gas recirculation (FGR)
o Low NOX burners (LNB)
o Reduced air preheat (RAP)
Each of these techniques is discussed below.
Low Excess Air
Process Description. Most industrial boilers are routinely operated
at higher excess air rates than are recommended by the manufacturers of the
boilers. Operation at higher than design excess air provides a cushion
against smoke emissions during sudden load changes, lessens the need for
careful operator attendance, and obviates the requirement for sophisticated
(and expensive) combustion air controls. High excess air also reduces the
thermal efficiency of the boiler by increasing the amount of hot gas ex-
hausted, and increases the amount: of residual oxygen available for oxida-
tion of fuel and atmospheric nitrogen.
Reduction of excess air in small, relatively simple fire tube and pack-
age water tube boilers is accomplished by closing down the inlet vanes on
the forced draft fan or by closing down the vanes on the windbox. On larger,
more sophisticated boilers, the forced and induced draft fan speeds can be
reduced.
3-21
-------
Development Status. Reduction in the excess air is being applied widely
to improve thermal efficiency and thereby reduce fuel costs.
Applicability to Industrial Boilers. Lowering excess air is applicable
to all industrial boilers that have been operated at higher than design air
levels. It i? recommended that automated air controls (oxygen trim systems)
be installed to ensure safe, smokeless operation.
Factors Affecting Performance. At any boiler load, the percent reduc-
tion of NOX emissions is directly proportional to the reduction in excess
air. In coal-fired boilers, a decrease of 1 % in the flue gas oxygen concen-
tration c?n result in a 5% reduction in NOX emissions.
Impact of Burning Hazardous Waste. It has been hypothesized that reduc-
tion in the available oxygen in the combustion zone might reduce the oxida-
tion of hazardous compounds for fie same reasons that it reduces oxidation of
nitrogen. This effect was not observed at one of the sites tested for this
effort where excess air was vari«d at constant boiler load. The destruction
efficiency of hazardous materials with the oil-fired boiler at 50% excess air
(7% 02) was not discernably different from the destruction efficiency with
the boiler at 17% excess air (3% 02). The determination of destruction effi-
ciency was made by comparison to the flow rates of the hazardous compounds
in the stack gas to their flow rates in the fuel stream. It is possible that
their destruction was by some means other than oxidation.
Staged Combustion
Process Description. Staged combustion consists of reducing the air flow
through the burner port (burner box) to a level flow needsd to completely ox-
idize the fuel. Secondary air is added to the flame zone to provide the addi-
tional oxygen required. The practice has two effects: it reduces the temper-
ature and oxygen concentration in the primary flame zone ?nd it diffuses the
flame into a larger volume. The reduced temperature and oxygen levels reduce
formation of NOX. The more diffuse flame provides a larger volume from which
heat is radiated, allowing more rapid cooling of tne flame and thus, A shorter
time for oxidation of nitrogen. Staged combustion is fiearly always applied
in conjunction with low excess air. The secor.dary (staged) air can be intro-
duced through the side wall or the top of the flame zone. The effect is in-
herent in coal-firdd stoker boilers but may be enhanced by reducing underfire
air flows.
Development Status. The status of development of technology to imple-
ment staged combustion in various types of industrial boilers is summarized
in Table 3.10.
Applicability to Industrial Boilers. Staged combustion normally can be
implemented on oil and gas-fired boilers. Installation of an oxygen trim sys-
tem to monitor and control the flow rates of primary and secondary combustion
air is necsssary to prevent excessive smoke and combustible gas emissions.
Implementation of staged combustion on pulverized coal-fired boilers
can result in increased slagging and corrosion and pose the same, poor com-
bustion efficiency problems. These again can be alleviated by installation
of automated combustion air controls, and by installation of compartmented
windboxes to assure equal distribution of combustion air among burners.
3-22
-------
TABLE 3.10
DEVELOPMENT STATUS OF STAGED COMBUSTION
FOR -APPLICATION TO INDUSTRIAL BOILERS - 1982
(Reference 2)
Coal-Fired
Oil- and Natural Gas-Fired
Industrial Boilers
Boiler
Type
Status
Packaged and
field erected
s tok e rs
Available
but not
implemented
Field erected
pulverized
Available
and
implemented
Industrial Boilers
Field erected
watertube
Available
and
implemented
Packaged
watertube
Available
and
implemented
Packaged
firetube
N/A
R&D status
N/A = not available
3-23
-------
Stoker coal-fired boilers appear to present the most difficult case for
retrofit of staged combustion. In general, the overfire air ports are not
adequately designed and positioned to provide efficient secondary combustion.
In addition if the primary lunderfire) air is reduced too much, clinker
£orr.-.s on the grates requiring additional labor tc manually break the lumps.
Emissions of NOX are reduced by implementation of staged combustion in most
stokers, but the emissions reductions are not consistent with increased stag-
ing.
Factors Affecting Performance. Utility boiler experience has shown that
if the secondary air ports are located too close to the connection bank high
steam temperatures result. If the secondary air is introduced too close to
the burners, the staging is compromised and optimum NOX reductions are not
realized.
With distillate oils or gas fuels, efficient smoke-free operation can
be attained with the primary air as little as 90% of theoretical combustion
air. Primary air must be 105% of theoretical combustion air to prevent slag-
ging and corrosion in residual oil- and coal-fired boilers.
Impact of Burning Hazardous Wastes. As with all approaches to reducing
NOX emissions, the intent of staged combustion is to effect lessened oxida-
tion of nitrogen by reducing the fla.-ne temperature and the availability of
oxygen in the flame. Mitigation of the flame conditions may be inimical to
destruction of hazardous compounds. To date, one boiler practicing staged
combustion has been tested. The unit was an 86 x 1O6 Btu/hr, gas-fired
•rater tube boiler that was co-firing aniline wastes. The waste stream was
spiked with chlorinated hydrocarbons for the test. The ORE of the wastes
exceedea 39.99% for all compounds (except benzene, a common PIC of fuel
burning for which the ORE was 99.98%) for all runs. There was no discernible
difference between the ORE observed during staged combustion and unstaged
comoustion runs.
The second potential impact is accelerated corrosion from hydrochloric,
nitric, and phosphoric acids formed when hazardous wastes containing Cl, N,
or P are burned. This effect will be most pronounced in residual oil- and
coal-fired boilers that are subject to enhanced corrosion during phased com-
bustion.
Flue Gas Recirculation
r*r
-------
The higher operating temperature impressed upon the forced draft fan
reportec'.ly causes it to erode more rapidly than normal (Reference 2). Flame
instability is lessened if recirculation exceeds 20 to 25% of the combustion
air.
Factors Affecting Performance. Recirculati in rate is the only variable
in this system. The percent reduction of NOX end ssions from distillate oil-
and gas-fired boilers is approximately linear with recirculation rate up to
30% flue gas recirculation. The percent reduction of NOX emissions is ap-
proximately 10 for flue gas recirculation rates and between 15 and 25% in
residual oil-fired boilers.
Impact of Burning Hazardous Wastes. The inorganic acids formed when
compounds containing Cl, N, or P are burned will contribute to corrosion of
the recirculation duct work and the forced draft fan. Addition of hazardous
wastes, especially those with high water contents, may contribute to the
flame instability that has been observed at high recirculation rates.
Low NO,, Burners
Process Description. Low NOX burners are designed to accomplish the
functions of low excess air flue gas recirculation end staged combustion
within the burner itself. They incorporate techniques such as delayed fuel/
air mixing, internal gas recirculation, flame cooling and dispersion sur-
faces, off-stoichioraetric fuel/air mixing to reduc« flame temperatures, and
oxygen availability in the high temperature zone. They differ from the pre-
viously discussed techniques only in that these functions are accomplished
internally (in the burner) rather than in the boiler.
Applicability to Industrial Boilers. Low NOX burners are available for
relatively small industrial boilers. The size of available low NOX burners
has inhibited their use in larger boilers because of the cost of construction
of multiple burner boilers. Single burner, shop-erected boilers up to 250
x 1O6 Btu/hr are less costly than field-erected multiple burner boilers of
eqxial size.
Factors Affecting Performance. Low NOX burners are affected by the same
operating parameters as the three techniques previously discussed. Greater
dispersion of the flame may cause it to impinge upon the back wall of the
furnace. Higher temperatures appear to cause increased NOX emissions.
Impacts of Burning Hazardous Wastes: Burning hazardous materials will
impact low NOX burners the same way it will impact other NOX control tech-
niques, i.e., possible: increased corrosion, low destruction efficiency and
flame instability. No testing of units burning hazardous wastes in low NOX
burners has been reported.
Reduced Rir Preheat
Process Description. Most boilers larger than 50 x 106 Btu/hr recover
sc.fe heat from the flue gases, either by combustion air preheaters or by
feedwater heaters (economizers). Selection of an economizer rather than an
air preheater lowers the heat input to the flame and thereby reduces oxida-
tion of atmospneric nitrogen. It does not appear to reduce NOX formation
3-25
-------
from fuel-bound nitrogen and, therefore., is less effective for residual oil-
fired and coal-fired boilers.
Development Status. Both air preheaters and economizers are well-devel-
oped technologies. In most cases, the purchaser of a boiler can opt for one
as easily as the other.
Applicability to Industrial Boilers. The technique is applicable to
new boilers. Retrofit of existing boii^rs is not likely to be economically
feasible.
Factors Affecting Performance. The degree of reduction of NOX emissions
is a function of the degree of reduction of air preheat. Normally, the air
preheater either exists or if does not, and the combustion air temperature is
either ambient or about 450°F. Gas and distillate oil-fired boilers with no
air pre-heat emit approximately half as much NOX as boilers with air preheat.
As was mentioned, thfre is little or no affact on formation of NOX from fuel
bound nitrogen. Therefore, the percent' reduction of NOX from residual oil
and coal-fired boilers is less, although the absolute reduction is= probably
equivalent.
of Burning Hazardous Waste. Other than possible increased cor-
rosion of the economizer and boiler, combustion of hazardous wastes should
have no effect on the performance of a boiler employing reduced air preheat.
Insufficient test data have been collected to demonstrate the effect of re-
duced flame temperature on the efficiency of destruction of hazardous com-
pounds.
POTENTIALLY APPLICABLE CONTROL EQUIPMENT
Available data indicate chat combustion of hazardous waste streams in
industrial boilers contributes no fly ash to the boiler exhaust, other than
the amount that is expected during combustion of virgin fuels of comparable
ash content. Compliance with EPA and state particle emission regulations
should be attainable by the same means employed for boilers burning virgin
fuels. There may be some wastes (e.g., paint manufacturing waste, degreas-
er waste) that contain abnormally high concentrations of ash and inorganic
natter that, when burned, emit excessive amounts of fly ash. Unless these
wastes also contain high concentrations of organic chlorine or toxic metals,
the application of one of the control devices discussed previously should
be sufficient.
Existing baghouses and electrostatic precipitators (ESPs), where they
are installed, are adequate to control the emissions of the relatively non-
volatile, toxic metals. ESPs are less efficient for removal of particles
of approximately 0.5 micron diameter than for other sizes; but, are approxi-
mately 90% efficient for these particles. Thus, even if as much as 80% of a
toxic metal is concentrated in particles of less than 1 micron diameter, the
overall penetration of the toxic metal will be less than 10%. The efficiency
of baghouses is relatively unaffected by particle size and should be in ex-
cess of 99%.
3-26
-------
The presence of hydrochloric acid (HC1) alone in the exhaust gas also
presents no difficulty to conventional control devices. Relatively simple
water flooded packed towers can achieve 98% efficiency if the scrubber watf:r
blowdown rate is sufficient to keep the HC1 concentration below 21 (Reference
12). Single pass scrubbers using caustic liquor can easily attain 99% effi-
ciency.
In the case that the hazardous waste contains high concentrations of
chlorine and toxic metals, presently installed equipment will prove inade-
quate. This case will require removal of vapor phas^ HC1 (and perhaps vapor
phase metals) and highly efficient removal of the small particles that will
contain most of the volatile toxic compounds. Of course, it would be possi-
ble to achieve adequate removal of both types of pollutants (vapors and fine
particles) by series installation of a scrubber and an ESP or baghouse; but
this aiay be more expensive than one of several recently demonstrated devices
that have demonstrated efficient removal of both particles and gases.
Emissions of nitrogen oxides (NOX) during combustion of wastes having a
high nitrogen concentration is the second case that may require application
of advanced control technology. Of the five combustion practices that are
currently implemented for control of NOX emissions, three (low excess air,
flue gas recirculation, and reduced preheat) are effective for reduction of
thermal oxidation of atmospheric oxyg«n but less effective for reducing oxi-
dation of fuel-bound nitrogen. Of the two techniques (staged combustion and
low NOX burners) that have demonstrated capability for reducing oxidation
of fuel-bound nitrogen only one, staged combustion, has been evaluated in a
hazardous waste service. These tests were performed on a gas-fired boiler
in which aniline production wastes (nitrobenzene and aniline) were te^ng co-
fired. The nitrogen contributed by the waste stream amounted to approximate-
ly 1.5% (wt/wt) of the total (gas plus waste^ fuel - a concentration that is
typical of coals. The test results showed NOX emission rates to be about 1.1
Ib NO /10 BCJ when combustion staging was not implemented and about 0.35 Ib
NOX/10 Btu wnen the combustion was staged.
Thus, staged combustion reduced NOX emissions sufficiently to effect
compliance with emissions limits placed upon a fuel with A comparable nitro-
gen content, but not sufficiently to comply with emissions limits for either
gaseous or liquid fuels. Note that the regulations cited apply only to large
(>250 x 10^ Btu/hr) boilers but that regulatory alternatives for smaller boil-
ers are presently being considered. Currently proposed regulation of indus-
trial boilers larger than 100 x 1O6 atu/hr heat input limit NOX emissions to:
NOX Emission Limit
Fuel (lb/106 Btu)
Gas and Distillate Oil 0.1
Residual Oil 0,3 - 0.4a
Coal (Stoker) 0.6
Coal (Pulverized) 0.7
Variable - depends upon fuel nitrogen content.
3-27
-------
Obviously, this boiler would not meet the proposed regulation for gas-fired
boilers. However, the proposed NOX regulation contains a provision that
allows holders of RCRA permits to obtain a variance based upon their demon-
stration that the best available technology will net bring them into compli-
ance with the NOX emission limit while maintaining adequate DRE of the haz-
ardous material. Such conflicts will be resolved on a case by case basis.
Advanced Control for Vapor ?lus Small Particle Emissions
Many novel, air pollution control devices have been developed and tested
over the past several years under the EPA Research on Novel Device for Par-
ticulate Control program. These were considered in relation to control of
emissions from the incineration of hazardous waste, by Branscome, et al. (Re-
ference 12). Several of these novel devices appear to be applicable to the
possible need for simultaneous control of vapors (HC1 and vapor phase toxic
metals) and the several toxic metals that appear to concentrate into parti-
cles less than 1 micron diameter. These novel devices are in various stages
of engineering development from bench-scale to full-size industrial instal-
lations. Pilot-scale models are available for nearly all, but only a few
installations exist. The discussion in thj s section is restricted to those
for which encouraging pilot-scale or full-scale results have be^n reported
tnd which show the greatest promise for practical, economical application
within the next year or two.
Wet Electrostatic Precipitators
Process Description. Electrostatic precipitation was discussed in the
previous section. Most of that discussion applies directly to wet electro-
static precipitators (ESPs). The boiler exhaust gas is cooled and saturated
with water vapor in a quench chamber upstream from the wet ESP. As the par-
ticle laden gas flows through the ESP it is subjected to a high voltage field
that impacts electrostatic charges to the particles in a manner analagous to
the dry ESP. The collection electrodes, however, are flushed continuously
with water (or reagent liquid). As particles migrate to the collection plate,
they become entrained in the liquid and are removed with it. The liquid may
be clarified and recirculated.
Because the collection plates are continuously wetted, the wet ESP also
serves as a packed bed gas absorber. Soluble gases (e.g., HC1, 50^, NO?)
in the water and are removed along with the particles.
Development Status. Several companies have produced pilot-scale wet
ESPs and a few fuli-scala units have been installed, mostly in service as
demisters.
Applicability to Industrial Boilers. Wet ESPs are thought to be di-
rectly applicable to industrial boilers. Because the incoming gases are
cooled and saturated, their volume is reduced, Theieiore, a wet ESP can be
smaller than a dry ESP. Pilot tests of two units, installed on a municipal
incinerator demonstrated overall particle collection efficiencies ranging
from 94.3 to 98.8%. Unfortunately, particle size measurements were not made
at the outlet so that expected efficiency of remova. f fine particles was
not confirmed. The efficiency of removal of sulfur dioxide ranged from 70
to 78%; of HC1 from 84 to 98%, and one unit reportedly removed 39% of the
NOX.
3-28
-------
Factors Affecting Performance. The factors affecting the performance
of dry ESPs will, for the m-ot part, affect the performance of wet ESPs. The
exceptions are that reentrai.ment due to low resistivity and gas scrubbing of
the collection plates will oe less in the wet ESP. Once a particle becomes
entrained in the liquid that is flushing, the collection plate no longer needs
to be held to the plate electrostatically. Therefore, low resistivity parti-
cles will not be as readily lost.
Since the wet ^SP will have been selected only in the case where acid
gas (HC1) removal is necessary, its design should incorporate corrosion re-
sistant materials.
Steam Hydro and Free Jet Scrubbers
Process Description. These devices are similar in concept; water is
injected into a gas stream as that stream enters a jet nozzle. The free
jet scrubber relies upon the turbulent mixing that occurs in the nozzle to
provide intimate gas/liquid contact. The steam hydro scrubber directs the
scrubber water flow .->nto a supersonic steam jet in order to break the water
into even smaller droplets. The steam ejector also provides the draft so
that no ID fan is needsd. Mixing occurs in the set nozzle throat providing
for absorption of dust particles and soluble gases. The free jet scrubber
is normally provided as a tandem, two jet nozzles with water sprays, in ser-
ies. The incoming gases are quenched by water sprays prior to entry into
either device. Both provide cyclone separators to remove the water droplets
(and entrained particles) from the gas stream. The devices are essentially
venturi scrubbers. Their manufacturer claims that the tandem free jet scrub-
ber has a lower energy requirement than other scrubber types for equivalent
particle removal efficiency. The use of supersonic steam in the steam hydro
scrubber results in a power consumption of about 10 times that of an air
ejector venturi, albeit particle collection efficiencies are better.
Development Status. Applications of the steam hydro scrubber are us-
ually limited to specialty problems that require reliability and efficiency.
They also are applicable where a source of waste heat can be captured to pro-
vide the needed steam. Free jet scrubbers have been commercially installed
on radioactive waste, PCB, and municipal waste incinerators. As mentioned,
both devices are essentially similar to venturi scrubbers and consist large-
ly of existing technology.
Applicability to Industrial Boilers. Either of these devices ought to
be applicable to industrial boilers though no known installation exists.
Both were pilot tested at a municipal incinerator. The free jet scrubtoer
reduced outlet particle concentrations to 0.014 to 0.032 gr/dscf at pressure
drops between 30 and 40 inches w.c. The steam hydro scrubber gave overall
particle removal efficiencies of 99.9%, the highest attained by any scrubber
in the EPA novel control device research program.
Factors Affecting Performance. As with other scrubbers, energy consump-
tion is the most important factor affecting the efficiency of particle re-
moval by these devices. The case where the boiler exhaust gas contains a
high concentration c': inorganic acids (HC1) can be accommodated by use of a
caustic solution in lieu of water scrubbing liquid and by fabrication of the
unit from corrosion resistant materials.
3-29
-------
Ionizing Wet Scrubbers
Process Description. Ceilcote has developed an ionizing wet scrubber
(IWS*). Tne device is conceptually similar to a tw->-stage ESP. The gas to
be cleaned is first quenched, if necessary, to lower its temperature; it then
passes through a high voltage ionizer section in which the entrained particles
become charged and, finally, through a cross flow packed bed scruober where
the charged particles are attracted to and absorbed by a grounded scrubber
liquor. The scrubber liquor may be water or water augmented with caustic
(lime) for improved HC1 collection.
Development Status. The IWS* is a combination of two existing technol-
ogies: electrostatic charging of particles and packed bed scrubbing. As of
March 1982, the company listed over 30 installations, including 13 at munici-
pal and industrial waste incinerators. The device can be considered to be
existing technology.
Applicability to Industrial Boilers. Although no installations are
known, the IMS* appears applicable to industrial boilers. Natural draft
i.3 lost because of the heat removal that occurs in this (or any) scruboer.
H-wever, the low pressure drop and total energy consumption reported are
attractive.
Facrors Affecting Performance. The incoming gas stream must be condi-
tioned to reduce its temperature, remove large (>5 microns) particles, and
to effect condensation of volatile organic material in order tc attain T.ax-
imum efficiency. Company literature states that a one-stage system will re-
mova 75 to 90% of 0.5 micron particles and that a two-stage unit will remove
93 to 97%. Results of tests of a two-stage unit in service on a reeractory
kiln showed only 50% efficiency of collection of 0.5 micron particles.
Dry Scrubbers
Process Description. The term dry scrubber refers to a device where-
in a powdered absorbing reagent is dispersed into the pollutant laden gas
stream before the gas enters a baghouse. For example, lime can be injected
to the exhaust of a combustion source to absorb, with reaction to CaCl2» the
HC1 generated -uring combustion of chlorinated compounds. There are several
designs which vary primarily in the means by which the reactant powder is
dispersed. Dispersion equipment used includes centrifugal atomizers, spray
nozzles, and venturi throats. At least one manufacturer injects the reac-
tant as a slurry which dries in the gas stream before removal by the bag-
house.
Development Status. Dry scrubbing is a relatively new technolcgy so
there is not a large base of published information about their performance.
A few installations (mostly on hazardous waste incinerators) have shown pro-
mise.
Applicability to Industrial Boilers. There is no known installation
on industrial boilers burning hazardous waste. Although several (about a
dozen) have been sold for the purpose of flue gas desulfurization, the tech-
nical and economic viability for controlling HC1 emissions from a boiler
using dry scrubbing has not been demonstrated.
3-30
-------
Factors Affecting Performance. Operation in an KCI laden gas has been
problematical. The HC1 is, of course, corrosive and ihe reaction product
lCaC!2) is delinquescent and corrosive. Hign chloride is damaging not only
to metallic structural members but also to tne oag material. Two materials,
polypropylene and Teflon4, appear to be satisf cictory . The former should not
be used at temperatures above 235T. Trie latter may oe used at nigner tem-
peratures but is more expensive. However, unless th^ stack gas temperature
is reduced, from the 350° to 400°l:' typical of small ooilers, the efficiency
of removal of volatile metals will not be optimum.
Ammonia Injection
Process Description. Anmonia (NH3) injection ivolves the noncatalytic
decomposition of NOX in the flue gas to nitrogen and water using ammonia as
the reducing agent. This technique is often referred to as selective non-
catalytic reduction or thermal DeNOx. At a mole ratio of 1.5 moles NH3 per
mole NOijj, over 40% of the NO can be reduced if the rerction is designed to
take place at a location in the boiler where the temperatura ranges from
17008 to 1800°F. Outside the range of 1650° to 20008F less than 10% of the
NOX in the flue gas can be reduced to nitrogen and water by ammonia injec-
tion.1 3 since ammonia must be injected into the section of the boiler that
is within the narrow optimal temperature window, some curtailment of load
following capability may result.
Development Status. Ammonia injection is applicable to all industrial
boiler types and fuels where there is access in the proper temperature range.
Although this technique is commerically offered, it is not currently applied
to any domestic operating industrial boiler.13 Ammonia '.njection has been
installed on three gas- and oil-fired boilers ranging in size f^om about 16
to 79 MW (55 to 270 x 1 O6 Bt J/hr ! thermal input in Japan. In the U.S. this
technique ha-s been investigated only on pilot-scale facilities, except for
one commercial installation on a crude oil-fired thermal enhanced oil recov-
ery steam generator. This installation is not currently operating because
of problems experienced with the steam generator.
Applicability to Industrial Boilers. Application of ammonia
injection to industrial boilers is complicated by the frequent load swings
these boilers typically undergo. Since the reaction between NH3 and NO is
efficient in only a narrow temperature range development of an injection
grid is necessary so that the injection can be made in the appropriate fur-
nace zone as load varies.
Factors Affecting Performance. The required reaction temperatures for
noncatalyf.c decomposition of NO with ammonia are found in different areas
of the. boiler depending on its design and operating load. For example, at
full load these temperatures occur in the convective section of both pack-
aged and field-erected watertube boilers. Changing boiler load, however,
causes a shift in the temperature profile through the boiler, reducing NOX
removal to below 30%. For small firetube boilers, optimal ammonia injec-
tion temperatures occur directly in the firebox. In this area of the boil-
er, cross-sectional flue gas temperatures are often not uniform, causing
significant degr?dation of the NO reduction performance to below 10%.
3-31
-------
For new units, multiple ammonia injection grids can be strategically
designed and located to compensate for temperature gradients and shifts in
temperature profiles with changing loads. This technique, however, has not
yet been demonst.vated.1 ^
Other factors affecting performance include 11H3 injection rate and resi-
dence time at optimal temperature. The optimal NH3/NOX molar ratio has been
established to be approximately 1.5, with no additional NO reduction gained
by increasing the ratio to 2.0. Maximization of the residence time at opti-
mal temperature can be achieved by proper location of the multiple injection
grids. A cross-sectional temperature profile will be required for each boil-
er design to identify these locations.^
Effects of Burning Hazarous Wastes. Sulfur-containing wastes present a
potential problem. The formation of ammonium sulfate or ammonium bisulfate
can cause plugging of an air preheater or corrosion of ooiler parts. In-
creased frequency of wcter washing will minimize this problem. To insure
that ammonia emissions to the atmosphere are minimized, ammonia sensors and
feedback control systems for the injectors may be required.
Selective Catalytic Reduction
Process Description. Selective catalytic reduction (SCR) is a technique
involving reduction of the flue gas NOX concentration by reacting NOX with
NHj in a catalytic reactor. With the exception of the use of a catalyst it
is similar to the ammonia injection NOX control technique described above.
In this process, ammonia, taken from a liquid storage tank and vaporized, is
injected at molar ratios of 0.7-1.2 moles NH3 per mole of NOX and mixed with
flue gas prior to the reactor. The flue gas passes through the catalyst bed
where NOX is reduced to ^2- Typically, a 1.0 mole ratio of NH3 to NO should
yield a 90% reduction in NOX emissions. The flue gas exits the reactor and
is sent to the air preheater and, if necessary, further treatment equipment
for removal of particulates and S02. Flue gas must enter the reactor at 350-
400°C since it is in this temperature range that the catalysts show the op-
timum combination of activity and selectivity. The catalysts used in most
SCR processes are oxides of non-noble metals which have shown the best com-
bination of high reactivity and resistance to SOj and 303 poisoning.
Development status. SCR is not considered a commercially demonstrated
control technology for coal-fired sources in the U.S. SCR processes have been
used commercially in Japan on gas-, distallate oil-and residual oil-fired
industrial boilers and SCR processes on coal-fired utility boilers are under
construction. Ongoing studies in the U.S. are investigating NOX only and NOX/
SOX SCR performance with coal combustion in pilot-scale facilities. There is
no full-scale U.S. or Japanese SCR installation with documented performance
in accordance with EPA test methods, although removals in excess of 90* have
been reported for Japanese gas- and oil-fired boiler SCR applications. EPA
is sponsoring two (2) pilot-scale evaluations of SCR technology on coal-fired
utility boilers. The Electric Power Research Institute (EPRI) is also spon-
soring a coal-fir»d utility boiler SCR pilot plant.
3-32
-------
Applicability to Industrial Boilers. SCR is applicable to all indus-
trial boiler types. The particular catalyst formulation and shape as well
as reactor type will be dictated by the fuel fired. Variables associated
with boiler type that car. affect the performance of SCR systems are flue
gas flow rate, NGX concentration, and boiler load variability.
The flue gas flow rate and NOX control level determine the catalyst
volume necessary. increases in either increase the required reactor size.
The NOX concentration is primarily a function of fuel type used.
The system will work well with gas- or oil-fired units using convention-
al catalysts and fixed bed reactor designs. With coal firing, potential ash
plugging problems dictate the use of one or more of the following approaches:
o Operation of the NOX control system downstream of an efficient and
reliable particulate removal device.
o Use of a moving bed design which permits the periodic removal of
catalyst for cleaning.
o Use of a catalyst shape that does not collect the entrained particles
present in the flue gas as they pass through the catalyst bed.
Factors Affecting Performance. An important design variable witn cata-
lytic systems is the space velocity which is expressed as the volume of ca-
talyst required to treat one volume per hour of flue gas. Space velocity
varies with catalyst formulation, catalyst shape, and control level. Both
NH3/NOX ratio and space velocity will range from approximately 1.5 to 8 for
control levels of 70 to 90%.1<* The operating temperature range for most of
these processes is about 300-500°C, though more efficient NOX removal usually
occurs in the higher portion of this range. To maintain the reactor temper-
ature at desirable operating levels during periods of reduced boiler load,
most process vendors recommend bypassing a part of the flue gas around the
economizer. In some pilot-plant and larger operations, auxiliary heaters
have been used to maintain reactor temperatures during turndown.
Impact of Burning Hazardous Wastes. The most probable impact of hazard-
ous wastes on SCR systems is catalyst degradation by metals and by the high
chloride concentrations that result form combustion of chlorinated compounds.
Disposal of spent catalysts may be an environmental concern since some of the
non-noble metals used in their formulations are hazardous. This problem will
be exacerbatad if combustion of hazardous wastes increases the frequency of
replacement. In addition, if metals in the hazardous wastes accumulate in
the catalyst the problems associated with catalyst disposal will be exacer-
bated.
3-33
-------
REFERENCES
1. Devitt, T., Spaite, P., and Gibbs, L. (PEDCo Environmental). Population
and Characteristics of Industrial/Commercial Boilers in the U.S. (Pre-
pared for the U.S. Environmental Protection Agency). Research Triangle
Park, NC, EPA-600/7-79-78a. Cincinnati, Ohio. August 1979.
2. Fossil Fuel-Fired Industrial Boilers - Background Information, Volume I,
Draft Environmental Impact Statement. U.S. Environmental Protection
Agency. EPA-450/3-82-C06a, March 1982.
3. Kemp, V.E. and Dykema, O.W. Inventory of Combustion - Related Emissions
from Stationary Sources (Second Update). Prepared by the Aerospace Corp.,
Environment and Energy Conservation Division for the U.S. Environmental
Protection Agency. EPA-600/7-78-100. June 1978.
4. Surprenant, N.F.. Battye, W., Roeck, D. and Sandberg, S.M. Emissions
Assessment of Conventional Stationary Combustion Systems: Volume V: In-
dustrial Combustion Sources.
5. Cowherd, C., Jr., Marcus, M., Guenther, C.H., Spigarelli, J.I,., and
Venezia, R.A. Hazardous Emissions Characterization of Utility Boilers,
EPA-650/2-75-066. July 1975.
6. Lyon, W. S. Trace Element Measurements at the Coal-Fired Steam Plant,
CRC Press, Cleveland, 1977.
7. Locklin, D.W., Krause, H.H., Putnam, A.A., Kropp, E.L., Reid, W.T., and
Duffy, M.A. Design Trends and Operating Problems in Combustion Modif.\ca-
tion of Industrial Boilers. Prepared by Battelle-Columbus Laboratories
for the U.S. Environmental Protection Agency. EPA-650/2-74-032. April
1974.
8. Air Pollution Control Compliance Analysis Report of Coal-Fired Industrial
Boilers. U.S. Environmental Protection Agency. 340/1-79-005. December
1976.
9. Smith, W.D.., Gushing, K.M. and McCain, J.D. Procedures Manual for Elec-
trostatic Precipitator Evaluation. EPA-600/7-77-059. (June 1977).
10. McCain, J.D., «t al. Results of Field Measurements of Industrial Parti-
culata Sources and Electrostatic Precipiators Performance. J. Air Pollu-
tion Control Association. Vol. 25, No. 2. February 1975.
11. Roeck, D.R. and Dennis, R. Technology Assessment Report for Industrial
Boiler Applications: Particulate Collection. EPA-600/7-79-17Bh. (De-
cember 1979).
12. Branscone, M.R., Wood, J.P., Allen, C.C., Turner, J.H. and Freeman, H.M.
Evaluation for Hazardous Waste Combustion, Final Report. U.S. Environ-
mental Protection Agency, Contract No. 68-03-3149. WA No. 12-1 (April
1983).
3-34
-------
13. Castaldini, C., et. al. Technical Assessment of Thermal DeNOx Process.
EPA-600/7-79-117. May 1979.
14. Jones, G. D, and Johnson, K. L. Technology Assessment Report for Indus-
trial Boiler Applications: NOX Flue Ga:= Treatment. EPA-600/7-79-117g.
December 1J79.
15. Maxwell, J. D.. et. al. Preliminary Economic Analysis of NOX Flue Gas
Treatment Processes. EPA-600/7-80-021. February 1980.
3-35
-------
bUD
SECTION 4
DESTRUCTION EFFICIENCY AND CONTROL TECHNIQUES
The EPA Office of Solid Waste recognized the need to collect emission
data from boilers burning hazardous wastes as fuel:*. Since little data ex-
isted, EPA developed a program to test representative industrial boilers
burning a variety of waste streams. This section evaluates the sampling
and analytical procedures used to collect the field data and summarizes the
results of test burns conducted at 15 sites. This section also includes a
discussion on the types of system modifications that will be required in
order to burn hazardous wastes in existing boilers. This will include mod-
ifications to the boiler, as well as the need for pretreatment and air pol-
lution control equipment.
EVALUATION OF TEST METHODS
The most widely employed procedures for measurement of the rates of
emissions of organic substances from combustion sources are the Modified
Method 5 (MM5;, the volatile organic sampling train (VOST) procedures, and
an adaptation of the VOST protocol for short sampling times and field anal-
ysis called Mini-VOST procedure. These methods are conceptually similar;
bo'ch sampling tra.rns consist: of a particle filter, a condenser, a bed of
porous polymer sorbent, and a condensate trap. Their differences lie in
their size, sorbent, and analytical techniques. Some of the salient at-
tributes of the two methods are compared in Table 4.1. The strengths and
weaknesses of these two methods are discussed in this section.
Modified Method 5 (MM5)
Description of Method
This method is an adaptation of EPA Method 5 (40 CFR Part 60) modified
to obtain samples foe organic compound analysis as well as quantification of
particulate matter emissions. The adaptation (Figure 4.1) is the addition
of a sample gas condenser and a sorbent resin module between the heated fil-
ter and first impinger of the Method 5 train. The sorbent resin most common-
ly used for hazardous waste combustion evaluation is XAD-2 which is highly ef-
fective at trapping organic compounds with boiling points greater than 10C°C.
The sample is collected by isokinetically drawing stack gas through a
heated glass or quartz probe, through a heated glass fiber particle filter
and then to the condenser/resin module. The sample gas is kept above 120°C
4-1
-------
TABLE 4.1
COMPARISON OF MM5 AND /OST PROCEDURES
Feature
Sampling Rate
Sorbent
Analysis of Condensate
Sample Recovery
Technique
Analysis
Boiler Point of Analytes
Sampling Duration
MM5
14-40 1pm
XAD-2
yes
sorbent
extraction
GC or GC/MS
>100°C
1 to 4 hr
V03T
0.5 to " Ipra
Tenax*-CC
no
thermal
desorptior
GC/MS
30° to 10C"C
0.3 to 1 h:.r
Mini-VOST
1 .0 1pm
Tenax®-GC
no
thermal
desorption
GC/HECD
30° to 100°C
10 min
4-2
-------
Tainpeialiu* Senior
Therinuiiiciur
Chuck Valve
Hcveite-Typc I'ilol Tulia
x TlieimomcUT
I lli.:i Holilor
Cll'ilJ :==U SoilMiit Tup
Pilot Manometer
liiipliujeii *• (no ilalli
Oy «V« Valve
Vacuum Line
Oiy Gat Muter Aii-liijlil I'uiup
FIGllKli 4.1
MOniUHD MH1IIOO 5 TRAIN (MMSI
-------
until it reaches the condenser where it is cooled to <20°C- Filtration tem-
peratures up to 205°C are used to minimize organic species condensation prior
'.o the condenser if this does not interfere with the determination of par'ci-
culate matter. The sample gas and condensate pass through a resin bed lo-
cated below the condenser allowing the condensate to percolate through the
bed and collect in an impinger or condensate trap underneath the resin module.
The sample gas is then bubbled through two more impingers in the conventional
Method 5 configuration for acid gas and additional moisture removal.
Samples are analyzed by performing solvent extractions on the probe and
filter material as one fraction, the resin as a second fraction, and option-
ally on the condensate and impinger catches as third and fourth ~r actions.
The solvent extracts are concentrated and may be combined and analyzed by gas
chroma tography using mass spectrometers, flame ionization detectors.-, or elec-
tron capture detectors, as appropriate, for the organic compounds o° interest.
Procedure Standardization
The construction and operation of a Method 5 train is well known and
well described in the literature. The train operation, sample recovery,
choice of sorbent resin, and analytical method to be used in a given appli-
cation of MM5 are not specified. At this time, a single description of how
the KM5 train is or may be used is not available. The sample collection
aspects including resin choice, are discussed in various publications, but
the sample recovery, analysis, and data reduction are not well described.
The Method 5 tr?.in and sample collection scheme has become the standard
for measurement of emissions of particulate matter and it is not surprising
that modifications to it have evolved in attempts to quantify other types of
emissions using the same equipment and techniques. Method 5 is written as a
compliance test method for certain specific categories of sources subject to
NSPS. This specificity has not been established for the MM5 procedure as ap-
plied to hazardous waste ORE sampling in part because at least some of the
measurement objectives will be different from test to test. Many of the ob-
jectives will be the same and these need to be identified and the method
written to ensure that the ooals are clear.
First, the relati"e priorities of measuring particulate matter and or-
ganic compound smissio is need to be set. Most often these will not conflict
and no sacrifice of one measurement for another is needed. In the cases
where some relatively significant quantity of organic chemical of interest
may be found in the "front half" (probe, filter, connecting glassware) of
the Method 5 train then the ac.^uracy of particulate matter measurement may
need to be sacrificed to obtain the S-VOC data. Examples are not brushing
the probe to avoid contamination, removing some particulate matter with the
solvent used for S-VOC recovery, and/or operating the heated portion of the
train at higher than normal temperature (e.g., 205°C) and vaporizing parti-
culate matter which would otherwise deposit in the front half of the train.
In the case of oil-fired boilers, much of the particulate matter is carbon-
aceous with little inorganic ash and the material contains straight chain
and aromatic hydrocarbons which may or may not be collected in the front
half depending on the filtration ttanperature. This is different from a coal-
fired unit where the ash is mostly (typically >95%) inorganic matter. If
4-4
-------
POHCs or PICs to be measured from hazardous waste combustion are to include
naphthenic or paraffinic hydrocarbons, then whether the boiler is oil- or
coal-fired will probably affect the sampling and analysis scheme.
Selection of POHCs and PICs also bears on the choice of a suitable sor-
bent resin. Generally, the MM5 is used to sample for PICs - the higher boil-
ing compounds, but it is also used to collect samples for certain Appendix
VIII compounds, a.g., toluene, monochlorobenzene. XAD-2 is the resin most
commonly used and recommended as a general purpose sorbent when solvent ex-
traction is the means of sample recovery. XAD-2 was selected for its sor-
bent properties, ease of cleaning, and sample recovery efficiency. Others
are also used, for example Tenax^-GC, if thermal desorption is the sample
recovery procedure. However, selection of a sorbent resin usually involves
a great deal of time and effort (literature and laboratory research) that
is usually not practical on a case-by-case basis.
The last major uncertainty in regard to applying the MM5 to boiler ORE
testing is analysis - compound identification and quantification. Currently
gas chronatography is used with one of three detection modes MS, BCD, and
FID. Each has its own benefits and disadvantages relating to sensitivity,
reproducibility, compound identification, interference rejection, and ana-
lytical cost. Each of those factors needs to be considered and the method
written to describe when each would be appropriate.
QA/CC Evaluation
Most of the QA emphasis has been on obtaining acceptable blank values
and preventing contamination. The large gas sample and relatively large quan-
tity of resin concentrate the sample which, in conjunction with the solvent
extraction, usually results in a greater analyzable mass than methods employ-
ing a lower sample volume. This greater analytical mass tends to decrease
the importance of trace contamination of the sample. The resin cleaning,
blank extraction, field trip, and laboratory handling blanks are adequate
to identify problems ark. likely causes.
There is one \mportant aspect lacking in the method as it is being v. sed
and this is use of field spikes to check on sample loss and recovery effi-
ciency. As currently practiced, the procedures does not provide a means for
determining target compound collection efficiency or for evaluating sample
recovery. The nature of the sample and the field conditions preclude the
usual option of splitting the collected sample and spiking one split with
a target compound to check loss and recovery. These spikes and replicate
analyses can be performed in the lab using the extractate with some loss of
sensitivity. It is also possible, a ad it seems highly advisable, to spike
the field samples with a tracer compound having properties similar to the
target compounds.
Because the volume of gas sampled by MM5 is large, the total quanti-
ties of the various semi-volatile organic compounds (S-VOC) in the samples
ranged from a few to several hundred micrograms. Thus the MM5 results are
not greatly influenced by even a few hundred nanograms of contaminants.
Two quality assurance acceptance criteria were applied to the MM5 data
before they were included in this document.
4-5
-------
1. The recovery of surrogate or spike compounds added to the sample
before analysis must have been in the range of 50 to 150%.
2. The rate of feed of any given compound rn^st have been 10,000 times
tne minimum detectable limit of the MM5 procedure.
The first of these is merely a demonstration of acceptable analytical
accuracy. The acceptable range is wide relative to normally attainable ana-
lytical precision. It is adequate for this analysis because the DR3 calcu-
lation is insensitive to an error of a factor of three (3) in the emission
rate measurement.
Th? calculation of ORE, the primary use of the data in this document,
is ba.<;ed upon the concentration of the various constituents in the waste
feed stream and in the exhaust gas. The accuracy and precision ot these
concentration measurements decrease when they are near the limit of detec-
tion of the analysis methods. Consequently, the dispersion of DRE values
that are calculated based upon these imprecise measurements becomes UP.ac-
ceptably large. The decision was made to include only those data chat are
as accurate as is possible using the available methods. The ruethod chosen,
the second of the quality assurance acceptance criteria, accomplishes this
goal.
The blank MM5 samples were, for the most part, uncontaminated. The sam-
ples contained many times more of each compound of interest than the blanks.
Correction of the sample values by subtraction of the blank values would have
had little effect on most of the results. Therefore, the decision was made
that no blank corrections would be done for the MM5 data.
At most sites a baseline run was performed. This consisted of the taking
of a MM5 sample while the boiler was burning only virgin fuel. In many cases
the emission rates of waste fuel constituents measured during the baseline run
was of the same order of magnitude as their emission rate during the co-fired
runs. No satisfactory explanation coulo be found for these observations. The
baseline emission rates were not subtracted from the co-fired emission rates
before calculation of DRE values. Although a case could be made for making
the subtraction, it was decided to take the conservative approach and to err
on the side of safety.
Volatile Organic Sampling Train (VOST) ?nd Mini-VOST
Description of Method
The basic details of construction and operation of the VOST are des-
cribed in the "Protocol for Collection and Analysis of Volatile POHCs Using
a Volatile Organic Sampling Train (VOST)" by Envirodyne Engineers for IERL
(Reference 1). The highlights of the procedure are described below. Stack
gas is drawn through quartz wool particle filter in a glass or quartz probe
heated to approximately 130°C, through a three-way stopcock and through a
coil condenser. Following the condenser, the sample gas passes through a
glass tube containing 1.6 g of Tenax*-GC, a condensate trap, a second con-
denser, and a second sorbent tube containing 1.0 g Tenax*-GC followed by
1.0 g of activated charcoal. A second condensate trap is next, followed by
a silica gel drying tube, and the pump/flow control system (sample lines,
4-6
-------
valves, flow meters, gas meters, etc,). All portions of the sample line
precedang the last condenser are glass, stainless steel, or Teflon*. Figure
-1.2 is a schematic depiction of tha train. Previous experiments have estab-
lished 20 liters as a maximum safe sample volume for this train. A greater
sample volume incurs trie risk of stepping sorbed POHCs off the resin. T;,=
sample rate for these tests was 0.5 L/niri *or a total of 40 minutes per pair
of tubes.
An ice water bath is used to circulate wacer through both condensers to
maintain sample gas temperature below 20°C through the sorbent tubes. Tem-
perature of the probe liner, first ccndenber oatlet, ambient air, and dry
gas meter are measured and recoried. Leak checks of the whole train and
each pair of sorbent tubes for each run are conducted and the resulting va-
cuum is released by allowing ambient air in through a charcoal filter con-
nected to the three-way
Tubes ased for the boiler tests were of :he inside-inside design that
are held in the sample train with stainless '3 tee? Swagelok® fittings and
ceramic-filled Teflon® ferrules. Other samplers have used the inside-out-
side design; a double walled sorbent cartridqe/shipping container that uses
0-rings and end caps to seal the cartridges. Stainless caps are used to
seal tube ends for shipment before and after sample collection. After sam-
ple collection, tubes are kept and shipped in chilled styrofoam containers.
The tubes were analyzed on a GC/MS using thermal desorption with trap
and purge. The method is described in the protocol and involves spiking
each tube or pair of tubes with an internal standard, thermally desorbing
the tubes into a water trap, and purging the water trap onto an analytical
column for component separation. Identification and quantification are made
by elation times, characteristic ions, and ion current profile using a com-
puterized data library.
The Mini-VOST method is an adaptation of the VOST protocol for short
sampling times and field analysis. The sample itself is similar to the VOST
procedure described in the previous paragraphs, except the second condenser
aud backup sorbent cartridge are eliminated. Also, a knockout was placed
after the first condenser to collect condensata. A sample volume of 10 li-
ters was used, which is one-half the VOST protocol volume. The sample was
collected at a rate of 1.0 L/min for 10 minutes.
Standardization of Procedure
The VOST protocol provides clear and specific directions about the sample
train to oe used and the method of sampling and the method of analysis. Re-
agent preparation, sample handling, QA/QC activities, calibration and calcu-
lations are all described in detail. The protocol states that conditioned
cartridges, as well as used ones with sorbed sample, be kept in ice water
before use and after sample collection. This ice water storage is net re-
quired if acceptable blank levels can be maintained.
Options regarding sample collection and recovery (analysis) efficiencies
are also provided along with evaluation criteria. The analytical procedure
is also described very specifically.
4-7
-------
Gluis Wool
P
txlicnul
ft I/mlu
Dry GUI
Meier
FKJdKH 4.2
Vulnl.Hi:
niini|i| lii|{ train (VOST).
-------
Mini-VOST procedures have not been standardized by EPA except for tnose
steps that are identical to the VOST protocol. This procedure is still being
evaluated by EPA.
QA/QC Evaluation
The following paragraphs describe the QA/£C criteria used in evaluating
the VOST data. No criteria were established for the mini-VOST results which
were accepted or reported as valid by the contractor performing the tests.
The pre-sampling QA activities are clear and direct. Tenax® and char-
coal cleaning, tube packing, desorption blanking, provide sufficient assurance
that the sample cartridges start clean. The trip blank, field blank, and lab
blanks are intended to provide a history and background levels of contamina-
tion and/or degradation so that the results of the sample analysis reflect
only POHCs present in the stack gas. This history is especially difficult
to create if the samples and blanks are not analyzed promptly. Sample degra-
dation/ tube cross-contamination, contamination from external sources (lab
air, ambient air, etc.), and calibration and response standard degradation
become more likely and less distinguishable.
As each analysis is a one time occurrence (no way to split), if an
analysis is bad, a tube contaminated or otherwise invalid, the data point
is lost. As the sample collection involves three pairj, the duplication is
inherent in the sampling. One could carry this suggestion further to say
that two backups are desirable to yield an average value and increase the
confidence in the results.
The protocol requires one exposure pair per six pairs of sample tubes.
These exposure (field or shadow) blanks are opened as if they were sample
tubes but are not installed in the train.
C.ie 2«". »ction which has not been done, for several reasons, is spiking
tubes in the field with one or more target POHCs or surrogates to establish
recovery efficiencies. The difficulty stems from two major areas. One is
the difficulty in maintaining reagent and tube purity in a field environment
and tne othr.r is not being able to split a single sampling and spike one por-
tion. The methods used to date have centered on lab simulations and the ana-
lytical process. A suitable field spike procedure would yield data on sample
loss (leakage out) and desorption efficiency as well as additional data on
contaoination, lab QA, and overall method validity.
The following list identifies areas where additional effort and investi-
gation could strengthen the VOST procedures.
o Spike blank cartridges in the field with a labeled compound to detect
potential leakage during field storage and subsequent transport.
o Analyze the sample immediately with as little storage time in field
and lab as possible.
o Conduct a detailed systematic evaluation of field handling, field
ctorage, shipping, and lab storage, to identify potential causes of
4-9
-------
contamination and/or leakage. Develop guidelines to detect and pre-
vent contamination and to leak check cartidges.
o Investigate the current seal design. Do temperature cnanges cause
leakage? How can one be certain the tubes are perfectly sealed.' How
can overtightening/undertightening be prevented? Can a pressure/va-
cuum tighc seal be obtained repeatedly on a large number of tunes
with no failures?
o Investigate cross contamination. Place spiked samples and clean
blanks in the same container, store one to two weeks, and analyze.
Do the above with loose fittings or cracked tubes and observe for
cross contamination. Place the samples and lab blanks with their
double seal in an atmosphere containing trace amounts of metnylene
chloride or waste fuel vapors, store, and analyze.
Quality assurance acceptance criteria were developed for the volatile
organic compound data produced during each test. These criteria could not be
identical for all sites since the methods used differed. In all cases only
those components of the waste listed in Appendix VIII (C5R 40 Part 261) were
included in the ORE results; even though there may have been other organic
constituents measured in the waste feed and stack exhaust streams.
For most of the test performed daring this program, three pairs of
tubes constitute a sample run, this run sample would also include one expo-
sure blank pair and one unopened (trip blank) pair. It is noted that not all
of the volatile organic compound sampling done during these tests were done
by the VOST procedure as has been described. The train used to sample vola-
tile organic compounds (VOC) at Site C consisted of two tubes of Tenax^-GC
in series. Sites 0 and E were sampled with one Tenax*-GC and one Tenax*-GC/
charcoal tube. No condenser or other sorbent temperature control was used
at Site C. Impingers were inserted in the train upstream and downstream of
the Tenax*-GC tube at Site D and E.
Site M was sampled by EPA Method 23. Samples were collected in inert
plastic bags rather than on sorbents.
The volatile organic compound sampling trains used at Sites D and E
varied from the VOST train in that an impinger (containing water) in an ice
bath was inserted in the sample line ahead of the Tenax*-GC cartridge. A
second impinger was placed between the Tenax*-GC and the Tenax*-GC/charcoal
cartridge at Site 0. Three quality assurance acceptance criteria were ap-
plied to these data. They were:
o The contents of both impingers must have been analyr-4.
o Both sorbent tubes must have been analyzed.
o At least 70% of the total quantity of any compound found on the
sorbent tubes must have been found on the first (Tenax*-GC) tube.
The first two (2) of these are completeness criteria. While analysis
of the condensate is not normally a part of the VOST protocol, it is neces-
sary in these instances because of the location and the temperature of the
4-10
-------
condensate trap. The volume of condensate obtained from a 20-liter sample
of boiler stack gas is approximately 1 .5 ml. Even though the compounds of
interest (mostly chlorinated hydrocaroons) are normally considered to be
insoluble in water they are miscible to a small but measurable extent. A
compound soluble to 1 mg/liter is said to be insoluble yet that translates
to 1500 ng/1.5 ml of condensate which is large relative to the analytical
quantities of interest. The VOST train causes the condensate to be drawn
through the resin bed. The resin should remove the compounds from the con-
densate. Since there was no contact between the condensate and the resin
at these sites, it was necessary to analyze the condensate.
The third acceptance criteria was included to eliminate contaminated
samples from the data. Persons who have sampled surrogate stack gases spiked
with chlorinated hydrocarbons under controlled conditions have reported that
at least 90% of these compounds are sorbed on the first resin trap. This is
not true of highly volatile compounds, e.g., vinyl chloride, but it is for
the compounds of interest in this document. Tenax*-GC has sufficient affi-
nity for these compounds to remove them nearly quantitatively from the sample
gas stream. The charcoal, used as a back-up sorbent in the second cartridge,
has a much greater sorbent capacity and affinity for these compounds. Thus,
if the cartridges are exposed ~.o contaminants, the second tube should sorb
them at a higher rate. Therefore, setting the acceptance criteria at 70%
allows acceptance of some contamination but rejects grossly contaminated
cartridges.
At the remaining sites (F, G, H, I, J, K, M, and N) the VOST train (Fi-
gure 4.2) was used. Acceptance criteria »1 is not applicable but the other
trfo are.
The procedure employed herein for making blank corrections follows that
suggested by the VOST protocol. The mean an-! standard deviation of the quan-
tity of each compound of interest was found from all blank tubes (field, trip
and laboratory) analyzed. If the quantity of a compound was greater than the
average blank value by an amount equal to or greater than three (3) times the
standard deviation of the blanks, then (and only then) the average blank vol-
ume was subtracted from the test run value.
The effects of application of these quality assurance criteria on the
data are discussed for each test in the following pages.
TEST BURN RESULTS
The initial surveys that were performed during the early stages of de-
velopment of this background information document (BIO) revealed that ade-
quate information about the destruction and removal efficiency of hazardous
compounds by industrial boilers did not exist. EPA has also developed ex-
tensive data that reported the ORE (ORE includes both thermal destruction
and removal by control devices) of hazardous compounds by incinerators.
There are significant differences in the two processes. Incinerators typi-
cally hold their combustion gases in an oxidizing atmosphere at temperatures
ranging from 1800 to 2500°F for times ranging from 2.0 to 3.0 sec. The com-
bustion zone temperature in boilers is typically higher (2250 to 3000*F) but
4-11
-------
the retention time is typically less (approximately O.S sec). However, kine-
tic theory predicts that elementary reactions should be faster at the higher
boiler temperature by a factor ranging from 4 up to 20,000, depending upon
the activation energy of the particular reaction. This range has been con-
firmed by thermal destruction analytical system (TDAS) data for many common
hazardous compounds. These data demonstrate that rates of destruction in-
crease by factors that range from 17 to 12,000 when the temperature is
raised from 1900 to 3100T.
EPA undertook a series of tests to determine whether the destruction
of hazardous materials by co-firing in industrial boilers is adequate to
protect the environment. Beginning in March 1982, EPA performed tests at
12 industrial boilers that are representative of the range of boilers, co-
firing liquid hazardous wastes. The results of these tests are described
in this section.
The sources that have been tested are characterized in Table 4.2. The
boilers range in size from 1 to 32 kg steam/sec (8000 to 250,000 Ib steam/
hr). With the exception of the boiler at Site G they were standard indus-
trial boilers that are representative of the various boiler types currently
in use. The boiler tested at Site G was specially modified to burn hazard-
ous compounds. The modifications consisted, primarily of rearrangement of
internal baffling so as to maintain surface and gas temperatures favorable
for destruction of hazardous compounds. The remaining boilers were indus-
trial equipment that had been modified only to the extent of providing a
means for injection of hazardous wastes.
Destruction and Removal Efficiencies of Principal Organic Compounds (POHCs)
The data from all test sites have been reviewed and subjected to the
quality assurance acceptance criteria that were presented previously. The
results at each site are presented individually. An overview of the impli-
cations of all of the data are summarized following the individual test dis-
cussions.
The data are summarized, by site, in Tables 4.3 through 4.30. Each
table heading reiterates some of the source characterization data that were
presented in the previous section. The compounds listed are those that were
measured in the waste and that are named in Appendix VIII (Reference 1).
Site A
The boiler tested at Site A was an old coal stoker that is used to burn
waste wood and waste creosote at a wood preserving plant. Waste wood chips
and saw dust are transported to the boiler by a conveyor system. Waste creo-
sote trickles onto the wood from a storage tank that is suspended above the
conveyor. No means to measure either the wood waste or the creosote waste
flow rate was available during the tests. The test team attempted to measure
the creosote flow rate by recording the change in the liquid levels in the
creosote storage tank. The tank was large, however, so the rate of change
of the level was small (approximately 2 inches/hour) and the rate measure-
ment was unreliable. More uncertainty was added by the observation that
4-12
-------
TABLE 4.2
COMPLETED FIELD TESTS ON IIAZAKUUUS WASTE CO-FIKING
*>
OJ
Site
Desig-
nation
A
B
C
Boiler Type
Keeler CP 308-hp
(10,000 Ib/hr of
steam) water tube
boiler
Cleaver-Brooks
250-hp (8,400 lb/
hr of steam) fire
tube boiler
Babcock & Wilcox
29-kg/s (230,000
Ib/hr of steam)
multiburner
wa te r tube
Primary
Fuel(s)
Wood chips,
Lark, and
sawdust
Natural gas
Natural gas
Waste Description
Creosote sludge containing
chlorinated aromatics in-
cluding pentachlorophenol
and chlorinated dibenzo-
dioxins
Alkyd wastewater with paint
resin containing toluene,
xylenes, and several acids
Phenolic waste containing
phenol, alkyl-benzenes, and
long-chain aromatic and
aliphatic hydrocarbons
Emission
Control
Source
Multiclone
for
particulate
collection
None
None
Sampling Protocol
o Flue gas
— Modified EPA Method b for
se;ni- & nonvolatile orgar.ics
-- Tenax** train fur volatile
organics
-- Continuous monitors O^ , CO^,
CO, NOX, and TUHC
o Bottom ash
o Cyclone collected ash
o Ambient hydrocarbon levels
o Flue gas
— Modified EPA Method 5 for
semi- & nonvolatile organics
— Tenax"* train tor volatile
organics
— Continuous monitors U^i CO^,
CO, NOX, and TUHC
o Flue gas
— Modified EPA Method b for
semi- & nonvolatile organics
— Venax® train for volatile
organics
— Continuous monitors (>2 , CO^,
CO, NOX, and TUHC
o Ambient hydrocarbon levels
-------
TABLE 4.2—Continued
I
*.
Site
Desig-
nation
U
Boiler Type
Babcock & Wilcox
11.4-kg/s (90,000
Ib/hr of steam)
multiburner
water tubea
Primary
Fuel(s)
No. 6 oil
Waste Description
o Waste stream no. 1
— Mixture of methanoJ ,
xylenes, and tetta-
chloroe thy lent!
o Waste stream no. 2
— Mixture of toluene and
bis ( 2-chloroethy 1 )
ether
Emission
Control
Source
Noneb
Sdinpliny Protocc
o Flue yas
— Modified EPA Metlioc
semi- k nonvolatile
— Tenax1* train tor vc
oryanics
— Modified EPA Mc:tho<
HCi emissions
-- Cj to C(j hydrocarb<.
chromatoyraphy
— Continuous monitors-
Co, NOX, and So^
o Fly ar'i
0 tor
Combustion Engi-
neering 13.9-kg/s
(110,000 Ib/hr of
steam) single
burner packaged
water tube
No. 6 oil
and natural
gas
o Waste stream no. 1
— Mixture of methyl
methacrylate, methoxy
butanone, methyl
methoxy butanone, and
fluxing oils
o Waste stream no. 2 and
waste stream no. 1 arti-
ficially spiked with:
— Carbon cetrachloricle
— Chlorobenzene
— Trichloroethylene
o Waste stream no. 3, mix-
ture of toluene, and
methyl methacrylate
None
o Flue gas
— Modified EPA Method 5 tor
semi- & nonvolatile oryanics
— Tenax* train for volatile
otganics
— Modified EPA Method 6 for
HCL emissions
— Cj to C(, hydrocarbons by yas
chromatoyraphy
— Continuous monitors 02, (^(^2>
CO, NOX, and
-------
TABLE 4.2—Continued
Site
Desig-
nation
F
Boiler Type
Babcock & Wilcox
7.6-kg/s (60,000
Ib/lir of steam)
raultiburner
water tubea
Primary
Fuel(s>
No. 6 oil
Waste Description
o Purge thinner containing
mixed methyl erters.
butyl cellosolve acetate,
aromatic hydrocarbons.
aliphatic hydrocarbons
o Spiked with chloroben-
zene, trichloroethylene,
and carbon tetrachloride
Emission
Control
Source
None
Sampling trot
o Flue yas
— Modified liPA Met
semi- & nonvolat
— Volatile oryanic
train tor volati
— Modified 1-JPA Met
IICl emissions
— Continuous inonit
CO, NUX, *nd TUII
in
Modified fire
tube boiler 5.0-
ky/s (40,000 lb/
hr of steam or
1,2OO-hp)j ther-
mal hot recovery
oxidizer (THROX)C
None -
natural gas
used only
for startup
o Mixture of chlorinated
hydrocarbons containing
up to 55% by weight chlo-
rine; major components:
— Bis(2-chloroisopropyl)
ether
— Propylene trichloride
— Epichlorohydrin
o Spiked with carbon tetra-
chloride
Two chloride
recovery/re-
moval water
scrubber
columns in
series
o Flue yas
— Modified EPA Method 5 tor
semi- & nonvolatile oryanios
— Volatile oryanic samp liny
train for volatile oryanics
— Modified EPA Method 6 for
HCl emissions
— Continuous monitors
U2, C02<
CO, NOX, and TUI1C
o Makeup scrubber water
— Chloride
o Scrubber discharyes
— Chloride
— Volatile oryanics by purye
and trai (EPA Method b24)
-------
TABLE 4.2—Continue.!
Site
Desig-
nation
H
Boiler Type
Combustion Engi-
neering tangen-
tial NSPS coal-
fired 31.5-kg/s
(250,000 Ib/hr
of superheated
steam) boiler
Primary
t'uel(s)
Pulverized
bituminous
coal
Waste Description
Crude methyl acetate;
spiked with trichloro-
e thane, carbon tetrachlo-
ride, and chlorobenzene
Emission
Control
Source
Cold side
electro-
static
precipi tator
Sampling Protocol
o Flue gas
— Modified EPA Method 'j lor
semi- & nonvolatile orgaiu
— Volatile organic sampling
train for volatile organic
— Modified EPA Method 6 for
HC1 emissions
— Continuous monitors 03, C(
CO, NOX, and TUHC
o Bottom ash and fly ash for
nonvolatile organics
Foster Wheeler
type AG252 bent
tube boiler 7.8-
kg/s (68,000 lb/
hr) of steam
Fuel oil o Waste gas (mostly me-
or gas thane)
o Steam stripper waste from
aniline production spiked
with chlorinated hydro-
carbons
None
o Flue gas
— Modified EPA Method 5 for |
semi- & nonvolatile organics
— Volatile organic sampling
train for volatile organics
— Modified EPA Method 6 for
HC1 emissions
— Continuous monitors CO,
O2, NO
X'
and TUHC
o Composite liquid waste samples |
_ I
-------
TABLE 4.2--Continued
Site
Desig-
nation
J
Boiler Type
North American
model 3200X fire
tube boiler \ .06-
kg/s (8,400 lb/
hr) of steam
Primary
Fuel(s)
Natural gas
and No. 6
oil
Waste Description
Blended for test 9U% tol-
uene spiked with carbon
tetrachloride, chloroben-
zene, and trichloroethylene
Emission
Control
Source
None
Sampling 1'rotocol
o Flue gas
— Modified EPA Method 5 ft
toluene and chlotobenzei
— VOST for carbon tetrach.j
ride and r.rictiloroe'ehylt
— Modified EPA Method 0 tc
HC1 emissions
— EPA Method 2J for all w<
components
— Continuous monitors (J^i
CO, NOX, and TUHC
co2,
t^
-J
K
Combustion Engi-
neering VU-10
wa te r tube 7.6
kg/s (60,000 lb/
jhr) of ste^m
Heavy and
light fuel
oils
o Waste solvent
o Light oil spiked with
chlorinated hydrocarbons
None
o Flue gas
— Modifier) EPA Method 'j for
semi- & nonvcl.iti le organics
— Volatile organic sampling
train for volatile oryauics
— Modified EPA Method 6 for
HC1 emissions
— Continuous monitors CO, CO^,
°2» N°x- and TUIIC
o Composite liquid waste samples
-------
4.2—Continued
Site
Desig-
nation
L
/
Boiler Type
Combustion Engi-
neering 13.9 kg/s
(110,000 Ib/hr)
of steam single
burner packaged
water tube
Primary
Fuel(s)
No. 6 oil
and natural
gas
Waste Description
Methyl inethacrylate rtill
bottoms spiked with carbon
tetrachloride and chloro-
benzene
Emission
Control
Source
None
Sampling Protocol
o Flue gas
— Modified EPA Method b tor
semi-volatile PICs
— VOST for volatile POHCs and
PICs
— Field analyzed Tenax"" tubes
for volatile POtiCs and PICs
— Total organic chlorine by
electrical conductivity
detector
— Continuous monitors for O->,
CO2, CO, TUHC, NOX, S02,
and smoke
oo
Combustion Engi-
neering 44.2 kg/s
(350,000 Ib/hr)
of steam, multi-
pile burners
Natural gas
o Waste stream No. 1 -
butanol/propanol heavy
ends spiked with carbon
tetrachloride, chloro-
benzene
o Waste steam No. 2 —
process - waste gas con-
taming principally, CH4,
C2H6» C2H4' co» H2' and
aldehydes
None
o Flue gas
— Modified EPA method 5 for
semi-volatile PICs ar.d HCl
— VOST with post test analy-
sis for volatile POlics and
PICs
— Mini-vost with onsite anal-
ysis for volatile POHCs and
PICs
— Continuous monitors 02,
NOV
and TUHC
C02. CO,
o Waste oil
— Grab samples with post test
ultimate and POHC analyses
a Boiler originally stok-sr coal-fired converted to oil burning.
b Some particulate collected by existing hopper cavities.
c Patented process for neat generation and chemical recovery of highly halogenated hydrocarbons.
-------
TABLE 4.2--Continue]
Site
Desig-
nation
N
Boiler Type
Riley spreader-
stoker coal-
tired 19.0 kg/a
(150,000 Ib/hr
of steam)
Primary
Fuells)
Coal
Waste Description
o Inorganic sludge consist-
ing primarily of water
(79.5%) and served inor-
ganic elements
o Spiked with trichloro-
ethy lene
o Spiked with 1,2,4-tri-
chlorobenzen, lead, and
chromium
o No. 2 fuel oil spiked
with trichloroethy lene
o No. 2 fuel oil spiked
with 1 ,2,4-trichloroben-
zene
o No. 2 fuel oil spiked
with 1 ,2 ,4-trichloroben-
zene, lead, and chromium
Emission
Control
Source
Mechanical
collection
and baghouso
in series
Sampling Protocol
o Flue gas
— Modified EPA Method 5 for
particulate mass, trace
metals, and seini -vola t ile
organics
-- Vost with jxjstteiJt analysis
for volatile POIICs and I'iCs
-- Anderson ir.ipactor for parti-
cle size distribution and
trace metal partitioning
— Mini-VOST with ons i te anal-
sis of volatile chlorinated
organics
-- EPA Method 5 for particulate
matter
— Modified EPA Method 6 for
HC1
-- Continuous monitors O-^i '-^2
CO, NOX, SO2, and TUIIC
— Coal for ultimate and proxi-
mate analyses arid metal
-- Waste fuel (sludge or No. 2
oil) for POIICs, metals, and
chloride
— Baghouse ash for trace
metals and semi-volt ile
— Mechanical collection ash
for trace metals and semi-
voltile
— but torn ash for metals and
semi-volati Jes
-------
TABLE 4.2—Continued
Site
Desig-
nation
O
Boiler Type
Modified Combus-
tion Engineering
coa L stoker 2.8-
Jcg/n (22,000 lb/
hr ) steam
Primary
Fuei(s)
Natural gas
a nd No . 6
oil
Waste Description
Alcoholic still bottoms
with methanol, methyl ace-
tate, and methyl chloroform
Emission
Control
Source
None
Sampling Protocol
o Flue gas
— EPA Method 23 for w.iste
com^xinents
— Continuous monitors CO^, CO,
O2, and TUUC
— Modified EPA Method 6 for
HCl emissions
I
K>
O
-------
the wood waste apparently was contaminated with creosote. This; observation,
made qualitatively by the test team members based upon the odor of the wood,
was supported by the results of the gross calorific value analyses of the
wood waste. These analyses, together with typical heating values for wood
found in the literature, led to the conclusion that the wood waste may have
contained 25% creosote by weight. The relatively high erru ssion rates of
phenol and naphthalene (Table 4.3) during the baseline nan add credence to
the supposition that the wood waste contained creosote. Unfortunately, it
is not possible to verify that the wood was contaminated. Therefore, the
DREs of the various components are based upon the reported creosote ff»sd
rates.
The boiler thermal efficiency as determined by the ASME heat loss method
was only 63%. The carbon monoxide concentration in the flue gas (Table 4.16)
averaged 1,200 ppm. These observations imply that the boiler did not main-
tain good combustion conditions.
All of the hazardous components of the waste stream had boiling points
in excess of 100°C so all of the POHC emission rates were based upon MM5 test
results. Table 4.3 gives the feed rates, emission rates, and DREs for the
seven Appendix VTII compounds for which the data met the quality assurance
acceptance criteria. The report included data for 12 compounds that a-e
not listed in Appendix VII. The DREs of these compounds was approximately
99.99%. The data for two compounds, 2,4-dimethyi phenol, and 4-nitrophenol
were omitted because their concentration in the feed was too low. The data
for one compound, pentachlorophenol, were omitted because its spike recovery
(33%) was outside of the QA acceptance range.
Less than 1% of industrial boilers burn wood waste (or bagasse or other
hog fuels) so it is not representative of a large number of boilers. Its op-
eration (high CO, low efficiency) was not representative of good boiler com-
bustion. The rate of feed of hazardous materials was not well documented;
it probably was under-estimated by more than a factor of two. The DREs were
marginally acceptable. This boiler may be operating just outside of the
range of conbusticn conditions that provide adequate DRE of hazardous mate-
rial. It, and others similar to it, seems to have the potential to destroy
hazardous material but a trial burn demonstration of that potential should
be required in individual cases.
Site B
The boiler at Site B was a gas-fired fire-tube in which alkyd resin
waste was co-fired. The waste stream was largely water. The waste holding
tank was nearly full during the tests so that the agitator could not be turned
on without causing it to overflow. The lack of mixing allowed the organic
material to separate and float to the top. Since the waste feed line was
below the phase boundary the waste fed during the tests was more than 99.7%
water. It is interesting to note that the water, which was 14% of the mass
of the natural gas (Table 4.4), had no obvious deleterious effect on the
boiler combustion. Neither the CO nor the total unburned hydrocarbon con-
centrations in the flue gas increased over their concentrations during the
baseline run when the wastewater was co-fired. The test data do not allow
4-21
-------
TABLE 4.3
SITE A
.nufacturer: Keeler CP Design Steam Rate: 10,000 lo/hr
•pe: Water Wall (solid fuel) Design Steam Pressure: 250 psig
lei: Wood Waste Test Steam Rate: 10,000 Ib/hr
.ste Stream: Creosote Waste Fraction Waste Fuel
Mass: 17%
Heat Input: 40%
Comoound
phenol
2, 4-dir.itrotoluene
2 , 6-dini troto luene
naphthalene
f luorene
chrysene
bis(2-chloro«thoxyl)
methane
mass-weighted average
Quantity Found (ug) Test
Feed Base- Aver- Test Emission Blanka
Rate line age Runs Rate Cor-
mg/sec Run Blank (avg) (ug/sec) rected
44.7 350 ND 11 80 No
13.0 ND ND ND <0.6 No
21.7 ND K'J <0.7 No
507 39 23 165 90.5 No
283 SD ND 23 <22.3 No
40 ND ND 29 3.0 No
8 ND ND ND <0.6 No
ORE
99.821
>99.995
99.997
99.982
99.992
99.993
>99.994
99.378
NOTES: 1) ORE m (Feed Rate) - (Test Emission Rate) x , Oo*
Feed Rate
2) Waste feed rates were probably underestimated by approximately
125%. If the feed rates are increased by 125%, the mass-weighted
DRZ becomes 99.991.
a Indicates whether or not'the results of laboratory and field blank analy-
ses were subtracted from the results of sample analyses prior to calcula-
tion of DRE.
4-22
-------
TABLE 4.4
SITE B
Manufacturer:
Type:
Fuel:
Waste Stream:
Cleaver Brooks
Firetube
'-aS
Paint Manufacturing Waste
Design Steam Pace:
Design Steam Pressure:
Test Steam Rate:
Fraction Waste Fuel
Mass:
Heat Input:
8,400 lo/hr
150 psig
2,000 lo/hr
14%
Compound
Quantity Found (ug) Test
Feed Base- Aver- Test Emission Blank3
Rate line age Runs Rate Cor-
mg/sec Run Blan>. iavg) (ug/sec) rected
ORE
naphthalene
pentachlorophenol
toluene
0.0187 WASTE FEED RATES ARE TOO LOW TO
0.0065 ALLOW CALCULATION OF ACCURATE
4.670 DESTRUCTION AND REMOVAL EFFICIENCIES
NOTE:
m (Feed Rate) - (Test Emission Rate) x 1
Feed Rate
a Indicates whether or not the results of laboratory and field blank analy-
ses were subtracted from the results of sample analyses prior to calcula-
tion of ORE.
4-23
-------
a conclusion as to the ORE cf hazardous compounds to be made. T^e rate of
feed of the hazardous compounds was insufficient to contribute mea^uraole
stack gas concentrations of tne compounds of interest. The results re-af-
i^rm the need to assure that ?OHC feed rates are high e.-.^ugh ••hat the resi-
due (after 99.99% DRE) will exceed the limit of detection of the method to
be used for their measurement in the stack gas.
Site C
The boiler tested at Site C was a wall-fired steam generator with a
capacity of 230,000 pounds of superheated (250 psig) steam per hour. The
boiler has six burners each having a gas ring ana an oil gun. The oil guns
were used to inject liquid waste while either natural or waste process gas
is fired through the gas rings. The boiler operated well during the testa.
There were no upsets reported. The boiler was operated at approximately 25%
of its design capacity so the percent excess air was high during all test*.
The concentrations of CO, NOX/ and unburned hydrocarbons were low.
Only two compounds listed in Appendix VIII were found in the waste feed
and exhaust gas. The waste streaa contained approximately 5% phenol and ap-
proximately 0.004% bis(2-ethylhexyl)phthalate. The ERE of phenol (Table 4.5)
was in excess of 99.99%. The DRE of the phthalate was less than 96%. The
phthalate was detected in all four of the stack gas samples. Its measured
emission rate during tne baseline run was twice its average co-fired emission
rate, whereas the baseline emission rate of phenol was only 3% of its average
co-fired emission rate. It appears that all of the phthalate measurements
may have been due to contamination. Further evidence that the phthalate re-
sult was anomolous is provided by the non-Appendix VIII compound results.
The DRE of nine other compounds identified in the waste feed averaged 99.999%,
the same as the ORE of phenol. It is scarcely conceivable that one compound
was not destroyed while 10 others were.
The conclusion is that a wall-fired boiler, operating at 25% of its de-
sign capacity with 40% of the fuel heat provided by waste, destroyed more
than 99.99% of the hazardous organic material that was co-fired.
Site D
The boiler tested at Site 0 was a balanced draft, field erected water-
tube with a rated capacity of 90,000 pounds per hour of 260 psig saturated
steam. The boiler was operated at approximately 80% of full load with (on
the average) 40% of the heat provided by waste material and 60% by No. 6
fuel oil. The boiler has four B*w oil/gas burners. One of these was used
to inject the waste material during the co-fired runs. The plant produces
several different waste streams. Two were selected for testing based upon
their relatively high contents of chlorinated hydrocarbon*. One of these
streams was primarily (approximately 70%) methanol with 15% xylene and 5%
perchloroethylene (perc). The second was 90% toluene with 6-7% bis(2-
ch 1 oroethyl)ether (BCEE). Three co-fired test runs were done for each
•waste stream.
4-24
-------
TABLE 4.5
SITE C
Manufacturer: Baococ* t, Wilcox
Type: Wall Fired
fuel: Gas
Waste Stream: Ph-nol'.c Wastes
Design Steam Rate:
Design Steam Pressure:
Test Steair Rate:
Traction Waste Fuel
Mass:
Heat Input:
230, OOU in/hr
250 psiy
S'J.OOO Ib/hr
38%
Quantity Found (ug) Test
Feed Base- Aver- Test Emission blanX*
Rate line age Runs Rate Cor-
Cymppund mq/sec Run Bl&nic (avq) (ug/sec) rected ORE
phenol
bis(2-ethylhexyl)
13,300
27
49
93
No
NOTE:
(feed Rate) - (Test Emission Rate) x ^QO%
Feed Rate
99.999
phthtlate
mass-weighted ORE
11 270 C2 154 233 No 97.3b2
99.999
• Indicate* whether or not. the results oi laboratory and field blank analy-
ses were subtracted fro* the results of sample ana'.yt.es prior to calcula-
tion of ORE.
4-25
-------
Scene upset conditions occurred during the field tests. These were
mostly during the first co-fired run for ths first waste stream. These
upsets, which included smoke formation were caused by flame outs of the
waste fuel burner which in turn were caused by waste feed disruptions and
improper flame scanner settings. Testing was -suspended during most of the
upsets, though some testing did occur during upset conditions.
Perc and toluene were sampled by a variation of the VOST procedure.
The train used included two impingers, one upstream and one downstream of
the first sorbenc cartridge. Considerable difficulty was encountered dur-
ing the laboratory analysis of the Tenax*-GC samples. None of the impin-
gers were analyzed for toluene and only 15 of the 21 were analyzed for perc.
None of the seven impinger blanks were analyzed for either compound. These
data were rejected because of their failure to satisfy quality assurance
acceptance criteria #1. Both sorbent cartridges were analyzed for toluene
in 6 of the 21 pairs and in only three of these cases was more than 70% of
the sorbent-bound toluene found on the first cartirdge. Only one of seven
pairs cf so r bent cartridge blanks was analyzed for toluene. Both tub*s of
9 of the 21 pairs of cartridges were analyzed for perc. Twelve pair did
not satisfy quality assurance acceptance criteria #2. More than 70% of the
sorbent-bound perc was on the first tube in only two of these nine pairs.
Thus, only two runs satisifed all three quality assurance acceptance cri-
teria. Only four of the seven sorbent blanks were analyzed. The amounts
of these two compounds detected during runs when they were a fuel consti-
tuent was not significantly different from the amounts detected during runs
when they were not.
All of this is strong evidence that the perc and toluene detected in the
sampling train was the result of contamination. Even so, the calculated ORE
for both compounds exceeded 99.99%. However, not enough data satisfied the
quality assurance acceptance criteria to provide confidence in these numbers.
Therefore, for the purpose of this document only the BCEE results (obtained
by MM5) will be considered. These results (Table 4.6) show that the ORE of
BCEF was much greater than 99.99%.
It is concluded that this boiler destroyed mere than 99.99% of the haz-
ardous compounds that, were co-fired. Test runs 2, 3, and 6 were chose during
which most of the upset conditions occurred. Sampling during upset conditions
was most prevalent during run 2. No effect on the emission rates or CREs of
either hazardous or other con pounds is evident in any of the data. These up-
sets apparently did not interfere with the destruction of the hazardous mate-
rials.
Site E
The boiler tested at Site E was a forced draft packaged water-tube with
a design capacity of 110,000 pounds of superheated steam per houx. The design
steaa delivery pressure w«js 425 psig. The boiler is equipped wi th a dual air
register COEN burner that has a gas ring and a No. 6 oil gun. This burner
had been modified by the addition of two waste fuel guns. These were located
at opposite ends of a diameter of the burner approximately midway between the
oil gun and the gas ring.
4-26
-------
TABLE 4.6
SITE D
Manufacturer:
Type:
Fuel:
Waste Stream:
Babcock & Wilcox
Water Wall (field erected)
Oil
Methanol and Toluene
Wastes
Design Steam Rate:
Design Steam Pressure:
Test Steam Rate:
Fraction Waste Fuel
Mass:
Heat Input:
90,000 Ib/hr
260 psig
70,000 Ib/hr
40%
37%
Compound
Quantity Found (ug) Test
Feed Base- Aver- Test Emission Blank3
Rate line age Runs Rate Cor-
mg/sec Run Blank (avg) (ug/sec) rected
ORE
bis (2-chloro-
ethyl) ether
7,600
ND
NO
3.7
<4.7
No
>99.9999
NOTE:
DRE
(Feed Rate) - (Test Emission jtate)
Feed Rate
x 100%
Indicates whether or not the results of laboratory and field blank analy-
ses were subtracted from the results of sample analyses prior to calcula-
tion of DRE.
4-27
-------
Two different waste streams were co-fired during the tests. One (TSB
wa.ste) consisted largely (approximately 80%) of fluxing oils and contained
approximately 1% methyl methacrylate (MMA) and 7% and 11% of x-hydroxy-
methylisobutyrate (MOB) and x-hydroxy-methylisobutyrate methyl ether (MEMOB),
respectively. The second (toluene waste) waste stream was approximately 80%
toluene and 20% methyl methacrylate. A third waste stream (Cl-TSB) was pre-
pared by adding approximately 2% each of carbon tetrachlori'ie, chlorooenzene,
and trichloroethylene to the TSB waste.
Nine test runs were done, they wsre:
o 1 baseline run - No. 6 oil only
o 1 co-firad, No. 6 oil plus TSB waste
o 5 co-fired, No. 6 oil plus Cl-TSB waste
o 1 co-fired, natural gas plus Cl-TSB waste
o 1 co-fired, natural gas plus toluene waste
Of the five co-fired (No. 6 plus Cl-TSB) runs, three were at 50% design steam
load, one was at 37% load, and one was at 73% load. One of the gas-fired runs
was at 50% load the other was at 40% load.
MMA, MOB, and MEMCD were all sampled and analyzed by MM5, the chlori-
nated hydrocarbons were sampled and analyzed by a modified VOST procedure.
The modification consisted of placement of an empty impinger (in an ice bath)
upstream of the first Tenax®-GC sorbent tube.
The Tenax*-GC sampling and analysis for the volatile hydrocarbons was
only marginally successful. Of the total of 27 pairs of tubes exposed, the
quality assurance acceptance criteria (QAAC) were satisfied for: carbon te-
trachloride - 6 pairs, trichloroethylene - 3 pairs, chlorobenzene - 5 pairs,
and MMA - 13 pairs. The number of analyses rejected by the various QAAC are
shown below:
Number of Samples Rejected for QAAC Failures
QAAC #1 QAAC #2 QAAC #3
(impinger (analyze (70% on Accepted
Compound analysis) both tubes) first tube) Samples
CC14 0 5 16 6
TCE 0 5 19 3
chloro-
benzene 0 5 17 5
MMA 05 9 13
Based on the data that satisfied the QAAC, the average DREs for these
compounds for all tests were:
Test Average ORE
Compound QAAC Data ORE All Data
carbon tetrachloride 99.9988 99.9996
trichloroethylene 99.9969 99.9986
chlorobenzene 99.9957 99.9981
methyl methacrylate 99.9587 99.9910
4-28
-------
The third column, ORE calculations based upon all data, is similar to the
second. Even though most of the measurements appear to be the result of
contamination rather than compounds sampled fror. the flue gas, the DREs are
all in excess of 99.99%. Such a conclusion cannot be unequivacaisle because
there is a second possible explanation for the observation o:: large quanti-
ties of the compounds on the back-up sorbent tube, i.e., breakthrough. If
the sampling train conditions were such that the first sorbent tube did not
collect compounds efficiently, then finding large quantities of material on
the back-up tube is expected. Determination of how much material may have
broken through the back-up sorbent tube cannot be done quantitatively. It
is unlikely that breakthrough caused hig^ loadings on the back-up tube, but
the possibility cannot be totally discounted. Therefore, the data can be
accepted only with qualification. Only the DRE for methyl methacrylate is
listed in Table 4.7 because it was the only compound for which there were
enough successful test runs to afford confidence to a conclusion. The ORE
of this compound appears to have been almost exactly 99.99%.
The two non-Appendix VIII compounds (MOB and MEMO3) that were present
in the TSB waste in significant quantities were sampled by MM5 during these
tests. The XAD-resin was maintained at 60°C during the sampling. At this
resin temperature and the total volumes of gas sampled, the resin would have
retained only 1/3 to 1/2 of these compounds. If the measured emission rates
of these compounds are tripled the poorest calculated DRE becomes 99.996%.
The data from Site E are not sufficiently sound to allow definitive
corrections between DRE and operating conditions to bo made.
Site F
The boiler tested at Site F was a balanced draft watar-tube with a rated
capacity of 60,000 pounds per hour of 200 psig steam. The boiler was origi-
nally constructed to burn coal, but had been converted by the addition of two
BSW circular burners to the burning of either oil (No. 6 or No. 2) or gas
fuels. Waste solvent is injected through a separate y-jet gun near the cen-
ter of the lower burner. The waste solvent normally cc—fired in this boiler
consists of paint thinner that has been used to purge paint spray guns. The
waste was spiked with trichloroethylene, carbon tetrachloride, and chloroben-
zene for the purposes of these tests. Waste thinner was 12% of the mass of
fuel fed and provided 9% of the heat input.
The boJler had been out of service for repair of refractory before the
test program. It had been operated solely on gas fuel since the repair at
the time the tests began. Apparently, the lower oil gun had not been rein-
stalled properly after the repair. It became encrusted with coked fuel to
the extent that is caused the boiler to shut down after the third test arid
again near the end of the fourth test. The burner misalignment had no ob-
servable effect upon the concentrations of any of the continuously monitored
gases. Neither the carbon monoxide nor the total unburned hydrocarbon in-
struments were on-line during Test 4, so no data for these two gases at the
time of failure are available. There were no significant changes in the con-
centrations of any of the combustion gases from test to test. The fraction
of the particle emissions that was attributable to fuel ash dropped somewhat
during Test 4. Between 92 and 106% of the stack gas solids were accounted
4-29
-------
TABLE 4.7
SITE E
Manufacturer: Combustion Engineering
Type:
Fuel:
Waste Stream:
Water Wall (package)
Oil
Methyl methacrylate
manufacturing waste
Design Steam Rate: 110,000 Ib/hr
Design Steam Pres.: 425 psig
Test Steam Rate: 55,000 Ib/hr
Fraction Waste Fuel
Mass: 28-75%
Heat Input: 21-52%
Quantity Found (ug) Test
Feed Base- Aver- Test Emission Blank3
Rate line age Runs Rate Cor-
Coropound mg/sec Run Blank (avg) (ug/sec) rected DRE
me thy line thacry late
12,210 1.5 0.287 2.920
1378
Yes
99.989
NOTE: 1) DRE = (Feed Rate) - (Test Emission Rate) x 100%
Feed Rate
2) Low ORE may be the result of sample contamination; see text.
a Indicates whether or not the results of laboratory and field blank analy-
ses were subtracted from the results of sample analyses prior to calcula-
tion of ORE.
4-30
-------
for by fuel ash during the baseline and the first two co-fired runs. This
fraction dropped to 80% during the final co-fired run (Test 4). This implies
that the poor burner alignment was producing soot particles during Test 4 and
was probably causing poor combustion efficiency.
Stack gas samples were taken by both VOST and MM5. Butylbenzyl phthalate
was the only semi-volatile organic compound listed in Appendix VIII that was
found in the waste stream. Its; concentration in the thinner was approximately
2.5%. Irs mass flow rate was approximately 0.85 g.n/sec. It was not detected
in any of the stack gas samples. Therefore, its ORE was in excess of 99.999%.
The only other MM5 compound detected in both the thinner and any stack gas
sample was butane dioic acid dinethyl es*.er. It was detected in the stack
gas only during Test 3. Its DRE during that test was greater than 99.999%.
Several other Appendix VIII compounds were detected in some stack gas samples.
Of these only bis(2-ethylhexyl)phthalate was detected at levels significantly
greater than its minimum detectable limits. The amount of this compound de-
tected in the blank exceeded the average amount detected in the co-fired run
samples so its detection is judged to be due to contamination.
The chlorinated hydrocarbons (carbon tetrachloride, trichloroethylene,
and chlorobenzene) that were added to the thinner were sampled by the VOST
procedure. Three pairs of VOST cartridges were used during each of the first
two and two pairs during the third co-fired test runs. Of these eight pairs
the QAAC were met by three carbon tetrachloride samples, three trichloroethy-
lene samples, and three chlorobenzene samples. Two runs were rejected because
of failure to analyze both tubes (QAAC #2). One analysis was lost because
of failure of the GC/MS. Two analyses were rejected because the amounts of
compounds found on the first tube was less than 70% of the total (QAAC #3).
The DRE of only chlorobenzene appears to have been less than 99.99%. Its
concentration in the thinner was low, only 20% of the planned concentration,
however. Because the feed rate was so low the CHE calculation is subject
to undue influence by small amounts of sample contamination. The DREs of
the other two compounds (Table 4.8), both of which are more refractory than
chlorobenzene, were greater than 99.99% as was the mass weighted average DRE.
Most (5 of 9) of the accepted data were from Test 4, the final co-fired
run. It was during this run that the boiler was shut down by coking of the
lower oil burner. The mass weighted DRE of all the chlorinated hydrocarbons
during this test was 99.997%. It is difficult to compare this run to the
other two co-fired runs since the data recovery from them was poor. The
final co-fired run does demonstrate acceptable DRE under less than ideal
operating conditions, however.
Site G
The boiler at Site G was a three-pass wetback scotch marine packaged
fire-tube with design capacity of 26 million Btu/hr heat input. This boiler
was originally rated at 40 million Btu/hr heat input but was derated in con-
junction with its modification to a hazardous waste combustor. The modifica-
tions included a change from positive to negative pressure in the firebox and
changes in internal baffle configurations needed to maintain surface temper-
atures that retard acid gas attack. The boiler is fitted with a two-stage
4-31
-------
TABLE 4.8
SITE F
Manufacturer: Babcock * Wilcox Design Steam Rate: 60,000 Ib/hr
Type: Water Wall Design Steam Pressure: 250 psig
?uul: Oil Test Steam Rate: 50,000 Ib/hr
Waste Stream: Paint Solvents Fraction Waste Fuel
Mass: 12%
Heat Input: 9%
Compound
trichloroethylene
chlorobenzene
carbon tetrachlorida
mass-weighted ORE
NOTE: DRE
Quantity Found (ug) Test
Feed Base- Aver- Test Emission 31anka
Rate line a^e Runs Rate Cor-
mg/sec Run Blank (avg) (ug/sec) rected DRE
1,100 0.398 0.041 0.104 25.7 Yer 99.998
109 0.029 0.010 0.060 15.3 Yes 99.986
806 0.116 0.107 0.264 70.3 Yes . 99.991
IT
;>9.992
„ (Feed Rate) - (Test Emission Rate) x IQO%
Feed Rate \
a Indicates whether or not the results of laboratory and field blank analy-
ses were subtracted from the results of sample analyses prior to calcula-
tion Of DRE.
4-32
-------
scrubber system to remove and recover halogen acids from the exhaust gases.
The boiler was equipped to burn neural gas and either liquid or gaseous
wastes. It would operate on waste halogenated hydrocarbons only, if the
heat value of the waste exceeds 9,500 Btu/lb.
The waste being fired during the tests contained 40% bis{2-chloroiso-
propyDe'cher (BCPE), 30% propylene dichloride, and 17% epichlorohydrin. Four
percent carbon tetrachloride (CC14) was added to the waste for the purpose of
the tests. The waste was approximately 43% chlorine. Its average heating
value was 8,990 Btu/lb, slightly less than the heat value specified by the
manufacturer for waste only firing. In spite of the nature of the fuel, the
boiler operated with out incident at 82-83% thermal efficiency while burning
100% waste throughout the tests.
Carbon monoxide (CO) and total unburned hydrocarbon (TUHC) concentra-
tions decreased slightly during the tests. The average concentration of CO
decreased from 170 ppm (all concentrations correctcxd to 3% O2) during Test
1 to 146 ppm during Test 3; TUHC decreased from 0.7 ppm to 0.3 ppm. The
changes, though small, were attributed to better atomization of the w«ste
during the later rur.s.
One Appendix VIII compound (BCIE) was sampled and analyzed by MM5. It
was detected in one of the three samples at an emission rate of 1.2 micro-
grams per second.
Three Appendix VIII compounds were sampled and analyzed by VOST proce-
dure. They were: CCl^, epichlorohydrin, and trans-1,3,dichloropropene
(T-DCP). Epichlorohydrin was not detected in any stack gas sample. T-DCP
was detected, in trace amounts, in two of the eight pairs of VOST tubes an-
alyzed. CC14 was detected in all eight pairs of VOST tubes but three pairs
did not satisfy QAAC #3 (less than 70% on first tube). Its emission rate
averaged 259 micro grains per second, over half of which was observed during
the first part of the second test. No measured aspect: of the boiler opera-
tion account? for this one high emission rate result.
All compounds were destroyed with greater than 99.99% efficiency (Table
4.9) during all test runs. This was a special purpose boiler. The results
can be extrapolated to similar units designed for the purpose of destruction
of hazardous materials.
Site H
The boiler tested at Site H was a pulverized coal-fired boiler built in
1975, with a rated steam capacity of 250,000 Ib/hr of superheated steam at 600
psig and 740°F. This tangentially fired boiler was equipped with three le-
vels of pulverized coal burners in each furnace corner. The three levels
were separated by two levels of steam-atomized oil burners. Generally, or-
ganic wastes were injected into the furnace by means of one or more of these
oil burners. Typical steam loads are 250,000 Ib/hr which is the rated steam
capacity of the boiler. Primary and secondary combustion air is preheated
by means of a regenerative air prcheater. The unit is equipped with a cold
side ESP for fly ash control. Fly ash is collected and removed from the hop-
per continuously via pueumatic system.
4-33
-------
TABLE 4.9
SITE G
Manufacturer:
Type:
Fuel:
Wa.ste Stream:
THROX
Fire Tube
None
Chlorinated Solvents
Design Steam Rate:
Design Steam Pressure:
Test Steam Rate:
Fraction Waste Fuel
Mass:
Heat Input:
26,000 Ib/hr
250 psig
15,000 lo/hr
100%
100%
Compound
Quantity Found (ug) Test
Feed Base- Aver- Test Emission Blank3
Rate line age Runs Rate Cor-
mg/sec Run Blank (avg) (ug/sec) rected
ORE
carbon tetrachloride 9,800 NR
trans-1,3-dichloro-
propene
epichlorhydrin
bis(2-chloroiso-
propyDether
mas s - we i gh ted
average ORE
31,600 NR
40,400 NR
1 07,000 NR
0.0143 2.58 259
ND 0.003 <2.0
<1 <1 <2
<1
<2.0
Yes
Yes
Yes
No
99.997
>99.9999
99.9999
>99.9999
99.9997
NR - No baseline run at this site, the boiler burned only hazardous waste.
NOTE: DRE , (Feed Rate) - (Test Emission Rate) x ^00%
Feed Rate
a Indicates whether or not the results of laboratory and field blank analyses
were subtracted from the results of sample analyses prior to calculation of
ORE.
4-34
-------
Several organic waste streams are produced and incinerated in the boil-
ers at this site. Firing rates of these waste streams are generally between
3 and 7 gpm when boiler loaas exceed 150,000 lo/hr. The waste stream of in-
terest was crude methyl acetate available in a 1,500-gal tank. The methyl
acetate was artificially spiked with chlorinated organic compounds, namely
CC14, chlorobenzene (C10), and 1,1,1-trichloroethane.
Four tests were performed: one baseline and three co-fire. All tests
were performed with the boiler set on manual control. A heat output of
246,000 Ib/hr of steam was maintained. No boiler upsets or transients were
recorded. For Test 1, the boiler was operated with pulverized coal (gene-
rally eastern Kentucky or south riest Virginia bituminous only). Furnace
excess air and soot blowing cycles were typical for this firing condition.
Tests 2 through 4 were the co-firing tests. The chlorinated methyl acetate
firing rate varied from 2.4 to 4.2 gpm. One steam-atomized oil burner was
used to input the waste which accounted for 2 to 4% of the total heat input
to the boiler.
The minimum concentration of any POHC in the waste fuel was approximate-
ly 2%. At this concentration and at the rate of waste feed cited the mini-
mum rate of flow of any POHC was approximately 2.5 g/sec. At 99.99% ORE the
flow rate of this POHC in the flue gas would have been approximately 250 ug/
s, which is easily detectable by the method of sampling and analysis used.
No test report is available for this site so it is not possible to com-
pare the methods and data to the £A acceptance criteria that were developed
earlier in this section. The data in Table 4.10 were abstracted from <* re-
port that summarized the results from all sites. The data available indi-
cate that the boiler achieved 99.99% ORE of hazardous materials.
Site I
The boiler tested at Site I was a forced draft bent tube capable of de-
livering 62,000 pounds of 175 psig steam per hour. The boiler was originally
designed for either oil or gas fuels but had b«en modified to burn waste gas
(largely methane) in combination with small amounts of organic liquid waste.
The boiler was equipped with two gas ring burners. Steam atomizing liquid
waste guns could be inserted into the center of both. The burners are ar-
ranged in a vertical plane. Most of the organic liquid waste burned by the
plant (and the only liquid waste burned during these testa) is a high nitro-
gen aniline production waste. The plant normally operates the boiler in a
svvged combustion mode while burning the high nitrogen waste in order to re-
duce formation of nitrogen oxides from fuel-bound nitrogen. Staging is ac-
complished by firing the high nitrogen waste through the lower burner and
reducing the flow of combustion air through the burner port.
Approximately 1\ each of carbon tetrachloride (CC14) and trichloro-
ethylene (TCE), and 3.5% chlorobenzene (MCB) and toluene were added to the
nitrogenous waste during the tests. The fuel, as fired, contained approxi-
mately 83% nitrobenzene (NB) and 2% each of anilane (Ab) and benzene. The
organic liquid waste contributed approximately 17% of the fuel mass and
approximately 8% of the fuel-heat input. The primary fuel burned during
the te.ts was natural gas.
4-35
-------
TABLE 4.10
SITE H
Manufacturer:
Type:
Fuel:
Waste Stream:
Comoustion Engineering
V'J-40
Pulverized Coal
Methyl Acetate Waste
Design Steam Rate:
Design Steam Pressure:
Test Steam Rate:
Fraction Waste Fuel
Mass:
Heat Input:
250,000 lo/nr
600 psig
246,000 lo/nr
6*
3%
Compound
Feed
Rate
my/seca
y/uantity Found (ug) Test
Base- Aver- Test Emission Blank''
line age Runs Rate Cor-
Runa Blank* (avg)a (ug/sec)a recteda ORE
carbon tetrachloride
chlorobenzene
methyl chloroform
mass-weighted
aveiag* ORE
99.98
99.992
99.994
99.991
a
b
Data not available
Indicates whether or not the results of laboratory and field blank analyses
were subtracted from the results of sample analyses prior to calculation of
ORE.
4-36
-------
Both MM5 and VOST samples were taken during the tests. The procedure
used for each POHC is summarized below:
Mrt5 VOST
chlorobenzene carbon tetrachloride
aniline benzene
nitrobenzene trichloroethylene
toluene
Toluene would normally be sampled by MM5 since it boils at 1~1°C. However,
since the VOST results for toluene were higher than the MM5 results they were
used to calculate DREs for toluene. Both benzene and chlorobenzene (MCB) were
also tested by both methods. The two methods gave comparable results for MCB.
Benzene is too volatile to remain in the sample during the MM5 solvent evapo-
ration. The test organization concluded that benzene was either a contaminant
in or a product of decomposition of the Tenax*. Therefore, the benzene re-
sults were deemed unreliable and are not reported in Table 4.11.
All MM5 data met both quality QAAC that were developed for them.
The test method used for the VOCs was VOST so OAAC No. 1 (impinger anal-
ysis) does not apply. All VOST samples satisfied QAAC No. 2 (both tubes ana-
lyzed). All the VOST samples «re analyzed in pairs, however - no individual
analyses of the Tenax* or Tenax*/charcoal tubes were done. Therefore, QAAC
No. 3 (70% on first tube) cannot be applied.
Combustion staging effectively reduced production of NOy from fuel-bound
nitrogen. Whereas 72% of the fuel bound nitrogen was oxidized to NOX during
the six unstaged combustion runs, only 22% was oxidized to NOX during the
staged combustion runs.
There was no difference in the exhaust gas concentrations or the DREs of
any of the compounds between the staged and unstaged combustion runs. Of the
compounds measured, only nitrobenzer.vi and CCl^ were emitted at a signific~>nt-
ly (95% CD higher rates during the co-fired tests than during the baselii.a
tests. Therefore, the ORE data have been combined. The averages shown in
Table 4.11 are for alJ six co-fired runs. The ORE of all compounds tested
exceeded 99.99% for all co-fired test runs.
It is apparent that the combustion staging practiced at this site was
effective at reducing the conversion of fuel-bound nitrogen to NOX but that
it had no adverse effect on the ORE of the co-fired hazardous compounds.
Site J
The boiler tested at Site J was a packaged, fire-tube capable of deliver-
ing 8,400 pounds of 150 psig steam per hour. It is designed to burn either
oil or natural gas but normally burns the latter. The fuel burned during these
tests was blended from nitration-grade toluene and technical grade chlorinated
hydrocarbons. The fuel contained approximately 1% each of carbon tetrachloride
and trichloroethylene and 0.5% chlorobenzene; the balance (97.5%) was toluene.
4-37
-------
TABLE 4.11
SITE I
Manufacturer: Foster
Wheeler
Design Steam Rate:
Type: Bent tube (staged
combustion )
Fuel: Oil or
Waste Stream: Aniline
Gas
Wastes
6d,OUU lo/nr
Design oteam Pressure: 250 psig
Test Steam
Rate :
4U,uuO lo/nr
Fraction Waste Fuel
Mass:
Heat Input:
Quantity Found (ug)
Compound
carbon tetrachloride
trichloroethylene
chlorobenzene
toluene
nitrobenzene
aniline
mass -weigh ted
average ORE
NOTE: DRE
Feed
Rate
mg/sec
797
797
1,457
1,571
37,980
1 ,070
« (Feed
Base-
line
Run
0.061
0.019
0.209
0.190
21 .1
10.6
Rate) -
Aver- Test
age Runs
Blank (avg)
0.028 0.133
0.043 0.026
0.0423 0.183
0.174 0.199
64.9
12.93
(Test Emission
Test
Emission
Rate
'ug/sec )
24
8
29
16
360
21
Rate) * !
17%
8%
Blanxa
Cor-
rected DRE
Yes 99.997
No 99.999
No 99.998
No 99.999
Yes 99.9998
Yes 99.998
99.9989
I 00%
Feed Rate
a Indicates whether or not the results of laboratory and field blank analyses
were subtracted from the results of sample analyses prior to calculation of
ORE.
4-38
-------
The operation of tne boiler was varied dunn-j the tests in an attempt to
discover what effects boiler load and excess air have on tne DkE of hazardous
compounds. The operating conditions are summarized below:
Test Nunioer Condition Steam Flew Excess Air
1 1 Half L-^ad Norifcai
2 2 Pull Load Normal
3 1 Repeat of Test 1
4 3 Full Load High
5 4 Half Load High
6 5 Full Load Low
Two of the compounds, chlorobenzene , and toluene were measured by MM5.
All of the toluene results satisfied both gAAC. The feed rate of chloro-
benzene was insufficient to provide measurable concentrations in the stack
gas at 99.99% ORE. Most (11 of 17) -chlorobenzene results were less than
detectable and the calculated DRE was 99.96%. The maximum calculable ORE,
based upon the limit of detection and the feed rate, was 99.98%. It was
judged that the data could neither confirm nor deny 99.99% DRE of chloro-
benzene so it was not included in this analysis.
Carbon tetrachloride and tnchloroethylene were sampled by VOST so
No. 1 (impinger analysis) does not apply. Six of the 55 pairs of VOST tubes
sampled were invalidated because one or the other tube was broken (QAAC No.
2) and therefore not analyzed. An additional 19 pairs were lost because of
problems in the GC/MS. Six pairs of tubes were analyzed individually in order
to indicate the fraction of the compounds collected on the first versus the
backup tube. Unfortunately, three of these analyses were lost, either by GC/
MS problems or broken tubes. The average fraction of the compounds of inter-
est found on the first tube cf the three remaining pair was 85%. Of the six
analyses (three for CCl4 and three for TCE) all showed more than 80% on the
first tube but one. The first tube contained only 61% of the total collected
TCE in one pair of tubes. Though the amount of data is small it was judged
that QAAC No. 3 was satisfied by the data.
The DRE (Table 4.12) of every compound during every run exceeded 93.99%.
The DREs of hazardous compounds at 50% load were no different from the DREs
at full load. No change in the DRE was found when the excess combustion air
was changed from a low of 17% to a high of 50%. The DREs measured at these
extreme values of excess air were the same as those measured at the normal
excess air of approximately 35%.
It is concluded that small fire-tube boilers can maintain adequate DRE
of hazardous compounds while operating between 50-100% of full load. It is
further concluded that the changes in excess air that occur when the load is
changed do not reduce the DRE of hazardous compounds.
Site K
The boiler tesved at Site K was a water-tube boiler that was designed to
deliver 60,000 pounds of 125 psig steam per hour. It is equipped with four
burners each of which can be used to fire either light or heavy oil by con-
necting the appropriate burner tip. Heavy oil is steam-atomized, light oil
4-39
-------
TABLE 4.12
SITE J
Manufacturer: North American
Type: I-'ire Tube
Fuel: None
Waste Stream: Blended (98% toluene)
Design Steam Rate:
6,000 lo/hr
Design Steam Pressure: 150 psiy
Test Steam Rate:
Fraction Waste Fuel
Mass:
Heat Input:
full load
100%
Compound
Quantity Found (ug) Test
Feed Base- Aver- Test Emission Blank3
Rate line age Runs Rate Cor-
ing/sec Run Blank (avg) (ug/sec) rected
ORE
carbon tetrachlonde J60
trichloroethylene 494
toluene 44,050
mass-weighted
average DRE
0.036 0.120 17.2 Yes
0.010 0.061 4.3 Yes
140 877 1,514 No
99.995
99.9991
99.997
99.997
NOTE:
DRE • (P*«d Rate) - (Test Emission Rate) x ^QQ%
Feed Rate
a Indicates whether or not the results of laboratory and field blank anal-
ys -s were subtracted from the results of sample analyses prior to calcu-
lation of DRE.
4-40
-------
is air-atomized. The boiler typically burns off-specification fuels and
waste solvents. The burners are located in a horizontal row along one side
of the boiler. During the three baseline tests, standard No. 6 oil was fired
through ail four ourners. During the three cc-fired tests a mixture of 70%
Nc. 6/30% waste solvent was fired through the two outside burners and a light
oil (a waste solvent mixture) spiked with chlorinated hydrocarbons was fired
through the two inside burners. The light oil contained approximately 0.5%
each of carbon tetrachloride, tuchloroethylene, and chlorobenzene. The
waste oils also contained approximately 5% xylenes and approximately 1% ben-
zene. Xylenes were not considered because they are not listed in Appendix
VIII; benzene was not considered because its concentration in the waste feed
is low (QAAC No. 2).
Only phenol was determined by MM5. Both QAAC were satisfied by all the
phenol data. The VOST method was used co measure the concentrations of car-
bon tetrachloride, trichloroethylene, toluene, and chlorobenzene. QAAC No.
1 (impinger analysis) is not applicable. Five of the 24 pairs of VOST tube
analyses were not accepted because of failure to analyze either the first
tub* or the back-up tube (QAAC Ho. 2). The laboratory attempted analysis of
individual tubes for four pairs of VOST tubes. Two of these attempts failed
because of loss of one or the other of the analyses. The average amount of
the various POHCs found on the first tube of the two pairs for which separate
analysis was successful was 93%. These analyses were taken as evidence that
QAAC No. 3 (70% on first tube) was satisfied oy all VOST analyses.
The boiler was operated at 80% capacity during these tests. The percent
excess air was invariant at approximately 2?%. Individual carbon monoxide
concentrations ranged from 65-300 ppm. Test long averages ranged from 87-150
ppm. The waste material contributed 57% of the fuel mass and 49% of the heat
input. Combustion efficiency (calculated from CO2 and CO concentrations) av-
eraged 99.9%. The ORE of all compounds (Table 4.13) exceeded 99.99% during
all co-fired test runs. The lowest ORE measured (99.998%) was measured for
carbon tetrachloride during Run 6.
Site L
The testing at Site L was undertaken as an attempt to define the range
of boiler operating conditions over which adequate ORE of hazardous compounds
can be attained. The boiler was controlled at the request of the test team
to provide ORE tests under conditions of: low steam load, low 02, high 02,
normal CO, high CO, rapid steam load changes, high fuel substitution rate,
various waste atomizer steam pressures, different waste atomizer orientations,
with No. 6 oil primary fuel and with natural gas primary fuel. Stack gas and
boiler operating parameters were measured in an attempt to find relationships
between them and poor ORE of hazardous compounds.
The boiler tested at Site L (Site L testing was done on the Site E boil-
er) was a forced draft packaged water-tube with a design capacity of 110,000
pounds of superheated steam per hour. The design steam delivery pressure
was 425 psig. The boiler was equipped with a dual air register COEN burner
that had a gas ring and a No. 6 oil gun. This burner hacf been modified by
the addition of two waste-fuel guns. These were located at opposite ends of
a diameter of the burner, approximately midway between the oil gun and the
4-41
-------
gas ring. The waste-fuel is normally atomized in a fan snaped pattern that
impinges upon the oil flame but they can be rotated so that the waste fuel
is sprayed away from the flame toward the boiler walls. The waste fuel was
methyl methacrylate wastes spiked with carbon tetracr.loride and chloroben-
zene. During these tests the waste material supplied between 12 and 56% of
the total heat input to the boiler. The ranges of conditions under which
the boiler was operated during these tests, are compared to good engineering
practice (Reference 22) below.
Range of Flue Gas Compositions for
Gas-
and Oil-Fired Package Water-Tube Boilers
Gas-Fired
Good
Flue Gas
Component
02 (%)
NO* (ppm)
C02 (%/
CO (ppm)
TUHC (ppm)
Engineering
Practice
2.5
70
9.0
145
25
to
to
to
to
to
5.5
100
10.0
170
45
•^ese Tests
2.2 to 5.3
a
a
90 to 1970
a
Oil-Fired
Good
Engineering
Practice
3.0
65
7.5
45
3
to
to
to
to
to
7.6
470
13.5
55
35
These
3.5 to
a
a
90 to
a
Tests
6.3
370
a Data not available.
The data for the hazardous compound DRE runs, Table 4.14, demonstrate
that with few exceptions the boiler destroyed more than 99.99% of both of the
hazardous components. The DRE of carbon tetrachloride (CCl,j) was less than
99.99% in 3 of 44 runs; the DRE of chlorobenzene (ClPhi) in only 2. These 5
runs are summarized below.
Dale
May 8
May 8
May 8
May 9
May 10
Site L
Summary of Tests Showing DRE Less than 99.99%
DRE
Time
1447
1660
2232
1311
1030
CCla
99.7
99.9996
99.998
99.987
99.989
ClPhi
100.0
99.98
99.97
99.995
99.993
CO (ppm)
92
345
120
325
128
6.33
Comment
Low waste flow
No waste atomization
Soot blowing
Low waste flow, O2
Waste startup
Two of these are related to waste feed disruptions (one startup and one test
with no atomization). Two occurrences of low DP.E occurred during periods of
low waste feed rate. These appeared to be caused by the relative instability
of the waste feed rate control system at low waste flow rates. Although only
two DRF.s at low waste flow were less than 99.99%, all DREs measured when the
waste flow was less than 1.5 gallons per minute were relatively low.
4-42
-------
TABLE 4.13
SITE K
Manufacturer: North American Design Stea-n Rate: 60,000 Ib/hr
Type: Fire Tube Design Ste^m Pressure: 125 psig
Fuel: No. 6 Oil Test Stean Rate: 49,000 Ib/hr
Waste Stream: Alcohol, Waste Solvents Fraction \.aste Fuel
Mass: 51%
Heat Input: 49%
Quantity Found (ug)
Compound
carbon tetrachloride
trichloroethylene
toluene
chlorobenzene
phencl
mass-weighted
average ORE
Feed
Rate
mg/sec
2,090
1 ,800
8,290
1 ,945
49,760
Base-
line
Run
0.022
0.008
0.081
0.013
None
Aver-
age
Hlank
0.065
0.009
0.089
0.024
Test
Runs
( avg)
0.081
0.044
0.143
0.025
5
Test
Emission
Rate
(ug/sec)
18.0
9.0
14.6
1 .4
<370
Blanka
Cor-
rected
No
No
Y«s
Yes
No
ORE
99.999
99.999
99.999
99.999
>99.999
99.999
a Indicates whether or not the results of laboratory and field blank anal-
yses were subtracted from th« results of sample analyses prior to calcu-
lation of ORE.
4-43
-------
TABLE 4.14
ORE GROUPED BY TEST CONDITIONS
SITE L
DATE TIME GC I FUEL LOAD
25-Apr
27-Apr
10-May
10-May
10-May
10-May
10-May
10-May
09-May
08-May
27-Apr
25- Apr
OS-May
25-Apr
27-Apr
OS-May
12-May
11 -May
11 -May
27-Apr
25-Apr
11 -May
10-M«y
OS-May
11 -May
11 -May
08-May
09-May
09-May
10-May
10-May
08-May
11 -May
10-May
10-May
i.O-May
09-May
09-May
10-May
09-May
09-May
09-May
09-May
10-May
09-May
09-May
09-May
OS-May
09-May
OS-May
08-May
11 -May
10-May
OS-May
09-May
1408
945
716
1647
1750
1331
1410
1740
1626
1732
1430
1606
1755
1154
9
1528
?ii9
1446
1445
7
C313
1600
2225
2033
2224
20
2108
1030
1917
2108
1430
2134
2012
2055
1503
2301
1452
2153
921
1702
1946
1132
1301
1311
1142
1710
1038
1642
1447
2118
2222
2332
2346
MM5
MrlS
FULL
FULL
FULL
FULL
FULL
FULL
FULL
FULL
FULL
96
90
87
101
103
104
57
48
VOST
VOST
36
VOST
VOST
35
VOST
144
VOST
VOST
VOST
122
118
40
VOST
1 TRAP VO
FULL
FULL
FULL
VOST
53
77
92
105
VOST
1«3
109
107
108
69
83
97
81
57
VOST
76
94
74
66
62
46
60
42
38
1 TRAP VO
111
50
36
GAS
GAS
OIL
OIL
OIL
OIL
OIL
OIL
GAS
GAS
GAS
GAS
GAS
GAS
GAS
GAS
OIL
OIL
OIL
GAS
GAS
OIL
OIL
GAS
OIL
OIL
GAS
GAS
GAS
OIL
OIL
GAS
OIL
OIL
OIL
OIL
GAS
GAS
OIL
GAS
GAS
GAS
GAS
OIL
GAS
GAS
GAS
GAS
GAS
GAS
GAS
OIL
OIL
GAS
GAS
59000
59000
27000
28000
32000
44000
46000
50000
52000
54000
57000
57500
58000
58000
60000
60000
60000
60000
61000
62000
64000
50000
50000
57000
60000
60000
61000
73000
25000
23000
47000
60000
62000
47000
47000
50000
25000
35000
40000
55000
63000
25000
27000
27000
28000
51000
53000
55000
57000
59000
60000!
60000
46000
62000
32000
CONDITIONS
TS8 BASELINE
TSB BASELINE
LOW LOAD BASELINE
LOU LOAD BASELINE
LOW LOAD BASELINE
BASELINE OIL. MODERATE
HIGH GPM/MOOERATE LOAD
MODERATE 02/HIGH GPM
GAS BASELINE
LOW 02 BASELINE
BASELINE
TSB BASELINE
TSB HIGH GPM BASE
TSB BASELINE
BASELINE
TSB HIGH GPM BASE
HIGH 02
HIGH BASELINE
HIGH 02
BASELINE
TSB BASELINE
MIN. USABLE ATOM
MIN. USABLE ATOM
NO WASTE ATOMIZER
POOR ATOM
REDUCED ATOM.
LOW ATOM STEAM (VARY)
EA TRANSIENTS
GPM TRANSIENTS
WASTE STARTUP
GPM TRANSIENTS
WASTE OIL STARTUP
WASTE STARTUP
INVERTED SPRAY
START INVERTED SPRAY
INVERTED SPRAY
LOAD REDUCTION
LOAD DECREASE
LOAD INCREASE
LOAD INCREASE
START UP DUAL VOST
LOW LOAD, MODERATE 02
LOW LOAO/02/GPM
LOW LOAC/02/GPM
LOW LOAD, LOW 02
LOU 02 I GPM
LOW 02 I GPM
LOW 02 WASTE INCREASE
LOW 02
LOW_GPM/02
UUCjGPM
SOOTBLOW
SOOTBLOW
SOOT BLOW
SOOTBLOW LOW LOAD
CCL4
NA
NA
99.996
NA
NA
I 99.9988
) 99.9996
99.99997
99.998
99.9997
99.997
NA
99.9995
I.A
100
99.?9990
99.99994
99.98
99.9997
99.996
NA
99.9997
99.9998
99.998
99.9998
99.99989
100
100
99.9990
99.989
99.9997
99.9997
99.998
99.9990
99.9989
99.9994
99.997
99.9995
99.998
99.9997
99.99994
99.9997
99.998
99.994
99.998
99.987
99.999904
100
99.9998
99.998
99.7
100
99.9993
99.998
99.9996
CHLORO-
BENZENE
NA
NA
99.9996
NA
NA
99.9997
99.9998
99.99993
99.9986
100
99.99991
NA
39.997
NA
99.99988
99.99991
99.99988
99.998
99.9998
99.99989
NA
99.99989
99.998
99.98
Cs.3998
99.9997
100
99.994
99.9994
99.993
99.99989
99.997
100
99.99992
99.9998
99.99994
99.9990
99.9997
99.9996
99.99993
99.9994
99.9997
99.998
99.995
99.9939
99.995
100
99.9998
99.9989
99.998
100
99.9998
99 9998
99.97
99.9998
MMA
99.9997
99.999987
99.99992
99.99989
99.99987
99.999897
99.9997
99.9956
99.9998
99.9997
99.9997
99.9983
99.99t
99.9996
99.9997
99.9995
4-44
-------
Low steam production did not reduce ORE. Between 1508 and 2018 on May
9, the boiler operated at approximately 24% capacity. DREs during this time
average 99.999%. The load reduction from 52,000 to 25,000 pounds of steam
t-er hour during the early part of this period did not affect the ORE. The
load increase from 25,000 to 55,000 pounds of steam per hour at the end of
this period did not reduce ORE.
ORE did not correlate with CO concentration. Three of the five poor
DREs occurred during periods when the CO concentration was within the range
of good operating practice (less than 170 ppm). Yet, on May 9 at 1508 when
the CO was 1970 ppm, the ORE was approximately 99.999%.
The principal investigator reported (Reference 23) that "the boiler was
more forgiving than we expected". It was his observation that acceptable
(greater than 99.99%) ORE was attained at virtually any steady state operat-
ing condition. He reported that disruptions in the waste feed flow caused
ORE to be reduced but not dangerously even at extreme conditions. For exam-
ple, on May 8 at 1600 the steam to the waste atomizer was turned off. Even
with the stream of waste reaching the boiler floor, causing relatively high
CO emissions and stack gas opacities in excess of 20% approximately 25% of
the time, the overall DRE of chlorobenzene was 99.98% and carbon tetrachlo-
ride DRE was 99.9996%.
The conclusion, based on these results, is that this boiler is able to
maintain excellent DRE of hazardous components under any operating condition
within the range of good operating practice. The parameters that control DRE
are atomization of the waste fuel and stability of the waste fuel flow rate.
Poor atomization, direction of the atomized waste away from the oil flame to-
ward the bcjler walls, the fluctuations and rapid changes in the waste flow
rate all produced lower DREs. For the most part these lower DREs are not
less than 99.99% but are lower than typical values obtained during normal
operation.
These tests we::e done at a packaged water-tube boiler. The conclusions
are believed to be applicable to any suspension fired boiler. Inversion of
the waste spray (away from the oil flame toward the boiler wall) resulted in
slightly lower destruction of hazardous compounds and increased generation
of PICs. Similar effects might be predicted for fire-tube boilers. However,
the DRE did not fall below 99.99% during the inverted spray experiments, nor
was the DRE of the fire-tube boilers tested (Sites G and J) less than 99.99%.
The destruction of the hazardous waste appears to occur in the flame zone of
the boiler. Maintenance of a steady flame at a sufficiently high temperature
appears to assure destruction of hazardous compounds.
Site M
The boiler tested at Site M was a forced draft water-tube with a rated
capacity of 350,000 Ib/hr of superheated steam at 620 psig and 700°F. The
unit is front-fired with four CE R-type burners arranged in a square pattern.
Each burner is capable of firing gas and/or fuel oil or liquid wastes. Fuel
gas which is fired in all four burners is a mixture of natural gas and pro-
cess waste gas. Heavy ends wastes from a butanol/propanol production unit
are fired in the lower two burners. The butanol/propanol waste stream con-
tains surface oil from a waste-retention pond.
4-45
-------
The objective at Site M was to provide confirmatory tests to validate
Site L results using an alternative boiler design. The liquid waste stream
was spiked with carbon tetrachloride (CC14), chlorobenzene (MCB), and 1,2,4-
trichlorobenzene (TCB) to quantify DREs for parameters testing (nonsteady and
off-design operating conditions) during confiring in an industrial ooiler.
Several boiler loads, excess air levels, waste gas and waste oil flow,
and various transient conditions were investigated using waste oil selectively
spiked with individual or a combination of tracers (CC14, MCB, and TCB). The
concentration of the tracers were also varied by adjusting their rate of in-
jection. Boiler transient and off-design conditions such as sootbiowing, low-
excess air, lightoff, reduced atomization, unsteady oil flow, and use of dif-
ferent burner locations were aimed at identifying the POHC ORE failure limits
for the boiler waste combinations. Poor combustion and smoky conditions were
produced to determine whether ORE values were affected. In one tsst, the oil
injection locations was moved to the two top burners which produced high smoke
levels even when the excess oxygen level was increased significantly. In many
of of the tests, only one burner (lower right) was working, producing condi-
tions which were unfavorable for efficient combustion since the combustion air
was uniformly distributed to all four burners. Near the end of some tests
days, the waste oil feed was shut off, and the emissions where checked for
any residual PICs and POHCs which might be caused by hysteresis in the boiler
(emission of PICs and POHCs after these materials are no longer being intro-
duced into the boiler caused by POHC deposition on soot, and subsequent ela-
tion, etc.). POHC and PIC measurements were by the full-VOST (FV), mini-VOST
(MV), and EPA modified Method 5 (MM5). Summaries of the test conditions are
presented in Table 4.15 through 4.20.
Carbon tetrachloride and monochlorobenzene were measured by the full VOST
and mini-VOST. Compliance of the full-VOST data with the quality assurance ac-
ceptance criteria could not be fully assessed because the analytical results
were reported as a single value for the pair of VOST traps rather than for the
individual traps. No peaks were detected for Run FV-9, and therefore this run
is not considered valid. Since only one trap was analyzed for Run FV-6B, the
results of this run do not meet the second QA acceptance criteria and there-
fore are not included in calculated the average ORE presented Table 4.21.
The ORE (Table 4.22) of the three compounds used for spiking waste fuel
exceeded 99.99% during every run except one. For the lone exception (Run
MV-30 in which soot blowing was being conducted), the ORE was greater than
99.98%. As with the Site L test, the high ORE values measured for the chlor-
inated PCHCs did not seem to be affected by the various trst conditions aimed
at producing poor combustion with high smoke and CO emissions. Values of the
DREs measured under these conditions were similar to baseline values measured
under normal load and excess air conditions.
Benzene was detected as a measurable POHC in the waste oil. Because the
concentration of benzene in the waste oil was very low (barely exceeding 1000
ppm) and because benzene is readily produced as a PIC for the waste oil and
natural gas combustion, benzene ORE values calculated from the test results
were not considered an accurate indicator of the actual DREs during the waste
oil combustion and are not reported here.
4-46
-------
TABLE 4.15
MINI-VOST TEST DESCRIPTION
SITE M
'-•037 I
HIE I
TIDE SMPU9
LOAD
OIL nmcxs
n OIL AIM
o: MSTE sria FUESS P«SS
ID EAS Oil RCt CCI4 TCI (|ii|l lpu»'
ttatt
OVERALL KSOtlPTION
Ml UM trm
rt kottM «•
W-l
w-2
w-j
iw-4
n»-S
W-4
ll»-7
(N-«
mM
WHO
HV-ll
w-12
IW-IJ
m-\4
W-13
»»-u
W-17
*HI
IfiMf
W-20
W-21
w-a
fiv-rj
BV-24
•w-a
IW-2»
w-n
W-3
IP-2T
W-3»
w-;i
IW-E
w-:i
IW-J4
W-33
IW-U
7/25/B
7/a/B
7/24/B
7/24/B
?'»/B
II2UK
7/24/B
7/24/B
7/27/B
7/27/B
7/77/B
7/27/B
7/27/B
7/27/B
7/27/B
7/27/B
7/71/B
7/21/B
7/21/B
7/a/B
7/a/B
7/n/B
7/a/B
7/21/B
7/21/15
7/W/B
7'CT/B
7/71/B
7/WB
7/21/B
7/JO/K
7/M/K
7/M/B
7/M/B
7/JO/B
7/M/I3
woj-«i:
1000-IOJO
07W-07U
113-1143
U10-I3M
I440-I4J*
It^-IMO
IUM702
M2tKH»
«I4«-«BO
mOH)«4«
OT57-I007
1)00- 1 1)«
1S4-I404
l«2»-l«»
irtj-ira
»»iikt»4<, IM IM*
LM IMI
LM lot*, IM ur
Milk oil
-------
TABLE 4.15—Continued
V«l 1
BV-37
W- .'1
«V-;T
IK -40
IW-4!
IW-42
n»-45
m-44
IKMJ
IW-4*
W-47
W-41
MMt
IW-JO
IW-31
M-J2
W-J3
IW-54
W-JS
KTE I
7/JO/B
7/JO/B
7/JO/B
7/JO/B
7/50/B
7/J1/B
7/JI/I5
7/Jl/B
7/JI/B
7/JI/B
7/JI/B
7/31/B
I/OI/B
I/OI/B
I/OI/B
1/01 /B
I/OI/B
I/OI/B
l/vl/B
T1K SAWLEI
i4ri-i4:s
isc5-u:j
1330-IMO
144V-1430
1703-1713
ms-ota
1202*1212
12:1-1241
1403*1413
UIMW7
1433*1443
1110- ICO
•CtHXM
Of|4*«T7<
IOS-1043
12*0-1*50
1410-1420
I443-I4S3
HKw»»
LOW
lor Ml
N^ Ml
**TMl
HOTMl
»irriM
Lw
Mrt4l
Mrul
•TMJ
Nroil
MTNl
Mrul
MTMl
Mr Ml
Mr Ml
Mr Ml
MrMl
Mr Ml
l*
OH M«*S
OK Olt. 4W
02 MSTE SPUE PUSS PttSS
(I) US OIL Kl C£)4 TCI (ttiol If HI'
llMt Ml U
rt tottw on
4.4
4.4
_
4.1
4.f
4.f
2.'
J.O
:.t
4.»
J.4
1.3
t.l
4.1
4.0
7.2
7.2
I
I
.
on
i
!
I
I
I
a
31
a
jt
w troo
•l
IM
IM
IM
IM
IttIM Ml
40
4*
40
30
30
30
15
M
M
30
53
33
33
IM
IM
IM
IM
IM
IM
IM
IM
IM
IM
IM
IM
IM
OVEUU Kscitmw
lw CD 4
Hi-* B14
iMitMl PONC k PIC Ck*Ct
taitMl PONC 1 PIC c»fcl
IMIIM) POHC i PIC ckttk,
WttlMr M] Mw, IW I Ml
NlM TCI
Htfk TO
Ntf* TO, lw Mr, uotr
ItwtM TCI, lw *ir, ittr
tuteti TCI, lw Mr, utr
Sottllwtif
iHltMl PONC 1 PtC CMCt,
HIM Mr, M 1*1 IH
Hit* Mr, tit* Kt 1 C14
lw 4ir, MIT itxt
lw Mr, MotT. MOD Id I
itif» ro i KM
Mot Kt i CCJ4
M Ml
M til
CC14
4-48
-------
TABLE 4-16
MINI-VOST TEST CONDITIONS
bITE M
VPST
UUHBER
nv-i
flv-j
nv-~,
nv-4
nv-s
nv-4
nv-7
nv-i
nv-»
nv-io
nv-ii
nv-i2
nv-i3
nv-i4
nv-i3
nv-u
nv- 1 7
nv-i8
nv-it
nv-jo
nv-2i
«v-::
nv-:;
nv-2*
nv-23
HV-t*
nv-27
nv-:«
nv-2t
nv-50
nv-3l
nv-::
nv-33
nv-34
nv-53
nv-3*
nv-37
MllER
OUTLET
TEHP
< Fl
It!
34i
580
383
580
383
540
340
3»0
3»0
380
380
383
380
380
370
530
343
340
380
573
573
600
380
541
384
580
600
373
573
570
373
373
373
570
370
370
STEAK IOHER
LOAD o:
IIOO" (11
16/hr)
ISO
ISO
18*
184
182
188
:33
130
180
180
180
180
180
17*
17*
1*0
140-1*0
128
i:*
W
184
180
1*9- l»0
170-185
170-183
110-163
180
176
176
17*
173
170
172
177
172
• 182
172
i.:
i.:
4.5
3.5
3.1
3.3
3. 5
3.4
6.3
6.3
3.3
3.3
3.3
5.3
6.0
3.2
4-8.3
8.0
4.1
6.7
«.i
5.0
3.0-8.0
1.5
3.0
8-10.2
6.0
».3
3.0
4.8
8.0
6.2
7.3
8.0
6.2
*.4
6.4
fUll HOU RATES TRACER INJECTION RATE
NAT HASTE HASTE
GAS 5AS OIL CC14 nd TCI
(ICK'G (H'OO <9P» (Pp*> <1P»
O»C (qp«>
140
140
1)5
115
US
120
no
110
130
130
130
130
123
i:3
1:3
ISO
110
no
110
135
130
130
125
210
1*0
133-170
1*0
1*0
1*0
1*0
130
150
130
130
1*0
1*2
1*2
75
73
67.5
67. S
tt
*f
0
0
45
43
45
43
44.2
44.2
44.2
47.23
0
0
0
0
0
0
0
0
48
0
0
0
0
0
N
0
0
0
0
0
0
c.o
o-:.:
6.0
3.3
3.3
3.3
3.3
3.5
5.7
3.7
5.7
5.7
5.4
S.»
5.»
0.0
3.8
4.0
:.7
11.5
11.6
11.8
11.6
0.0
0.0
7-13
.*
.0
.0
.0
.0
.2
.4
.8
.?
t •
.8
0.00
0.00
O.l>0
0.00
0.00
o.oo
0.00
0.00
o.co
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.17
0.18
0.23
0.34
0.30
o.;j
0.2*
0.00
0.00
o.:s
0. IS
0.15
0.13
0.13
0.00
0. 1»
o.:*
0.64
0.64
0.10
0.10
0.00
0.00
0.00
0.63
0.61
O.*l
0.54
0.34
.6*
,04
.0*
.">*
.03
0.0*
0.0*
0.00
0.03
0.03
0.04
0.14
0.13
0.13
o.i:
0.00
0.00
0.08
0.0*
0.03
0.03
0.03
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
g.oo
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
o.oo
0.30
0.2*
0.2*
0.2*
0.2*
0.00
.00
.00
.00
.00
0.00
0.00
FLUE
6AS
(1000
att«ht
46.5
30. t
31.8
50.8
53.6
33.5
3*.*
30.2
35.;
35.3
-------
TABLE <. 16—Continued
to inn
••'031 OUt. El
MuntC* TEnf
( Fi
nv- ;a
nv-*«
nv-40
«V-4|
flV-42
(IV- 43
nv-44
nv-43
*V-4*
l»V-47
NV-41
«V-4»
HV - 3<>
WV-51
«v-s:
nv-55
nv-34
nv-53
170
370
570
179
5*0
570
370
570
943
9*5
519
370
S»0
515
570
570
5*0
5*0
FUEL FLOW RftUS TRACER INACTION RAU
STEAK I01LEK NAT MASTE HASTE
LOAD o: s*s SAS on c:i4 nci TCI
11000 H) (10VO (1000 <9P«I (gp*J <9P«'
!6/hr) d«Cfh) S»:»h) lgp«)
170
llv
171
1 33- It3
HI
17:
17:
17:
110
110
i?:
119
170- in
110
175
J?3
1*5
1*8
4.4
—
..
—
4.3
«.»
4.?
:.«
J.O
2.f
».T
;.4
1.5
M
4.1
4.0
7.:
7.2
it:
u:
1*2
l»3
1:2
150
15"
ISO
150
130
190
130
t;o
i:5
121
in
1:1
1:0
0
0
0
9
0
0
0
0
0
0
0
*J
*4.S
4*. 5
4*. 5
4*. 3
4*. 3
0
*.f
*.3
*.S
0.0
:.7
s.»
5.7
fe.4
*.l
*.o
*.o
0.0
*. 1
*.o
*.o
*.o
*.2
*.2
0.*4
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
o.co
0.00
0.00
0.00
o.::
0.1»
0.22
0.22
0.22
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.24
0.17
0.24
0.24
0.24
0.00
.00
.CO
0.00
0.42
o.:7
o.;7
O.J7
0.42
0.42
0.21
c.^o
0.00
0,01
O.C1
.0
-------
TADI.E 4.17
VOST TEST DESCRIPTION
3ITE M
ui
VOSI i
MIE I TIKE SAHPLEI
SIEAN
IOAI
OIL BURNERS
ON Oil MOD
CICESS MSIE SPIKE PRESS PRESS
•2 111 GAS OIL NCI CCI4 ICI lp«i|l
(Boil
OVERALL MSCRIPTION
oil HON Iroi
rt boltoi gun)
fV-l
FV-2
FV-l
FV-4
FV-5
FV-6A
FV-61
FV-7
FV-B
FV-9
FV 10
FV-II
FV-12
FV-IJ
7/24/83
7/2i/i5
7/27/13
7/21/85
7/21/83
7/29/83
7/10/85
7/10/83
7/30/85
7/31/BS
7/31/85
1/01/85
8/01/85
8/01/83
H3S-II5S
1410-1500
0930-0150
I05S-III3
1130-1210
1600-1420
1200-1220
I2S2-I3I2
1530-1610
0915-0935
1405 H 25
1035-1055
1240-1300
1410-1430
Nor ill
Nor ill
Nor ul
ION
Varying
Mortal
Nor ul
Norul
Mortal
UN
Mortal
Mortal
Nor »l
Mortal
3.5 1
3.3 I
3.7 I
4.2
t.0-9.1
9
1.2
4.2
4.3
4.4-3.0
3.1
9 I
4.8 1
4 I
1 I
I "
I I
I I I
I I 1
I 1 I
I I
I 1
I
I
I
I I 1
I I 1
I I I
R
L,R
R
l,R
I L
I R
R
ft
R
I It
I R
R
R
R
22
24/28
SO
25/25
ISO
72
SO
38
38
40
SO
60
50
55
Ui
OFF
ISB
140
140
160
160
140
140
140
160
140
160
160
NCI
Lou
lOM
ION
Lou
High
Hl,h
High
ipltf iiarl
«ir atoiiiation
air, haiy
load, ION air
load, light oil,
air
CCI4
CCI4
Risldual POHC 6 PIC
Usttady oil (Ion,
High
High
ICI, ION air,
air, high HCI
off
iioky
check
I OK load
iioly
I CCI4
LON iir, hny
LON
air, high stokt
, high KCI 1 CCI4
-------
TABLE 4.1U
VOST TEST CONDITIONS
S1TK M
\
fc.
I
tn
BOILER STEAM BOILER
VOST OUTLET LOAD 02
NUMBER TEMP (1000 (X)
(deg F) Ib/hr)
FV-I
FV-2
FV-3
FV-4
FV-3
FV-6A
FV-6B
FV-7
FV-8
FV-9
FV-IO
FV-ll
FV-12
FV-13
585
585
580
545
580
600
575
570
570
560
570
385
570
570
184
1B8
180
128
108-184
176
177
170
175
140-li.
172
176
180
155-180
3.5
3.3
3.7
4.2
9-9.8
9
8.2
6.2
3
Vi
>>.!
9
4.8
4
FUEL FLOW RATES TRACER INJECTION RATE
NAT HASTE HASTE CC14 MCB TCB
GAS GAS OIL (qpa) (qpn) (gp«)
11000 (1000 (gpn>
dscfh) dicfh)
113
120
130
110
135-170
160
ISO
160
162
132.5
ISO
125
128
128
67.5
69
45
0
0
0
0
0
0
0
0
46.5
46.5
46.5
5.
5.
5.
4.
3
4
7
1
0.2-12
6.
6.
6.
6.
3.
6.
6.
6.
6.
2
4
9
3
7
6
0
0
2
0.00
0.00
0.00
0.03
O.IS
0.16
0.64
0.64
0.00
0.00
0.00
0.22
0.19
0.22
0.65
0.60
0.06
0.04
0.05
0.06
0.00
0.00
0.00
0.00
0.00
0.24
0.17
0.24
0.00
0.00
0.00
0.00
0.18
0.2/
.00
0.00
.00
0.42
0.37
0.01
0.01
.00
FLUE
GAS
(1000
dtrfM
51.8
51.6
49.8
39.5
64.9
70.0
50.6
46.3
4B.O
46.0
50.1
63.7
49.7
49.0
-------
-A
TABLE 4.19
MODIFIED METHOD 5 TEST DESCKll'TION
SITE M
EICESS
OIL BURNERS
ON OIL ATOM
row •
imiC * 1
inc annrLCV
LUHV
IK »H'
(X) 6AS
OIL HCB CC14 TCB
(Noit
rut 33 rrtai UVLKHU utsiiuriiuR
Ipsigl (ptig)
oil floit (rot
rt bottot gun)
IW5-1
HH5-2
BH5-3
HHS-4
IM5-5
M15-t
IM5-7
HN5-8
HN5-9
HN5-10
HN5-11
HI15-12
HH5-13
W5-14
7/26/85
7/26/85
7/27/85
7/27/85
7/28/85
7/29/85
7/29/85
7/29/85
7/29/85
7/31/85
7/31/85
7/31/85
7/31/85
7/31/85
0740-840
1232-1332
0926-1026
1352-1459
1055-1155
1134-1238
1334-1434
1337-1641
1810-1910
0910-1010
1131-1231
1346-1447
1608-1708
1810-1858
Mortal
Mortal
Mortal
Mortal
Lou
Varying
Mortal
Mortal
Mortal
Lou
Mortal
Mortal
Mortal
Mortal
3.9 I
3 I
3 I
S.S
4.5
5-10
5.7-8
9
5
5
4.9
3.1
3
6.9
1 L,R
I 1 L,R
I I R
I I R
III L,R
I I I
I I I
I I I
I I I
I
X
I
I
I
L.R
R
' R
L.R
R
R
R
R
R
28
28
SO
65
25/22
ISO
70/70
72
70/65
45
40
50
50
45
152
166
158
160
158
160
160
160
160
160
160
160
160
160
Biiilini, norul load t t\r
Lo* air, itoky
LON air, haiy
SootblONing
LON load, ION air
LON load, lightoff,
Baseline Nith HCB,
High air
SootblONing
Unsteady oil (lew,
High TCB
ION air, high TCB,
Reduced TCB ION air
SootblONing
Sioky
CC14, TCB
ION load
uoky
i n'?y
-------
TAfaLE 4.2C
MODIFIED METHOD 5 TEST CONDITIONS
SITE M
MB 1 BOILER
OUTLET
TEHP
I f)
fl!fi-l
IWS-2
IW-3
IW5-4
mS-5
BB5-6
IWS-7
HIS-B
MB-9
IW5-10
F.S-11
W-12
BR5— 13
IW-14
580
580
580
580
540
580
580
MO
575
55?
570
567
565
585
STEM
(1000
lb/hr)
184
182
183
176
126
105-200
!80
176
176
145
170
176
180
172
02
C)
3.9
3
3
5.5
4.5
5-10
5.7-8
9
5
5
4.9
3.1
3
6.9
FUEL
NAT
GAS
(1000
dscffi;
115
115
125
125
110
135-170
160
160
160
132.5
150
ISO
150
150
FL£» RATES
HASTE HASTE
SAS OIL
(1000
tfSCfil) (QPI)
67.5
68
49.5
0
0
0
V
0
0
0
0
0
0
0
6.0
5.6
5.5
5.9
3.7
0.2-14
6.3
6.0
6.0
3.5
5.8
6.1
6.0
6.0
TRACER INJECTION RATE
£14 WCB TCB
(gpi) (gpi) Igpi)
0.00
0.00
0.00
0.00
0.25
0.3
0.15
0.15
0.15
o.ot-
0.00
0.00
0.00
0.00
0.00
0.61
0.06
0.06
0.04
0.08
0.06
0.05
0.05
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.30
0.2?
0.26
0.26
0.42
0.37
0.37
0.42
0.28
FLUE
SAS
(1000
dscfh)
58.2
53.6
49.8
58.5
39.5
64.9
54.1
70.0
55.0
46.0
54.9
50.1
50.4
62.0
4-54
-------
TABLE 4.21
SITE M
Manufacturer:
Type:
Fuel:
Waste Stream:
Combustion Engineering
VU-60, Water-Tube
Natural Gas, Oil, Waste Gas,
and Waste Oil
Butanol/Propanol Heavy Ends,
Process Waste Gas
Design Steam Kate:
Design Steam Pres.:
Test Steam Rate:
Fraction Waste Fuel
Meat Input:
3bU,000
t>20 psig
200,000 Ib/hr
36 - 4B x 10fe Btu/lir
Compound
Carbon Tetra-
chloride
Chlorobenzene
1 , 2,4-Trichlo-
robenzene
Feed Rate
mg/sec
3,018-
64,395
2,095-
45,391
23,717-
38,312
Quantity Found (ug)
Baseline
Run
MV
Oa
Oa
FV
b
c
MM 5
0
Average
Blank
MV
0.0005
0.103
FV
0.03t
0
MM5
0
Test Runs
(Avg)
MV
0.073
0.129
FV
0.166
0.057
MM 5
11.7
Test
Emission
Rate
(mg/sec)
MV
184.6
285.5
FV
MM5
191.9
Blank
Corrected
MV
No
Ho
FV
No
No
MM5
No
UHK
MV
99.9995
99.9998
FV
99.9997
Mf
99. ';
No CC1. or CgHcCl was detected during first three baseline runs. During runs made to test for hysteresis (i.e.,
after POHC feed stopped: MV-16, MV-24, MV1-5, MV-31, MV-39, MV-40, MV-41, MV-49, and MV-50) the collected CC14
ranged from 0.0025 to 0.365 ug rfith an average value of 0.072 ug while C^H^Cl ranged from 0 to 0.024 u
-------
TABLE 4.22
ORE GROUPED BY TEST CONDITIONS
SITE M
7-r »
nv-r
MS-?
HV-32
«v-::
«•;-«
HV-10
nv- 1:
FV-bj.
iv -28
NKrZ
HV-5
HV-11
fiv-12
FV-3
flV-21
flV-22
"V-52
rv-33
FV-13
FV-12
WI5-12
HM5-13
FV-10
HV-6
FV-2
HV-7
HV-6
HV-IB
HV-19
FV-4
HV-26
FV-5
wis-6
M5-10
HV-35
HV-54
KV-31
FV-11
HV-i4
OVESAU BESCRlPTiOk
fcaseime «tr, Kb, CCK, TCt
tiseline *itn HCfc, CCK. TC(
horiii loiC inC air
Norii] lead ind air
Rax tan cacacity
na>
Lo« air, naz*
Lo» air, nuy
Lou air, na:r
Hi on oil firing, siory, Jo« air
Kign oil tiring, s»o»y, lo» air
Lo« »ir, nary s:»c>
LOM air, iioM, high HCB i CC14
LOH air, hign siote, high IO I
Lot* air, haiy
Lo« air, high TCB, sioky
Reouctu TCB lo* air, ha:\
High TCi, lo» air, saoky
Lw «ir, atoi;:anon ot-
Lou air atonzation o44
Lo« load, no «ast( gas
5§oK>, lew loifl t »r, no »astc
Lo* load
Lou load, lo« air
Lo* load, Ion air
Lo« load, hghtof4, sioiry
Lou load, h;MoH, sioky
Lo* load, lightoH, sioky
Unsteady oil flo«, IM load
Hi oh nCB I CCH
High nCB ; CCK
Hioh air, hioh HCB ', CCH
Hi or- 2ir, high nCB V CC14
Hi on CC1*
DKI
ecu
9?.0(»?61
99.99B3?
99.957*:
—
100
9?.9??8B
—
99.99949
99.99940
99.99937
99. 99932
CC14 99.99968
99.99967
_
—
—
9»s __
99.99933
99.99951
99.99681
95.99922
100
—
99.99755
99.99963
95.99375
99.99976
99.99931
DRI
IO
95.99M
99.999^
IOC
99.9995:
99.996B1;
9S.99B8£
—
99.99491
99.99924
99.9997E
99.99862
99.9995e
99.9985S
99.99947
99.99990
99.99938
99.99860
__
_
99.9974«
100
99.99166
100
99.99705
100
99.99789
99.99967
99.99973
—
99.99996
99. 99994
99.99981
99.99884
tf,: "jr~.
Kt BchlENE
W- I
.-
_ _
~ 95.76771
iw- ~
~ 94.7164E
_
_ __
~ 97.4939E
_ 99.68780
99.99773 _
95.9997* ~
98.64495
_ _
~ 9S.2914T
~ 95.9977?
99.99991
99.99934 ~
~ 99.86947
4-56
-------
TABLE 4.22—Continued
J"WVSWMB
TE." 1
FV-er
fl » ' •> J
FV-'
HV-3P
HK-1I
ir.-;e
m-:7
«v-4
FV-!
CVc-VU. DEsCiiJrTIDli
Higr CCH
Higr, CC14
Hijn CCH
Hi:r. CCJ4
hi;" TCi
L9* CC14
1C. CC14
Ki ss ike star:
RC5 SBik* start
Dfit P't
C3.4 BCr
99.99911
99.9992B
99.9986'
99.99981 _
—
99.99454
95.9979; ~]
— 9?. 997«;
.. 9?.99;9;
C-: Df:
TCfc HHZEH:
99.3917E
~ 9E. 69712
«9.99S9? "
.-
99.10e(-:
IK~\i Oil hgntoH. lo> loac
HV-J4 So&ttlowing
HV'ij SQQt^lOHinQ
HP!'-14 5oot61o»m5
BV-29 Scstbloaing
BV-30 Social owing
W!-9 Sootblo»ing
HV-20 High oil firing
BV-23 Sao>y 2 too burners only
FV-9 Usteady 014
- 9?.9999i
IOC
99.9998C
100
99.99891
100
99.9997* 99.9997c
99.99975 99.99941
AVERASt ORE
99.9991 99.9957 99.9994 98.8573
4-57
-------
I* is concluded from the results that the co-fixing of hazardous waste
:• the tested boiler produce very high DREs of the waste compounds witnin
trv'" normal range or boiler operatinc conditions including transients such as
startup, insufficient combustion ciir, and sootLlowing.
Site N
The boiler tested at Site N was a coal-fired spreader stoker with a rated
capacity of 150,000 Ib/hr of saturated steam at a design pressure of 290 psig.
It is equipped with a fly ash control system, consisting of a mechanical col-
lector in series with a baghouse. The boiler is used to dispose of industrial
sludge which is injected at opposite sides of the furnace at approximately 3
fact above the traveling grate.
The objective at site N was to provide ccnfirmatcry testing to validate
parameters testing and Site L and M by measuring principle organic consti-
tuents (POHC), ORE and PIC emissions for a typical coal-fireei spreader stoker.
Chlorinated organic and metal traces were added to the sludge to broaden the
range of POHCs available fc>r ORE quantification and to determine the parti-
tioning of hazardous metals to the ash and flue gas streams generated. Tests
were also performed with oil used as a carrier for these trace pollutants.
Eighteen boiler operating conditions, shown in Tables 4.23 through 4.26,
were investigated during the site N test program. The test matri:; consisted
of four distinct series of tests. The initial four baseline test conditions
established the background emission data with coal combustion only and coal
with sludge co-fired. Background emissions with coal and No. 2 oil feed were
measured during test condition 5. For these background tests, the sludge and
No. 2 oil were not spiked with either organic or inorganic tracers. Both nor-
mal and low steam load test conditions were investigated. Excess combustion
air was maintained at normal operating levels during these tests corresponding
to approximately 7% excess ©2 at the boiler economizer exit. Baseline tests
were repeated during test conditions 14A and 17 to determine the reproduci-
bility of initial baseline results during and at the conclusion of the test
program.
The second series of tests, 6 through 10, consisted of tests with coal
and sludge spiked with TCE and TCB, alternatively. Bo:.ler operation was set
on manual and tested at normal load and excess air levels (test condition 6
and 71, low load and high excess air (test condition 8 and 8A), low load and
low excess air test condition 10). Combustion excess air was varied by ad-
justments in both undergrate and OFA.
The third series of tests, conditions 2 throuqh 14A, were performed with
the boiler co-fired with No. 2 oil spiked with TCE and TCB, alternately. These
tests were performed to provide a comparison of POHC, ORE, and PIC emissions
with an alternative liquid stream. As in the case of the sludge co-fire test
series, variations in combustion excess air and boiler load were investigated.
The fourth and final series of tests, represented by test conditions 15
and 16, were performed to investigate the partitioning of metal tracers (lead
and chromium) spiked in the sludge and No. 2 jil liquid streams. During these
tests, TCB was also simultareously injected along with the mefaJ tracers. ORE
calculations were performed on the TU2 and TCB POHCs spiked in the simulated
4-58
-------
TABLE 4.23
SITE N TEST MATRIX
l»Ul»t r«l« l»r-t
,1111 •ttlrr lo«*l tt»r*4Ut«n tl»» U/*r) *lr tt'ilftf ttt*tt Of
iMpllfff k*fh*Mt*> InlM
\««vMno| b«<|>HMM« iMllo?t
1
1
*
1
i
r
i
M
1
10
It
11
14
|*J"
11
It
IM 140
IM 140
00- W
on to
IM 140
IN-140
130 140
M to
00 W
Ml W
to M
MM
Ml to
1 Ml 140
Ml M
190-140
ftu«l
H.rm.1
Ural
IhTMl
•or-Ml
Hl«k
It.
•OMMI
ftorwl
m*.
It.
Mortal t
fturiMl
litfh
UfMl •• -- II 1*111
H.KMI »<•*»» «-t - -- •- 1 1 1 f 1 1 1
Motttlt -- •- .... . | .. ,. j |..|
>>rMl Sll»ltt 4 0 ..-• -- 1 1 -Jill
AMMl •>. t til 4.0 .... .. 1-1)4
tenMl SIM* '.0 O.t •• •- 1 -- 1 1 1 1 -. -.
H~~l Sl«l* 4.0 - ».? - 1 .... | 1 |
Nlfli llXtt 4.0 O.I -- •- - •- 1 I 1 1
tto iUiV «0 O.I •- > 1 I » • 1 .-
IkKfMl JU .
Hiyi ••. f til 4.0 - o.r •- 1 -- -. f i
(n> «.. till 4.C 0.? -- 1 1 1 » 1 1 1*
Hoi alii Hottll 40 -- .- •> * t 4
(Mr-., Ho. t til "• -- O.f O.I 1 -- < 1 1 1 1
• l»» Iln4tr '« -- O.t O.I 1 1 » t 1 |
Htr«4l Slw^fc 40.-*. -. 1 | 1 t •
1
1
1
1
t
1
1
1
1
1
••«!* lr>4 ••< .hrp.1^. •llctlt >lll tt IxjcclKl IK I »•'«• ttUIIwi II < In S»etl«« I.
•SotltUott l» "-II.-J 1 Kill) 'oilllt* K«tl>«4 S It liKUO IMlitUlllt Iff4tr Kl ••Illlvil
«<%.H'I~ »J!it''< i»/«/. rn,. ro. NT. \o>. 1^1 w.
-------
TABLE 4.24
BOIUEH TliST (JONUiflUNS — BASEL.INK TKSTS
SITE N
o>
SlfM
•Hltlo* Itll 4>lllt* IIO1 It/hr) U
dli !M"I^' {ill!" '!»•!"
]
4
4
l««
•• ut.« i»4. tt to m
Ml tr.ll |0t|
» U IVt
«>l .-4 il.rt^ (01) IHM
l|>> M-»0»
(Wt|
m
m
inn
0 H|0|
«*••!»»• fr««l «»«r rvlitjrctl
t.O
(I
It
If
(I
l.t
(1
t.l
l»
I.I III II II 10
II
11
0.« It 1 It H It
tl
t.O It It II It
M
t.t II II 14 10
•1
l< If t.t l< 10
It
1)
•nil "«)•
t»ll.(l.r fr,i
t* (In N}0| (In
-00 4.1-
II
< 1 1.4
11
41 l.t
(1
-I.I 4.J
('
• \.» •
Met •"!
r
HjO) Cl| I.I.I
1.0 4110
M
'1
I.I MO
M
I.I US
1)
4.1 loo
tl
« lit
11-11
(Ml
'• -i;
I'M
M c. > oil
..'(•"II <»r-i l.,J.cll»o I«|"
1.5 1.0
111)
HOI I'.tl II M
1.0 1.0
If tl
1.0 I.I t.l I.I 1.1
C.ni U.It
1.0 1.1 •- -- >.« I.I
III!
1 -- I • 0 l.«
II II II II
-------
TABLE 4.25
SLUDGE CO-?IUE TEST CONDITIONS
SITE N
III •««!..««• It.
>»»t
(•MUM
•4
t
14
...
1.11 4.1
• Ir. tMl
1.4 III
!•* I*M.
llr <«4»
«* M*
I— I.M.
<••'. •)•
III
!«. I.M.
ir'ni;
4 111
.'UU.
.-.ix,
Id
1*. .1.
• ~l
.1 III
J III
U»v* t.M ri*%t."
til 1H 1*1
lll»l«
•4*4 HI
41 f* 1*4
IM'
44 t* IM
IIMI
it n i"i
HI l« M 1' 1 N t 14 «• •• Mt !!-!•
(Ml
ni t.ii.l u • it n li t\ M* n it
ii» mi
m r* i« t ir i ii n % > > • tit 11 tt
II II INI
ni rti.' it i MI* it t •• •• it* ft ii
li ii mi
u «i
• • it
M i < 1.1 1 >
nn ii «i
ii i 1.1
li u
III! II M
It l.t 1 t
II II
III II* UI.H<
II II II III
II II • <*
It II
II til
II II
I.I I.I t If
II II
II II II II
I.I I.I III t l«
II II
-------
to.
Kl
TABLE 4.26
NO. 2 OIL CO-FIRE TEST CONDITIONS
SITE N
II
1)
In !«««. «ll ••>«
1(1
It- I.X. Mfk
•Ir, «ll |«J ICI
1C* ««»J *•!•!«
•II -Ilk Iti
«• frtti.ir m (H/l|
«t«« !«.<-.,« ~ '
t1 tfc/tr) (M*t) U»*t) U»M»M|| IV»4*r Irmt ft*«r rrlA)MllMi
M ,1 IM )»< HI (.1 l.t 11. t I.I I* It
Mil* H«l l*.'l
»IH IHttl 1H t K.I III It II. 1
Iti) l»ll
(i*0 U.'l (H M
(tl) (>tl| fr>K |<»t1K» Ml " " "~ " Ill lit *..!«
•« « l.t III. Ill II-N 1 I • |.» - nil
ll»l IKI
•1 t.t lit tin rt-M 11 i.t ii nn
DM* on
(I/I
(> "1 DM) I'll
»,
-------
liquid waste streams. Tables 4.27 tnrough 4.29 summarize these results. ORE
results on TCE from six test conditions and a total ot 15 individual measure-
ments clearly indicate that the destruction of TCE was approximately 99.999'J%
or well aoove current regulations for commercial incinerators. Some reduc-
tion in ORE is evident during test condition 8, where excess air was increased
well above the typical operating setting of. the boiler. ORE results for TC!i
averaged 99.997% indicating thermal destruction approximately equal to that
of TCE.
It is concluded from the results that co-firing hazardous wastes in *.he
coal-fired spreader stoker boiler produces high DREs of waste compounds within
the normal ranges of boiler operating conditions including low-load and hi^h-
excess air.
Site O
The boiler tested at Site O was a coal stoker that had been converted to
fire natural gas, fuel oil/ or liquid wastes. It is capable of delivering
22,000 pounds of 95 psig steam per hour. The exhaust gas from the boiler was
tested by EPA Method 23 (inert plastic bag samples). All other tests reported
herein were done by either VOST or MM5, or ooth. Analyses were done on-site
by GC/FID or GC/ECD. Samples from all other sites were shipped to the labora-
tory for GC/MS analysis. The waste fuel consisted of alcoholic still bottoms
to which the test team added methyl chloroform (1,1,1 trichloroethane). The
waste contained no other hazardous material (Appendix VIII).
The QAAC for this site was a demonstration that the method (Method 23)
would have detected methyl chloroform had it been present. Before going to
S'te O the testing organization added known quantities of methyl chloroform
to samples of the exhaust gas from a residential heater that was burning a
fuel similar to the Site 0 waste fuel. Recovery of the spikes was good (95
to 96%). There was a further demonstration that low concentrations (0.5,
0.005 ppm) of met.iyl chloroform were stable in the bags for up to 3 day^.
Ml sample analyses were completed within an hour of their collection.
These tests were taker as evidence that had methyl chloroform been present
in the stack gas at Site O, it would have been detected. Therefore, the
data were accepted.
All stack gas methyl chloroform concentrations were less than the limit
of detection of the analytical procedure (Table 4.30). The test average ORE
of methyl chloroform was greater than 99.999*. The boiler operated at the
same steady conditions throughout the seven test runs. The carbon monoxide
concentration in the stack gas during runs 2 and 3 (210 ppm and 110 ppm) was
higher tha;* the average (40 ppm) measured during the other five tests. Thare
was no discernible effect on the DRE of methyl chlrroform nor on the other
two non-hazardous compounds (methanol and methyl acetate) that were measured.
Products of Incomplete Combustion
Several means have been suggested by which compounds that were not pre-
sent in the fuel or waste fuel burned in a combustion source may appear in
the exhaust jas of that source. There ara two general mechanisms by which
products of incomplete ccmbustio.i (PICs) may be formed. One is formation of,
products of partial oxidation of fuel components. The formation of aldehydes
4-63
-------
TABLE 4.27
TCE DESTRUCTION
SITE w
TCE emission rate
TCE
TCE-ORE (percent)
injection
Test
no.
6
8
8A
9
10
14
Weighted
rate
(9/s)
17.62
18.55
40.80
21.33
20.40
19.47
average
Run
no.
1
2
1
2
1
2
1
2
1
2
1
2
Mini -VOST
(ug/s)
17
33
130
32
13
10
NO
NO
8
5
2.8
2.8
VOST
(ug/s)
38
72
NA
NA
NO
NA
Mini -VOST
99.99990
99.9998
99.9993
99.9998
99.99997
99.99998
100
100
99.99996
99.99998
100
100
99.99991
VOST
99.9998
99.9996
NA
NA
100
NA
99.9998
NO — Not detected, less tnan 1 ug/sec.
NA -- Not available, sample not taken.
4-64
-------
TABLE 4.23
TCB DESTRUCTION
SITE N
TCB TCBa
injection emission
Test rate rate TCB ORE
no. (g/sec) (yg/sec) (percent)
7
12
13
15
16
Wei
19.81
22.78
20.80
19.81
20.80
gnted average
17
120
59
32
74
99.99991
99.9995
99.9997
99.9998
99.9996
99.9997
aBaghouse inlet condition — emissions
are not blanic corrected.
4-65
-------
TABLE 4.29
SITE N
\
Manufacturer: Riley
Type: Coal-Fired Spreader Stoker
Fuel: Coal
Waste Stream: TCB & TCE Spiked Sludge and No. 2 Oil
Desiyn Steam Pate:
Design Steam Pres.:
Test Steam Rate:
150,000 Ib/hr
291) psi
-------
TABLE 4.30
SITE 0
Manufacturer:
Type:
Fuel:
Waste Stream:
Combustion Engineering
Converted Coal Stoker
No. 6 Oil and Natural Gas
Alcoholic Still Bottoms
Design Steam Rate:
Design Steam Pres.:
Test Steam Rate:
Fraction Waste Fuel
Mass:
Heat Input:
22,000 Ib/nr
95 psig
7,600 Ib/hr
100%
100%
Quantity Found (ppm) Test
Feed Base- Aver- Test Emission Blanka
Rate line age Runs Rate Cor-
Compound mg/sec Run Blank (avg) ( g/sec)
methyl chloroform 113.9 NR NA <0.02 <82.5
rected DRS
No >99.999
NR - No baseline run at this site.
NA - Not applicable.
NOTE:
ORE
(Feed Rate) - (Test Emission Rate)
Feed Rate
x 100%
Indicates whether or not the results of laboratory and field blank anal-
yses were subtracted from the results of sample analyses prior to calcu-
lation of ORE.
4-67
-------
during incineration of refuse is an example of this mechanism. Formation
of low molecular weight hydrocarbons.: (C-]-Cg) during combustion of coal and
heavy oil fuels is another. The secDnd general mechanism is a reaction of
free radicals or other molecular fragments produced in the high temperature
flame to produce different compounds. The products of these reactions may
have higher molecular weights than the fuel components. Benzene, polynuclear
aromatic compounds, and soot particles appear to be formed in this manner.
There are also raeans by which fuel-absent, compounds may appear in the
exhaust gas that are unrelated to the combustic-r. These include evaporation
of lubricating oils from mechanical equipment downstream of the furnace, com-
pounds that were present in the ambient air that was used in the furnace, and
compounds introduced with the compressed cir used in soot blowing. Where the
combustion sources are equipped with scrubbers the wa -.er may be a source of
organic compounds. These may exist in the fresh water supply, be introduced
by other processes where water re-use is preicti'ied, or they may have been
added in the form of fungicides or other waf.»r treatment chemicals. Com-
pounds from these sources are not PICs, but rather, are flue gas contami-
nants.
A third source of fuel-absent compounds in samples *:aken from the flue
gas is contamination of those samples during sampling or analysis. This con-
tamination can be external. For example, large amounts ?f Freon were iden-
tified in Tenax**-GC samples from one site. Apparently V-he refrigerator in
which the samples were stored was defective and Freon invaded the samples.
The contamination can also be internal. Several irvestigators have reported
that benzene appears to be a product of thermal degradation of Tenax®-GC.
Others have attributed the presence of several C^-C^ alcohols and ketones
in exhaust gas samples to the degradation of XAD resin. In addition, there
is evidence that compounds sorbed by porous polymer re-sins are not quantita-
tively removed during their preliminary cleanup. Successive cleanings re-
lease additional amounts of these compounds. Thus, it is possible that a
measured compound may be the residue of some past sample or contamination.
Ascription of the presence of a compound to its f^rmtvtion by incomplete
combustion of other compounds must be done with caution in light of the small
quantities of matter involved, the potential for contawination, and the poten-
tial for degradation of the organic polymer sorbents that are used for these
tests. The PICs reported at Sites A and C (see Table 4.17) i\re an example of
these difficulties. The limits of analytical detection for t/iese compounds
are typically 1 microgram. Even though some were detected at levels several
times their limit of detection they are, on the average, found at only a few
micrograms above this limit. Further, their concentration in tl.-e co-fired
samples is not greatly more (and in some cases less) than their concentra-
tions in the baseline run and blank samples. The difficulties with contami-
nation and high blank values are even more pronounced with tha VOST method
than they are with the MM5 procedure. Methylene chloride, for exanple, is
found in nearly every VOST tube analyzed. This compound is a common labor-
atory solvent. It is used for the MM5 extractions and clean-ups. Creation
of a protocol and an environment that would allow credible sampling and anal-
ysis of methylene chloride would require a monumental effort.
4-68
-------
Much of the PIC data (Table 4.31) are more demonstrative of the diffi-
culty of maintaining contamination-free VOST samples than they are of the
formation of PICs. Only at Site G rfere the amounts of PICs (chloroform, per-
chloroethylene, chlorobenzene, chlo :c.~c thane , and 1,2 dichloroethane) found
in the samples significantly higher than blanxs. Chloroform was found in
significant concentrations in the stack gas during all eight of the runs.
Over 80% of the total quantity of the other four compounds found was found
in one of the eight runs. The remaining 20% was distributed among five other
runs. The boiler operation, as characterized by the flue gas concentration
of combustion gases and particles, was not any different during this run than
during the other runs. The quantities of these compounds emitted were small.
If the total rate of emissions of these five compounds is added to the total
rate of emissions of the POHCs, the total hazardous compound emission rate is
less than 14 mg/sec (PICs =3.6 mg/sec; POHCs = 0.3 mg/sec). The ORE of haz-
hazardous chlorinated compounds would be 99.998%.
A different sampling and analysis procedure was used at Site L, M, and
N. At these sites Tenax® tubes that had been cleaned and sealed in the lab-
oratory were recleaned immediately prior to use in an in-field desorber/gas
chromatograph with a Hall electrical conductivity detector (GC/HECD). In
this way the lack of contamination could be verified immediately prior to the
use of a Tenax® tube. The tubes were then taken to the stack and samples ac-
cumulated. The tubes were analyzed in the field by GC/HECD within two hours
of sample collection. Blank tubes that were treated similarly (taken to the
stack, opened, closed, returned to the laboratory, and analyzed) showed in-
significant (2 to 5 ng total chlorinated compounds) amounts of contamination.
Several PICs were measured by this technique. Identification of the com-
pounds was made by matching their relative retention times on the column to
those of known compounds,.
The measured PICs at the three parametric testing sites (L, M, and N)
are summarized in Table 4.32 through 4-34. The PIC emissions were generally
on the same order of magnitude as those measured during baseline testing and
as the POHC emissions. Exceptions to this generalization were observed dur-
ing sootblowing and waste atomizer upsets when the PIC emissions were much
greater than the POHC emissions. Curiously, POHC emissions were lower than
usual during the sootblowins and aiorizer upsets.
Tables 4.32 and 4.33 give the PIC/POHC ratios calculated from the PIC
and POHC emissions for Sites L ani M. The use of PIC/POHC ratios without
their associated PIC and POHC emissions is considered a bad practice because
it can lead to misconceptions about the magnitude and risks associated with
PIC emissions. This ratio is calculated from both the PIC and POHC emission
rates and consequently, a high PIC/VOHC ratio does not necessarily indicate
a high PIC emission level. It could simply mean that the POHC level was very
low indicating a high ORE was achieved.
The emission rate of PICs was highest during periods of unstable waste
feed (low unstable flow of waste, waste startup, and waste spray directed
away from the oil flame). The amount of organic chlorine released as PICs
ranged from approximately 0.02 to 0.0002% of the POHCs fed to the boiler.
There was no discernible relationship between PIC emission rate and carbon
monoxide or oxygen concentrations, boiler load, or load changes.
4-69
-------
TABLE 4.31
REPORTED PRODUCTS OF INCOMPLETE COMBUSTION
Site
A
B
C
D
E
F
G
H
I
J
K
0
Comoound
2-chlorophenol
2 , 4-dichlor ophenol
2,4, 6- trichlorophenol
nitrobenzene
dichlorobenzene
trichlorobenzene
none reported
butylbenzylphthalate
dibutylphthalate
di e thy Iph tha late
chrysene
f luoranthene
benzo ( a ) anthrace ne
carbon tetrachloride
chloroform
1,1, 2-tricnloroe thane
methylene chloride
chloroform
methylchloroform
perchloroethylene
none reported
chloroform
chlor ome thane
chlorobenzene
1 ,2-di chlor oe thane
perchloroethylene
Test
none found
none reported
substituted phenols
none reported
Baseline
ND
ND
ND
7
ND
ND
—
22
35
4
ND
ND
ND
ND
ND
ND
ND
ND
ND
1.1
—
NR
Report Not
ND
ug/Sample
Average
Run Blank
ND
ND
ND
ND
ND
ND
—
5
5
4
ND
ND
•ND
NA
0.032
ND
NA
0.27
0.064
0.51
—
0.057
0.21
0.68
0.01
0.18
Available
ND
Average
Cc -Fire Run
4
2
6
1
2
2
—
6
8
7
2
1
1
0.07
0.08
0.007
0.12
4,8
0.49
4.3
—
21
2.4
1 .8
7.4
4.0
trace
ND - None detected.
NA - Not analyzed.
NR - No baseline run at this site.
4-70
-------
TABLE 4.32
COMPARISON OF PIC AND POHC CONCENTRATION AT SITE
2i-AS-
27-A-'
Oi>-".»"
ub-",aj
Oi"Ma>
C: -".a*
0---Ma>
Qi- K j •
0?-".J>
0£->la»
OB-M»«
0?-<4a»
09-Ma»
0^-May
05-Ma»
O'i-May
09-hjv
39-Mav
09-Ma»
09-M«v
09-Mi.
0«-Ma«
05-Ma»
OS-Mat
09-Mav
10-Mav
10-M»>
IC-Maj
ID-May
10-M*v
10-May
10-H41
10-Ma<
10-Mav
10-Mav
10-Mav
10-hav
10-Ma>-
10-Mav
10-Mav
11-Mav
11 -Ma*
l)-Mav
13-Mav
ll-Ma>
62:
IS:
43;
"jii"
lij.
164
1 Ti'.
]7j;
211=
2224
•33:
OC2C
09;:
1035
IK"
1312
141C-
150?
1702
IK:
194t
2215
215;
2321
234t
071r
094:
103C
113.-
140S
145?
164:
175C
1331
1917
201"
205-
213-
222."
2213
000'
143C
152:
2033
2118
•" . v:r
FU- v:r
Fu.. vor
FULL V33T
fj.. vs::
FL1.: V3S'
3:
3i
3?
*:•
4~
4;
4;
ru.-. V2S*
FL.. V03T
EC
53
57
60
62
66
67
6?
FULL YOST
74
76
77
&:
E3
So
S7
90
92
94
96
97
101
1C3
104
105
107
IDE
1C"?
Ill
US
122
?43
144
1 TRAP YOST
1 TRA' YOST
t •'•
a.-.;
GAi
GAS
GA1
GA:
uC
u/-—
&».'.
GA.r
G*^
GA;
GAS
GAS
GAS
GAS
GAS
GAS
GA:
GAS
GAS
GAS
GAS
5AS
GAS
GAS
GAS
01.
OH
Cin
OIL
OIL
OIL
01.
01.
01:
OIL
01.
on.
01.
01-
01.
01:
o;.
OH
01:
OIL
r 'It-' "* ' * ! n*i
.Jr.. . . i — '1.
TS: P^SL' IK"
IS: K:G.- G1* EAS1
L0« G1**
K. ATO^'IE-
L0« 5Di^/C:
LOW cr WAST: IKCREAS:
LD. 02 SASL.!^'
ii.STE 01. STAT.U"
LC™ A10' STLA« iVART,
SOOT BLO.
ii. TRAhSlEKTS
STAX- UP DUAL YOST
LOW o;
LOW 02 t GO-
LD. 0? L Gf~
G^.S SASL"L!N."
LOAO SLD^:TIO^
LOW LOAD. M3ESATI 0.-
LOW LOA3, LOW c:
LOW LOAD/02/G"*
GP^ TRANSIENT:
LOAD 1NCSEAS:
LOAT DtCRE'.Sl
SOOT3L3W LOW LOAT
LOM LOAD BASIL INI
LOU LOAD BASCLIN:
MASTE STARTS
LOW LOAD/02/GPf
LOW LOAD BASEL IK:
LOAD INCREASE
BASELIN: OIL. NDDFRAT: LOA:
HIGn G^M.'MODERATI tOAi
MDOtRATL 02/HI5H GPf
GPK TRANSIENT;
START INVERTED SPRAr
INVERTED SPRA1
INVERTED SPRA'
SOOTBLOi
«1N. USABLE ATO*
M!K. L'SABLE AT»
WASTE STARTL"
HIGH BASEL1N1
REDUCED ATOf..
SOOT3LO.
in;/:
K:
K:
c ;
H:
C.I
i.:
2Sj.C
I.';
1.1
N:
0.!
0.:
1.4:
4.6
PKA
0.2
2.0
D.3
62. C
1S.C
27. C
4.1
2S.C
9.C
15. C
2.C
3.t
2.3
l.t
3.S
1Z4.C
42.0
Z4.0
12.C
1S.C
S.£
0.3
2.C
14. C
l.i
4.5
8.5
2.i
4.:
10.C
100.1
o.-
NT-
in;"! '.r
\:
K:
K:
C.i
0.'
r, r
6.'
11 .C
K:
62. C
r. -
C.i
K:
s.<
2E.6J
73. C
20. C
2.5
12.0
ND
31. t
20. t
12. C
7.C
19. C
10. C
12. C
C.'
2.C
i.r
o.t
l.C
9S.C
41. C
3.r
2.1
4.1
2.t
0.?
l.C
2.3
lie
0.:
2.7
33. C
I,"
HI
17. L
2.1
1.4
li
Z'X
i
~
3~
;
2-
3C
7
C.3
0.3
10
52
7
7
101
22
i
3
i
^
}
I
2
t
E
E
4
<
i
1C
;
23
37
2>
B"
1M
16^
14:
i
272
SI
K
4..
1 . !
C.:
w.i
r :
I. !
153.:
2. .
£ -
J . i
ol;-
0.0;
C.2
1.7
199.5
o.o;
o.:-
2.:
3."
c.;
o.:-
o.;
0."
0.:
c.:
1.1
0.^
0.1
0.1
o.:
c.:-
C.'
ill
1.3
10.-.
2."
4.(
1£ '
it.:
12. •
31..'
L.I
2.5
67 1:
124.:
indicate MiniVOSl analysn
As chlorine t>*su
4-71
-------
TABLE 4.33
PIC AND POHC EMISSIONS AT SITE M
TEST
1 V - 1
iv -2
1V-3
1V-4
IV -5
IV- S
IV-'
1V-8
pw-9
.IV -10
NV-ll
.IV- 12
flv-13
,1V- 14
1V-15
flV-16
nv-17
nv-is
1V-19
KV-20
nv-2i
nv-22
flV-23
1V-24
1V-25
1V-26
nv-27
«V-28
1V-29
nv -30
1V-31
•IV -32
nv-33
IV- 34
"W-3 j
IV- ~g
* •J - "7
"'•-73
IV -39
CONDITIONS
Baseline, low air, nc oil
«iast = o:i lightoff, low air
Baseline, normal load V air
id- soi *e start
Low air, s » o * t
low air, atoiiiiation off
Low load, no waste gas
SnoVy, low load i air, no waste
rt a x far. capacity
flax fan capacity
Low air, ha:y
Low air, na:y
Hi gd air
Soot&l owi ng
Soot-il owi ng
Residual POHC Si PIC check, no 01
Oil 1 1 ghtof f , 1 ow 1 oad
Low load
Low load, low air
High oil f ' r ing
High oil firing, s^oky, low air
High 01 i firing, si*o:
ff H i i 'J
--
--
f'. S
•>. 9
i'< ••
3 . 1
--
111.3
--
43.5
175.7
93.0
214.4
1365.:
— -
1.2
1.6
1.0
2. 7
i.:
2. 7
1 . 5
--
__
1.7
6. 2
3.7
—
0.6
-.
0.°
0. 3
i. :
o.:
H ",
) . :
i . ;
- —
4-72
-------
TABLr. -i.33--Continueo
TEST
CONDITIONS
TOTAL Ch'.or'iated PIC/f-OHC
POHC PICs RATIO
(u g •' s > ( u g / s >
MV-40
flV-4 1
1V-42
•1V-43
MV-44
flV-45
PIV-46
nV-47
nv-48
nV-49
nv-50
rw-5i
nv-52
nv-53
nv-54
nv-55
FV-l
rV-2
FV-3
FV-4
FV-5
TV-6A
FV-6B
FV-7
FV-8
PV-9
FV-10
FV-l !
FV-12
FV-l 3
RPSI
Resi
Unst
Hi gh
Hi gh
Hi gh
dual
dual
eady
TCB
TCfr
TCP
Reduced
Reduced
POHC
PCHC
0! 1 *
, 1 OM
TCB, 1
TCB, 1
k PIC
?. PIC
lOM,
air,
OM ai
OM ai
chec*
checi:, no oil
1 OM 1 oad
S/noky
r, hazy
r, hazy
SootOl OMing
Resi
High
Hi gh
LOM
LOM
High
Hi gh
dual
ai r
ai r
air ,
air,
HCB
nCB
POHC
I PIC
check , no oil
252
339
765
79
79
266
.. .
--
.0
.6
--
.6
.6
.6
.7
--
, no spikes
, high
ha?y
smoky
I CC1
& CC1
MCB
itack
8, CCU
, high nCP I CCU
4
4
ff-B spike start
LOM
LOM
LOM
LOM
High
High
High
Resi
Use?
High
High
LOM
LOH
ai r
air,
load
load
air
CC1
CC1
dual
*dv
TCB
ai r
air,
air ,
atoniz
hazy
, 1 OM
ati on
air
, lightoH,
4
4
POHC
oil i
, lOM
, high
hazy
hi gh
J. PIC
lOM,
air,
ncB
smote
oH
Sfioxy
Check
1 OM load
smoky
t ecu
, higl- MCB I CCU
290
130
163
90
103
39
55
153
31
120
564
859
252
765
244
189
i71'
.4
.0
. 5
.6
.6
.7
—
.7
.8
. 1
.8
m 7
.8
— —
.0
.t,
• J.
.5
*
632.
642,
2394.
642.
343.
515.
57b.
716.
676.
3.
1518.
104.
975.
157.
52.
263.
89.
426.
126.
163.
180.
438.
750.
1283.
713.
47.
3038.
596.
226
f^4.
6
8
1
2
9
8
5
7
9
7
2
5
4
9
5
4
0
3
1
4
5
9
7
8
t
5
j
Q
0
9.
1.
0.
7.
9.
2.
-
-
0.
5.
1.
0.
2-
:<
-
2.
1.
5.
3.
1.
1.
-
0.
4.
2.
* .
3 .
•-
c
9
•-
7
3
0
c
:-
'-
4
t
0
e
e
3
-
3
1
8
6
3
5
-
*
0
4
* '
'.
4-73
-------
POHC Emission ;-at{- PIC rjnissions* -yl/cscr
POH;
Test
conaition
,
2
2
4
5
6
7
8
8A
9
10
12
13
14
14A
15
16
17
POHC
tyoe
—
—
^ _
_.
TCE
TCB
TCE
TCE
TCE
TCE
TCB
TCB
TCE
—
TCB
TCB
._
f eearate
ig/seci
--
~
__
_.
17.62
19.72
18.55
40.80
21.33
20.40
22.66
20.71
19.47
—
19.72
20.71
_.
iB0/CSC.>
..
—
--
__
-.
1.0 to 2.3
(1.8)
0.9
2.0 to 8.2
14.9)
0.6 to 0.8
(0.8)
NO
0.4 to 0.6
(0.6)
9.6
3.4
0.20
--
2.0
4.7
..
d>g.' sec)
..
--
--
_,.
..
17 to 38
(29)
17
32 to 130
(78)
10 to 13
(12)
NO
£ to S
(7) .
12C
59
2.8
--
32
74
..
DRi
(percent)
„
—
--
__
..
99.9994
99.99991
99.9996
99. 99998
100
99.99997
99.9995
99.9997
100
--
99.9998
99. 9996
--
Chtor-.nateC
25 to 22C
(93:
20 to 25
?0 to 36
17 to '22
(20;
36 to 200
(97)
41 13 73
(61)
11 to 19
(15)
57 to 13C
(90)
J9 to 40
(40'
£.4 to 7.C
(6.7;
5.3 to 8£
(41;
91 tc 115
(100)
12 to 18
(15)
31 to 45
(38)
127
57
7.3 to 23
(15)
29 to 1300
(450)
'r3n:r.~ 0*1 natec
4800
NA
NA
NA
78C
118C
NA
46C
NA
NA
4."
NA
NA
NA
NA
NA
NA
290C
NA — Not available. Sampling and analysis not performed or loss of data due tc
analytical problems.
ND -- Sot detected.
'RCP.A — listed Appendix VIII, Volatile and Semivolati 1 e Comoounas — nonchiorinatea are
primarily oenzene, toluene, naontnalene, ana phthalates.
''Number in pa ren tries is is tne arithmetic average of multiple test measurements.
4-74
-------
Total unbumed hydrocarbons (TUHC) was measured at seven sites; C-)-Cg, hydro-
carbons were measured at five sites. Both measurements were made at one site
- Site G. The former, Ij^C measurements, were made continuously by a FID.
Filtered stack gas is drawn through the FID which has been calibrated with a
specific hydrocarbon (methane arid propane are commonly used). The C-j -Cg pro-
cedure is a GC analysis of integrated bag samples of the stack gas. The GC
(also with an FID), is calibrated with five to seven different low molecular
weight alkanes. The results of the two methods are not directly comparable.
At Site G (Table 4.35), for example, the hydrocarbon emission rates estimatec
by the C^-Cg procedure were higher than the TUHC results by a factor of 30.
The C-) -Cg procedure is more rigorous and should produce better data.
The unburned hydrocarbon emission rates summarized for the steady opera-
tion tests are approximate (^100%), but precise enough for the present dis-
cussion. They, plus the carbon monoxide that is emitted at approximately
the same rate, are the products of incomplete combustion of the organic fuel.
The most predominate species emitted is three-carbon (propane) which is not
a significant component of No. 6 (or distillate) fuel oil. It. is noted that
the DRE of the fuel is, except for Site E, greater than 99.9%.
During the parametric testing TUHC was measured continuously by the FID
method described above. An attempt was made at Site M to determine if corre-
lations exist between PIC emissions and TUHC, and between the POHC emissions
and TUHC. There appeared to be a slight increase in PIC and POHC emissions
as the TUHC emissions increased. The correlation is weak, however (least
squares correlation coefficient of 0.21), and more data is needed to substan-
tiate th_s trend.
Chloride Emissions
The emission rate of chloride was measured during most tests by draw-
ing sample gas through liquid filled impingers. The impinger solution was
variable (there is no reference method) but should be unimportant since HC1
is very soluble in water at room temperature. The various attempts to per-
form mass balances on the chlorine in the fuel streams were unsuccessful.
In general, the measured chloride emission exceeded the measured chlorine
input by 25%. The exhaust gas chlorine mass flow rate ranged from 69 to
168% of the fuel chlorine mass flow rate.
The method used by the contractors to determine the amount of chlorine
in the fuel and waste fuel samples was ASTM-D808 or a procedure similar to
it. This procedure consists of ignition of the sample in a Parr bomb that
contains enough water to dissolve the resulting HC1. The chlorine content
of the solution was dete-mined either by classical wet chemical titration
or by specific ion electrode.
Test Methods for Evaluating Solid Waste, Physical/Chemical Methods
(SW-346) contains no procedure for measurement of the chlorine content of
oil and waste oil. Persons in the Studies and Methods Branch (EPA/OSW/DC)
recommend that Method 9020 (for organic halides in water and wastewater) be
4-75
-------
SUMMAK* Uf UNBURNSD HYDROCARBON EMISSIONS
!NONPARAMETRIC TESTING)
Site
C
D
(Hi
Load)
D
(Lo
Load)
E
F
G
I
0
Fuel Plus
Waste Emission
Feed Rate Rate
(gm/sec) (ng/sec)
525
718
346
439
240
247
264
115
<25
NM*
NM
NM
1 .19
2.95
21 .8
NM
Fuel ORE
(%)
>99.995
99.995
99.999
99.992
™£x 100
Fuel
C.005
.....
O.C05
0.001
0.008
Emission
?,ate
(mg/sec)
NM
356
300
1606
NM
91 .6
NM
<28.5
Fuel DRE
(%)
99.950
99.913
99.725
99.963
>99.975
C,-C,
6 t 1 CM
Fuel
___ _
0.05
0.09
0.27
0.04
<0.002
a Not measured.
4-76
-------
used). This procedure requires pyrolysis or organic material followed by
nucrocoulometric titration of the liberated halides. It is not known whether
the two methods, ijive comparable results, nor has it oeen demonstrated that
SW-846 Method 9020 is applicable to fuel and waste fuel samples.
A methods development effort is needed to standardize these procedures
and to develop quality assurance criteria for them. The observed weaknesses
do not preclude use of the data for this project, however. The exhaust gas
measurement of HC1 emissions gives the higher estimated emission rate sc it
is recommended that this (conservative) estimate be used to assess the im-
pacts of chloride emissions.
The chloride emission rate data, presented in Table 4.36, are based on
the exhaust gas ncasurements. They range from 0.45 to 14.53 grams/second
13.5 to 115.3 Ib/nr) depending upon the size of the boiler and the percent
chlorine in the fuel. Site M was a relatively small boiler that was burning
a high chlorine content (approximately 12.5% CD fuel. Site D was a rela-.
tively large boiler burning a low chlorine content fuel.
Particle Emissions
Measurements of the emission rates of particles were made at only five
of the sites. The measurements were all done with the MM5 procedure, which
is not an EPA reference method. The results are useful estimates of the par-
ticle emission rates but cannot be considered highly accurate. The results
are summarized for all sites (except Site N which are presented in Table
4.38) in Table 4.37. The available data are not sufficient to allow assess-
ment of the impact of co-firing hazardous waste on the emission of particles
by boilers.
Particulate emissions were measured at Site N (parametric tests) at the
inlet and outlet of a baghouse to calculate particulate collection efficiency.
Significant variation in the pariculate emissions and the baghouse collection
efficiency was observed. The fly ash control system consists of a mechanical
collector in series with a baghouse. The baghouse typically treats only 70%
of the flue gas from the collector. The other 30% is bypassed to the stack.
DurJnq the tests the bypass valve was classed resulting in baghouse filtering
of all the flue gas. This valve positioning resulted in the baghouse operat-
ing at rates in excess of the design level which could account for the var-
iation in collection efficiency and the poorer than anticipated collection
performance.
The large variation in particulate emissions reported for the baghouse
inlet is not only a result of changes in the boiler but also reflects changes
in the performance of the upstream (of the baghouse) mechanical collector.
This device is affected by changes in flue gas flow rate and particulate
loading.
Grain loadings calculated from Table 4.38 range from 0.110 to u.776
grain/dscf at the baghouse inlet and from 0.014 to 0.031 grains/ascl at "he
outlet.
4-77
-------
TABLE -1.36
SUMMARY OF CHLORIDE EMISSION R£jULTS
Site
A
B
C
D
E
p
Ga
K
I
J
K
L
M
N
0
Chloride Average
Recoverv (I0,,,,, I1T
NM
NM
NM
1.11
.8'3
1 .68
1.52
1 .27
0.69
0.886
o.sn
0.389
0.728
1 .62
Coefficient Average Chlor
of Variai.ce Dur tng Co-F ''. \
,) (percfnt
47.0
76.8
49.7
3.8
Report Not
10.9
41 .0
37.3
55.9
45.2
36.2
23.0
) gm/ ser
NM
NM
NM
14.53
6.18
2.87
0.443
Available
5.18
0.51
3.41
0.94
2.48
6.52
12.98
ide Emission
ing (as HC1)
Ib/hr
NM
NM
NM
115.2
49.1
22.8
3.52
41 .1
4.03
27.1
7.672
19.7
51 .8
103.1
NM - Not measured.
'~A Source was equipped with two-stage scrubber.
4-78
-------
TABLE 4.37
SUMMARY OF PARTICLE EMISSION RESULTS
SITES A-M, AND 0
Site
Particle Emission Rates3
gm/sec
Ib/hr
ng/J lb/10b Btu gr/dscf5
A
3
— Not Measured —
— Not Measured —
— Not Measured —
Baseline (oil)
Average Co-Fired
Baseline (oil)
Average Co-Fired (oil)
Average Co-Fired (gas)
Baseline (oil)
Average Co-r'ired
100% Waste Fuel
H
I
J
L
M
0
630
785
304
441
157
0.0807
0.408
5.00
6.22
2.41
3.50
1.25
0.641
3.24
25.0
33.2
13.9
18.3
9.4
7.61
40.77
0.058
0.077
0.032
0.042
0.022
0.0177
0.095
0.026
0.039
0.105
0.152
0.068
0.009
0.042
960
3.36 86.9 0.202
— Not Measured —
— Not Measured —
~ Not Measured —
— Not Measured —
—> Not Measured —
— Not Measured —
0.099
a These results are based upon Modified Method 5 (MM5) samples. They should
be considered to be approximate.
Corrected to 7% O2 as required by incinerator regulation.
4-79
-------
FIGURE 4.38
PARTICULATE MASS EMISSIONS AND BAGHOUSE COLLECTION EFFICIENCY
=;. ~E N
">es: 1
cons-lion "ext oescnotion
: ni or load ojselint
coal oniv
T Hi on ioatf De&eline
coal ana siuoo*
2 LO» loatr
coal oniv
4 Lo» loac
coal and siuoge
5 Hi on load
coal and ol".
6 Hi an load coa'
s'luoae snc TCI
7 HI ah load coa"
sleaar and TC£
9 Lo» load. io» ai-
coal. siuogf and T"I
12 UM Ioao. coal.
or, and TCE
13 LOB toad, Moti air
coal. oil. and TCB
14 LO- Ioao. coa1. .
oil and TCI
IS* HI on load, coal . oil
TCfr and •ctai!
16 Lt>« load, nion 11-
coa), sitnoe. TCf.
and aetals
17 Hioti load. hxs*lin»
coal an* sludo*
^
LunpcrHur1
(•r-
37t
3 Si
355
363
397
382
J86
347
3s:
370
353
375
367
37?
baonouse inle". bagiouie ou'. »»:
«s 6as ias c"?^:^r
!10- ascfm) (16/hr) (T) ,'H'- asc«n>) Ht>/nr) ..oercfn:
34. 7i 116. C 355 3B.<7 10.25 91.:
35. W 105.2 3«i' 39.2« 6.92 92. e.
25.3i 23.32 319 30.42 i.3T li.r
25.14 34.8J1 325 ?9.25 S.7c £2. i
34.66 92.75 - - -
34.9? 2«5.* —
40. 4f 265.6
27. ZC 110.00 317 26.87 3.19 97.;
26.16 35.90 324 31.13 6. Si El. '
36.% 75.66
29.92 28. OS 323 30. «E 6.57 76. t
33.55 46.73 3S6 40.0E 9.41 79.9
33.49 48. K 337 37.8* 4.6! 90. t
34.W 96. 30
all
mints wert saw I re.
4-80
-------
Metals Emissions During Co-Firing
Sampling and analysis of the metals content of the waste fuels and stack
emissions was not a primary object of this test program. It was decided that
sufficient inforiration about the behavior of various fael-contained metallic
elements in boilers already existed. Section 3 of this document contains a
discussion of the volatility and small particle enrichment of the metals in
fuels. The limited aim of metals sampling during this program was to confirm
that *:he reported partitioning was applicable to the case of hazardous waste
combustion. Thus, the emission rates of hazardous trace metals were measured
at only four of the sites.
Of the eight metals measured ±~. one of these sites (Ag, As, Ba, Cd, Cr,
Hg, Pb, and Se) only two were found in higher concentrations during the co-
firing tests than during the baseline test. Mercury was higher by 180 pico-
grams per cubic meter and lead by 80 picograms per cubic meter. Only the
concentration of selenium exceeded 1 microgram per cubic meter. All others
were less than 400 picograms per cubic meter.
Fourteen metals were measured in the teed streams and exhaust gases at
the second site. There was no statistical difference between the baseline
run stack gr»3 concentration and the co-fired run baseline concentration for
six of these. Increased stack gas concentrations for seven of the other
eight corresponded to increased fuel concentrations. The emitted concen-
trations of the more hazardous of these metals is given below.
CONCENTRATION OF HAZARDOUS METALS EMITTED FROM SITE
Exhaust Gas Concentration
Metal
Arsenic
Cadmium
Chromium
Cobalt
Mercury
Lead
Antimony
The stack gas contained approximately 70% of the chromium in the combined
fuel. The partitioning of aietals between fly ash and boiler ash, as discussed
in Section 3, varies according to the vapor pressure of the metal and its com-
pounds at the boiler flame temperature. The more volatile elements measured
during this test (As, Cd, Pb, and Sb), demonstrate this effect.
In Site N testing the oil (test condition 15} and sludye (test condition
16) were spiked with chromium and lead to investigate the partitioning of
these metals between the bottom ash and the fly ash. To qualify the portion
in the fly ash, the metals were measured in the flue gas at the baghouse in-
let and in the mechancial collector ash. Results of these measurements are
summarized below.
4-81
Co-Fired
13.7
114.5
64.4
33.3
0.9
1050.8
4.6
Baseline
3.0
71.2
46.8
18.7
0.3
455.8
1.8
-------
Bottom Ash Fly Ash Inlet/
Test Condition g/hr g/hr Outlet
Cr Pb Cr Pi- Cr Pt>
1 Baseline Coal 0.73 0.73 9.22 16.1 2.92 0.00
2 Baseline Coal and Sludge 0.75 0.66 3.99 18.54 2.94 0.00
15 Coal Co-fired with
Metal Spiked Oil 0.73 0.73 88.37 207.22 2.58 1.15
16 Coal Co-fired with Metal
Spiked Sludge 0.33 0.24 30.45 58.67 7.45 4.0~
The chromium results are considered inconclusive because of the poor
closure of the material balance (less then 40% of chromium unaccounted for).
The single run (test condition 15) in which a good mass balance closure was
obtained, indicates that lead, which is a volatile metal, can be expected to
be emitted in the flue gases during combustion of lead-containing wastes.
Regulatory Implications of the Test Burn Dsta
Test burns were conducted at several types and sizes of industrial
boilers. They were:
Size Range of Tested
Boiler Type Boilers (MMBtu/hr)
Fire Tube 8.4 - 40
Package Water Tube 10 - 110
Field Erected Water Tube 230 - 350
Converted Coal stoker 22 - 90
Coal Stoker Burning Wood Waste 10
Pulverized Coal Tangential Fired 250
Coal Stoker Burning Coal and Sludge 200
Only one boiler typ<^, the coal stoker burning wood waste, failed to destroy
at least 99.99% of the hazardous material that were co-fired. This boiler
exhibited other signs of poor combustion efficiency; e.g., high (1250 ppm)
carbon monoxide concentration in the fine gas and low (63%) thermal effi-
ciency. Additional testing is needed to demonstrate that this type of boil-
er can effectively destroy hazardous waste.
The hazardous waste fuels burned during these tests were atomized into
t.ie boiler combustion chambers, at all sites but one. This one was the coal
stoker burning wood waste. The lack of waste atomization could have contri-
buted to the poor performance of this unit. It is deemed wise to specify
that atomizing guns be provided for any hazardous wastes to be burned. In
addition, one other boiler (Site F - a multiburner water wall boiler) failed
to destroy 99.99% of the hazardous waste. This failure was attributed to
improper orientation of the waste burner gun relative to the boiler flame
that caused flame instability, some smoking, and formation of coke on the
burner. Engineering specifications on the size, location, and orientation
of the waste atomizing gun should be developed to assure adequate destruc-
tion of hazardous wastes.
4-82
-------
The percent of total fuel comprised of hazardous waste materials ranged
from a low of 8% to a high of 52% for the boilers co-firing wastes. No ef-
fect of waste fraction on ORE was observed. One boiler, a fire-tube boiler
that was specially modified for the purpose of burning hazardous waste, was
able to destroy more than 99.99% of the hazardous components while burning
100% waste material. This was a special circumstance that should not be con-
sidered representative of boilers in general.
The minimum heat content of the water materials burned auring these tests
was 3700 Btu/lb. This, co-incidentally, was approximately the heat content
of the waste that was successfully destroyed by the specially modified boiler
firing without auxiliary fuel. The data support the conclusion that adequate
ORE will be attained when the hazardous material has a heat content of 8,700
Btu/lb or more.
Attempts to correlate ORE with operating conditions (e.g., CO, 02) failed,
The ORE of the hazardous components did not vary significantly and no corre-
lations were possible even though combustion parameters were varied during
some tests over a wide range of operating conditions.
No correlation of PIC emission with carbon monoxide was observed during
the testing either. However, below a carbon monoxide concentration of about
100 ppm the PIC emissions observed during the burning of hazardous waste
were approximately the same order of magnitude as those observed during the
burning of conventional fossil fuels (baseline conditions).
The minimum boiler steam load (as a percent of boiler capacity) tested
was 25%. The maximum load tested was 100%. ORE of hazardous constituents in
excess of 99.?9% was demonstrated at all loads. The data support the conclu-
sion that a boiler may be operated within this range while maintaining 99.99%
ORE.
MODIFYING BOILER SYSTEMS TO BURN HAZARDOUS WASTE
In order to fire hazardous waste, certain modifications to the boiler
system may be necessary. Some wastes may need v.o be pretreated before they
can effectively be used as fuel. Also, burner guns and combustion controls
may have to be adopted to handle the wastes. In addition to these changes,
certain modifications may be required to coaj,?y with requirements adopted
as part of a regulatory option. For example, waute feed, oxygen and carbon
monoxide monitoring is being considered to snr>ur« that adequate destruction
of POHCs is achieved in these boilers. This -subs.ection describes the ration-
ale and the types of equipment used for these modifications.
Waste Pretreatmert
Some hazardous waste streams are pretreated to improve their suitability
for firing in industrial boilers. An OSW sponsored mail survey of approxi-
mately 250 industrial boiler operators burning hazardous wastes revealed that
four types of pretreatment are common:
4-83
-------
o Blending of the waste stream with a conventional fuel cr another waste
stream
o Heating the waste to reduce the viscosity (thermal treatment)
c- Dewatering
o Solids removal
These four types of pretreatment are described below, Based on information
obtained from a follow-up telephone survey of 11 respondents to the OSVJ sur-
vey, other types of pretreatirent are expected to be rare.
Blending
In any facility, the quantity and composition of waste produced will
likely vary with time. Blending helps to normalize fluctuations in compo-
sition, heating value, and viscosity of waste fuel flowing to the boiler.
This reduces the potential for boiler upset and assures high efficiency in
destruction of hazardous wastes. Blending may also be the only way to re-
duce emission rates of ash, metals, chloriue, and similar contaminants, to
permissible levels.
Tests for incompatibility are performed before components are mixed on a
large scale to assure that no unexpected events occur. Several possible
upsets are:
o Release of dissolved water or formation of two organic phases
o Precipitation of solid materials
o Release of gases or fumes
o Sudden release of heat
It is theoretically possible to continuously blend wastes, and run the
blend to a boiler. However, the wide variety of wastes which might be ex-
pected at any given facili.-y and the unexpected problems which might arise
from mixing varying components indicate that tank -(lending is the better
mixing technique. The technology is simple, requiring only storage tanks
equipped with top- or side-entering agitators. The wastes are pumped to
the tank and agitated for several hours. After testii.g, the mixture can
be pumped to the boiler.
At least three mixing tanks should be provided, each equipped with s.n
agitator. One tank will be filling, one filled and on standby, and one emp-
tying to the boiler, each tank si'.ed for at least eight hours of burning ca-
pacity. Additional tanks may be required to store wastes with high metals,
ash, chlorine, or other contaminant levels. The contents of these special
tanks would be olended down in the working tanks. Blending may be combined
with settling, in which case the components would be introduced into cone
bottom tanks, agitated for blending, then allowed to remain quiescent to
allow separation of the phases.
Each tank should be equipped with level indicators, and with high and
low level alarms. These will assure that the tanks are neither overfilled
nor pumped dry.
4-84
-------
Thermal Treatment
Although waste fuels may be heated to assist in breaking emulsions or
to keep mixtures in a single phase, the primary reason for heating waste fuel
is to maintain viscosities at desirable levels. Very high viscosity materials
may only be pumped with difficulty, so it is generally considered desirable
to keep viscosity below 10,000 SS'J.
Keavy fuels (e.g., No. 6 fuel oil, bunker fuel and tars) are generally
stored in insulated tanks equipped with tank heaters. This proven technology
is directly applicable to waste fuels. The storage tank is usually insulated
with a two to three inch layer of suitable insulation and heated with a side-
mounted steam heat exchanger or steam coils installed near the floor. A gear
pump, designed for the expected temperature and viscosity is usually used to
transfer the fuel. The pump and pipe are insulated and heat traced to prevent
fuel from cooling in the lines. The main fuel pipe runs past the boiler, and
recirculates to the storage tank. A branch, as short as possible, carries
fuel to each burner.
In addition, it is necessary to regulate fuel viscosity at the burner.
Although some burners are capable of handling high viscosity materials (Refer-
ence 2), maximum viscosities as low as 100 SSU have been cited by one supplier
(Reference 3) and 250 SSU by another (Reference 4). As fuel viscosity at the
burner increasas, the likelihood of incomplete combustion and stack opacity
increases, so 750 SSU has been selected as the maximum desirable viscosity
at the burner for this study. This can usually be achieved by installing
a steam or electric heat exchanger adjacent to the boiler. Figure 4.39 is
a block diagram of a typical installation at the boiler. The storage tank
and tank heater are not shown. Temperature of the oil discharged from the
boiler is usually controlled, but if characteristics are variable, the fuel
should be tested frequantly to determine the temperature which will yield the
proper viscosity.
The waste fuel may be heated to 500°F (260°C). The temperature limit
is set by the possibility of charring and cracking the liquid stream, and of
volatilizing low boiling components.
Before an unknown waste is heated, small samples should be tested to be
sure that no undesirable reactions occur. These may include:
o Separation of phases
o Coagulat:on of components
o Release of vapors
o Explosive reactions
o Increased danger of ignition during storage
o Cracking
Dewataring
Water in a liquid waste fuel impacts a boiler in three ways. Free or un-
dissolver water in a waste stream almost invariably causes burner pulsation,
and frequently leads to flame failure with concomittant burne.- shutdown. The
problem is addressed in detail in Reference 5. Water also tends to lower the
4-85
-------
I'lCiUHli 4. 19
BLOCK DIAGRAM, EQUIPMENT FOR THERMAL TREATMENT
j
03
RECYCLE
TO STORAGE
STORAGE
TANKS ARE
INSULATED
AND HEATED
o
BACK-PRESSURE
REGULATOR
STEAM
FROM
STORAGE
D
ST
O
o
UPLEX
RAINER (
*
FUEL
HEATFR
» TO
BOILER
1
~.ONDENSATE
-------
neating value of thti organic we.ste fraction since a portion of the heat gen-
erated by its combustion is consumed in vaporizing and heating trie moisture
up to the boiler discharge temperature. A third effect of water in a waste
stream is to increase the moisture content of the combustion gas which lowers
its dew point. This will increase the potential for acid corrosion in the
flue gas handling equipment.
One large facility limits waste fuels to single phase mixtures contain-
ing less than 15% water (Reference 6,. However, because few facilities will
have the sophistication to handle this level of water, this discussion is
based on a maximum water content of 5%.
Blending and decanting are the means being used for reducing the water
content of waste streams fired in industrial boilers. Blending, which was
described above, can be used to lower the water concentration when the water
and organic fraction of the waste stream are highly miscible. Decanting is
used to remove undissolved or free water from a waste stream and is described
in this subsection, it does not remove water dissolved in the organic waste
fraction which is governed by solubility limits.
Decanting is a physical separation process where the waste is allowed to
stand until the water and organic phases form two separate layers. Each layer
is then mechanically drawn off. The process may either be a batch or contin-
uous operation.
Equipment for decanting is relatively simple. One or more common storage
tanks are often used for dewatering waste streams fired in industrial boilers.
For a continuous operation requiring a long detention time to effect the sepa-
ration, an arrangment like that shown in Figure 4.40 may be used. Three tanks
are used to provide the long detention time. Each tank is alternatively used
for receiving the water laden waste; holding the wasie until the separation
is achieved; and for a transfer tank from which the dewatered waste is pumped
to the boiler. In the holding mode, the tank is used to provide a quiescent
environment for the water to separate from the organic fraction, the tank de-
tention time required depends on the properties of the waste and may be deter-
mined by periodically withdrawing samples from the tank if it is not known
from past experience. Once the separation is achieved, the water is pumped
to the sewer cr the plant's wastewater treatment facility and the holding
tank then becomes the transfer tank. The tank originally serving as the
transfer tank becomes the new receiving tank and the original receiving
tank becomes the new holding tank.
Solids Removal
Although almost any waste material that can be pumped can be used as a
waste fuel, high solids materials cause significant operating problems:
o E'>-ner gun plugging
o Abnormal burner gun fouling and abrasion
o Deposition in the fuel train
o Deposition on boiler heat transfer surfaces
o Increased particle emissions
4-87
-------
TYPICAL SOLIDS REMOVAL SYSTEM
FROM PROCESS UNIT
SLURRY TO
DISPOSAL
•
1
^
X /
J
m__ f
\
-t- — /
1
\
«
•_— •»
V-
1
.1
\
I
settling
P|
settling
Tank
^~N
Settl ing
Tank
\
,f^
-------
To minimize these deleterious effects, this discussion is base-J on a maximum
of 5% solids in the waste fuel fed to the boiler. The more common processes
used for removing solids from waste streams incinerable in industrial boilers
include settling or sedimentation, straining or screening, ana filtration.
-settling is a physical separation process whereby particles suspended in a
liquid are made to settle by mean3 of gravitational and/or ;.nertial forces
acting on both the particles suspended in the liquid and the liquid itself.
A 'anety of devices are used for this process. For the size streams incin-
erated in industrial boilers, one or more conical bottom tanks are used to
provide the time and space for the solids to settle out. These sometimes
have provisions to prevent the incoming stream from disturbing the settling
solids.
In-line strainers consist of one or more mesh baskets housed in a vessel
which nay be one of a variety of geometric configurations. When the waste is
passed through the strainer, the solid particles are trapped in the basket.
Several strainer designs are commercially available, differing maimy in the
cleaning approach. A duplex strainer which permits the cleaning of one bas-
ket while another is on-line is a common choice for this application.
Filtration is a physical process whereby the suspended particles are
separated by forcing the fluid through a porous medium. As the fluid passes
throught the porous medium, the suspended particles are trapped on the sur-
face of the medium and/or within the body of the medium itself. A wide vari-
ety of filtration equipment is commercially avalaible to meet the many types
of process requirements. A cartridge type filter may be used for solids re-
moval from hazardous waste streams incinerated in industrial boilers. This
device consists of a vessel containing one or more cartridges constructed
from fiber glass, polyethylene, or other suitable materials through which
the waste is forced to flow. The particles are collected on the cartridges.
By using a duplex vessel containing two cartridges, either side may be shut
down and serviced while the other side continues to filter.
A strainer or a filter may be used if the solids loading is sufficiently
low that the cleaning fraquency is not excessive. When the solids content is
high, a combination of either settling and straining or settling and filtering
is used. The overall process may be a batch or continuous operation. For a
continuous operation requiring a long retention time, an arrangement like that
shown in Figure 4.37 may be used. Tvis arrangement is nearly indentical witn
that described above for dewatering. In this scheme, the wastes are pumped
fr~ a processing unit into one of thre« tanks, The three tanks are alter-
nately used in a receiving wdt, a holding mode to provide a quiescent envi-
ronment for settling, and a transfer mode from which the supernatant waste
is pumped to final solids removal by eithar screening or filtering. The col-
lected solids are transferred from the bottom of the tanks into either a tank
truck, drums, or carts for transport to safe disposal.
Burner Gun Assembly and Process Control Instrumentation
Burner Gun Assemblies
Burner gun assemblies are intended to provide intimate mixing of fuel
and combustion air, assuring complete combustion using the minimum amount of
air. Assemblies are available to burn combinations of gas, liquid, and solid
4-89
-------
FIGURE 4.41
BURNER COMPONENTS
—Gas Firing-
Multi-jet
Nozzles
Vortex Chamber
Igni T.or
—Liquid Fuel Firing-
/"^TV
n»
Atomizing Sun
—Solid Fuel Firing-
f 1 '1
• ••/.
Pu1verired Coal
an^gr C par) ^nT
—Air Registers-
Single
Dual
1
4-90
-------
fuels, or to burn two gas or liquid streams with another fuel. ,.lthougki there
are differences in design details, the following, based on Zurn Industries
Equipment (Reference 7), is typical. F..gure 4.41 shows basic burner compo-
nents. The ignitor maintains a pilot flame, when necessary, to ignite t.ne
main stream of fuel. Three types of gas feed s>stems are shown, as well as
a liquid fuel atomizing gun, and a solid fuel (pulverized coal) gun. The air
registers act to control air flow and to direct the air providing intimate
contact with fuel, and shaping the flame.
Figure 4.42 shows single fuel burner assemblies, and Figure 4.43 shows
combination fuel burner assemblies.
Most boilers are equipped to burr, one fuel at a tin.e, although dual fuel
burning is not unknown. If a boiler is to be dedicated to burn relatively
high heating value waste fuels which can support combustion, modification
might be as simple as replacing a burner nozzle (and resetting fuel:_ir flow
rates). It is more common to have wastes which vary in heating value and
water content. To assure good combusticn of these wastes, it is good prac-
tice to equip the boiler with two guns, one burning conventional fuel, and
the other burning waste fuel. Some furnaces are already equipped to burn
two fuels, in which case only the burner would have to be replaced. Single
fuel boilers would require more extensive modification, which might require
replacing the entire burner assembly and modifying the end of the boiler or
the burner ports. In any case, all boiler safety controls and interlocks
must be maintained.
Burner maintenance and operating problems associated with waste combus-
tion include pulsation and flameout, poor acomization, flame instability and
smoke formation, abrasion and fouling, coking, premature ingnition, and cor-
rosion. These problems usually occur when the waste contains significant
levels of water and solids, or when the burner design is not compatible with
the ph/sical properties or combustion characteristics of the waste (Reference
5). Many of these problems can be solved by co-firing waste fuel with conven-
tional fuel, but it is critically important that each burner be suited to the
tuel being burned. A nozzle designed for high heating value gas such as LPG
might not be suitable for burning medium heating value gas such as natural
grs, and would certainly not be suitable for burning low heating value waste
gas.
Burners for liquid waste fuel are sensitive to viscosity, solids content,
and particle size of the fuel. Burners can be selected fi/om the following
five types (Reference 2).
o Rotary cup atomization
o Single-fluid pressure atotnization
o Two-fluid, low pressure air atomization
o Two-flu'.d, high pressure air atomization
o Two-fluid, high pressure steam atomization
In air or steam atomizing burners, atomization can be accomplished inter-
nally, by impinging the gas and liquid stream inside the nozzle before spray-
ing; externally, by impinging jets of gas and liquid outside the nozzle; or
by sonic means. Sonic atomizers use compressed gas to create high frequency
sound waves which are directed on the liquid stream. The liquid nozzle diam-
4-91
-------
FIGURE 4.42
TYPICAL
SINGLE FUEL BURNER ASSEMBLIES
Liquid Fuel Firing
4-92
-------
FIGURE -5.43
TYPICAL
COMBINATION FUEL BURNER ASSEMBLIES
•Gas/Liquid Fuel Firing-
single Register
Multiple Gas/
Liquid Fuel Firing
iicScrD—
Dual Register
Solid/
Liquid Fuel Firing
Pulverized Coal Center-Feed
4-93
-------
eter is relatively large, and little waste pressurization is required. Some
slurries and liquids with relatively large particles can be handled without
pluggi.ig problems.
The rotary cup consists, of an open cup mounted on a hollow shaft. The
cup is spun rapidly and liquid is admitted through the hollow shaft. A thin
film cf the liquid to be atomized is centrifugally torn from the lip of the
cup and surface tension reforms it into droplets. To achieve conically shaped
flames an annular high velocity jet of air (primary air) must be directed ax-
ially around the >;up. If too little primary air is admitted, the fuel will
impinge on the siies of the furnace. If too much primary air is admitted,
the flame will net be stable and will be blown off the cup. For fixed fir-
ing rates, the proper adjustment can be found and the unit operated for long
periods of time without cleaning. This requires little liquid pressurization
and is ideal for atomizing liquids with relatively high solids content. Burn-
er turndown is ibout 5:1 and capacities from 1 to 265 gal/hr, (1-280 cm^/s)
are available.
In single--fluid pressure atomizing nozzle burners, the liquid is given
a swirl as it passes through an orifice with internal tangential guide slots.
Moderate liqur'.d pressures of 100-150 psi provide good atomization with low to
moderate liquid viscosity. In the simplest form, the waste is fed directly
to the nozzl-i, but turndown is limited to 2:5 to 3:1 since the degree of atom-
ization drops rapidly with decrease in pressure. In a modified form involv-
ing a return flow of liquid, turndown up to 10:1 can be achieved.
When ;his type of atomization is used, secondary combustion air is gen-
erally introduced around the conical spray of droplets. Flames tend to be
short, bushy, and of low velocity. Combustion tends to be slower as only se-
condary air is supplied and a larger combustion chamber is usually required.
Typical burner capacities are in the range of 10 to 105 gal/hr. Disad-
vantages of single-fluid pressure atomization are erosion of the burner ori-
fice and a tendency toward pluggage with solids or liquid pyrolysis products,
particularly in smaller sizes.
Two-fluid atomizing nozzles may be of the low pressure or high pressure
variety, the latter being more common with high viscosity materials. In low
pressure atomizers, air from blowers at pressures from 0.5 to 5 psig is used
to aid atomization of the liquid. A viscous tar, heated to a viscosity of 75
to 90 SSU, requires air at a pressure of somewhat more than 1.5 psig, while a
low viscosity or aqueous waste can be atomized with 0.5 psig air. The waste
liquid is supplied at a pressure of 4.5-17.5 psig. Burner turndown ranges
from 3:1 up to 6:1. Atomization air required varies from 370 to 1,000 ft^/
gal of waste liquid. Less air is required as atomizing pressure is increased.
The flame is relatively short as up to 40% of the stoichiometric air may be
admixed with the liquid in atomization.
High pressure two-fluid burners require compressed air or steam at pres-
sures from 30 to 150 psig. Air consumption is from 80 to 210 ft^/gal of
waste, and steam requirements may be 2.1 to 4.2 Ib/gal with careful control
of the operation. Turndown is relatively poor (3:1 or 4:1) and considerable
energy is employed for atomization. Since only a small fraction of stoichio-
4-94
-------
TABLE 4.38
KINEMATIC VISCOSITY AND SOLIDS HANDLING LIMITATIONS
OF VARIOUS ATOMIZATION TECHNO.UES
(Reference 2)
Atomization Type
Rotary cup
Single-fluid
pressure
Maximum
Kinematic
Viscosity, Maximum Solids
SSU fiesh Size
175 to 300 35 to 100
150
Mayimusi Solids
Concentration
20%
Essentially 0
Internal low
pressure air
U30 psi)
External low
pressure air
External high
pressure air
External high
pressure steam
100
200 to 1,500
150 to 5,000
150 to 5,000
200 (depends on
nozzle ID)
100 to 200
(depends on noz-
zle ID)
100 to 200
(depends on noz-
zle ID)
Essentially 0
30% (depends on
nozzle ID)
70%
70%
4-95
-------
metric air is used for atomization, flames tend to be relatively long. The
major advantage of sncn burners is the ability to burn barely pumpable- li-
quids without further viscosity reduction. Steam atomization also tends to
reduce soot tormation with wastes that would normally burn with a smokey
f lar.ie.
Table 4.38 identifies typical kineir-atic viscosity and solids handling
limitations for the various atomization techniques. These data are based on
a survey of 14 burner manufacturers. In evaluating a specific boiler instal-
lation., however, the viscosity and solid content of the wastes should be com-
pared with manufacturer specifications for the particular burner employed.
Whatever nozzle is selected, proper operation requires that the fuel
arrive at the boiler at a reasonably constant viscosity.
Process Controls
Direct-connected controls are found on many packaged boilers. A single
actuator-operated jackshaft (mechanical linkage) is used to open and close
fuel and air valves (Reference 8). Typically, the jackshaft is positioned
proportionally to boiler pressure, and linkages from the jackshaft regulate
fuel supply valves and air dampers. If waste fuel is fed to a boiler of this
type at a constant rate, limited to about 30% of the expected maximum heat
load, and a conventional fuel is co-fired to take up the boiler swings, the
direct-connected controls can be set up to fire without reaching reducing
conditions or too much excess air in the stack gases (Refevence 9). If wide
swings in waste fuel heating value are expected, waste fual flow might be
limited to 5% of the boiler heat input, or an oxygen (e\cess air) analyzer
might be installed in the stack gas duct. The output signal from the oxygen
analyzer would be used to trim the position of the combustion air dampers,
permitting firing of higher proportions of various heating value waste fuels.
Larger boilers are usually equipped with metering combustion control
systems utilizing conventional instrumentation (Reference 10). The controls
may be set up to fire single or multiple fuels, usually based on known heat-
ing values or air:fuel ratio requirements for each fuel (Reference 11) or on
manually measured stack gas oxygen levels if waste fuel characteristics do
not vary rapidly. The trend in conventional boiler instrumentation is to-
ward a metering combustion control system with automatic adjustment of fuel:
air ratio as a function of the target oxygen set point (Reference 10). This
technique is directly applicable to burning waste fuel. Oxygen (excess air)
control is necessary if high boiler efficiency is to be maintained. Automa-
tic excess air trim systems which have been used succesfully in conventional
boilers are available from several vendors for use on spreader stoker and
packaged oil and gas boilers (Reference 12). These systems are reliable,
requiring only about one hour a week of maintenance.
Carbon monoxide control is usually installed in response to regulatory
requirements. An instrument senses carbon monoxide concentration in the
stack gas and outputs a signal which may be used to reset the oxygen control
set point. Alternatively, carbon monoxide concentration may be indicated
and oxygen level adjusted manually.
4-96
-------
Waste Feed Rate Monitoring
It may be necessary to restrict the flow of some highly toxic waste
streams to a snail fraction of the total fuel input to minimize the health
risks associated with POHCs. If such restrictions are adopted, waste feed
rate monitoring will be needed. Similarly, a trial burn may be advisab1e
to demonstrate the capability of a boiler to achieve an adopted ORE. The
quantity of POHC being fed to the boiler is needed for the ORE determina-
tion. This subsection describes some of the more useful flowmeters that
may be used for this application. Detailed information on these and other
flowmeters can be found in References 13-16.
The orifice meter, the positive displacement meter, and the flow tube
meter are well suited for measuring the flow rate of liquid hazardous wastes.
All three instruments are moderately inexpensive, are capable of the level
of accuracy needed, are of relative simple design, and can be used over a
large range of flows. The orifice meter and flow tube are differential pres-
sure type flo,' measurement devices. This type device indirectly measures
flow velocity oy measuring a differential head (pressure) across an obstruc-
tion in the flov stream which increases the velocity of the fluid, thereby
decreasing its pressure. Flow equations relate the velocity change to the
pressure change. In an orifice meter, the differential pressure between the
upstream and downstream sides of an orifice plate is measured with pressure
taps on either side of the orifice plate.
One disadvantage of the orifice meter for use in this application is
that suspended matter in the fluid may build up at the inside of the orifice
plate (which will affect its accuracy). This can be avoided by keeping the
solids content low- If it is not practical to reduce the solid content, the
flow tube may be used. The flow tube is basically a venturi without the down-
stream recovery cone. Because it does not restrict the flov. to the extent an
orifice plate does, it is applicable to streams with appreciable solids con-
tent. It has a very constant discharge coefficient and is considered to be
highly reliable. It is not as expensive as the venturi but considerably more
expensive than the orifice meter.
The positive displacement type flowmeters have one or more moving parts
positioned in the flow stream. The main devices are reciprocatory piston,
rotary piston, rotary-vane meter, and nutating disk. Of these, the nutating
disk meter is probably used in greater quantities than all the others com-
bined. This device consists of a movaole disk mounted on a concentric sphere.
The disk is contained in a working chamber with spherical sidewalli and top
and bottom surfaces that extend conically inward. The disk is restricted
from rotating about its own axis by a radial partition that extends across
the entire height of the working chamber. Each complete movement of the disk
displaces a fixed volume of liquid. The liquid enters through an inlet port
and fills the spaces above and below the disk, which fits closely and precise-
ly in the measuring chamber. The advancing volume of liquid moves the piston
in a nutating motion until the liquid discharges from the outlet port.
The vortex shedding meter works on the vortex shedding principle. In
this device, the gas stream is forced past an obstruction (shedding bar) which
sets up vortices (eddies) in the gas. These vortices cause vibrations in the
shedding bar which are proportional to the flow. These vibrations are mea-
4-97
-------
sured by a piezoelectric crystal which creates a voltage that is amplified
and transmitted to an electronic scaling nodule. Advantages of these instru-
ments inci ide accuracy, no moving parts, and relative inexpensive price.
The turbine meter is a mechanical type measurement instrument. It oper-
ates on the turbine principle; i.e., the volume is measured by the movement
of a wheel c-r turbine type of impeller. The blades of the turbine, which are
positioned within a chamber, rotate as the gas passes through them. The ro-
tor can be positioned so that it can be driven by radial or axial flow or a
combination of both. The rotor's motion can directly drive a register. This
device can be used to measure continuous high flow rates with minimum pres-
sure loss.
Oxygen and Carbon Monoxide Monitoring
Some continuous indicator of the combustion performance of a boiler burn-
ing hazardous waste is essential. Combustion performance depends on operat-
ing parameters such as temperature, feed rate of waste, and air flow rate,
but monitoring those operating parameters does not indicate what is actually
being accomplished in the boiler in terms of the waste destruction. Monitor-
ing oxygen and/or carbon monoxide levels in emissions doei> give a continuous
assessment of the effectiveness of combustion. This point is supported by
the industry trend to install 02 and/or CO monitors as part of excess air
trim systems to save fuel costs through increased combustion efficiency. CO
monitoring is being required on hazardous waste incinerators as an indicator
of the completeness of combustion.
Instrumentation for both 02 and CO monitoring of boiler flue gas is
commercially available, is considered to be reliable (Referencs 12), and
is already installed on many boilers as part of the excess air tiim system.
Both in-situ and extractive systems are being used. A variety of analyzers
are used in these monitoring systems. These are reviewed in Reference 1">
which also presents a list of vendors.
BOILER OPERATING CONDITIONS PROVIDING ACCEPTABLE DREs
Most of the testing that was undertaken during this program was accom-
plished at stable boiler operating conditions. In fact, test runs were sus-
pended or aborted if the operation of a boiler became unstable during a test
run. The data generated demonstrated conclusively that the ORE of hazardous
compounds by boilers exceeds 39.99% under stable conditions. EPA recognized
that boilers are not always operated at steady state and that the ORE micht
be less during unsteady state operations. For these reasons, special test-
ing for the purpose of determining the effects of boiler operating conditions
on the DRE of hazardous compounds were done at Site L. The objects of these
tests were to determine whether DRE fell to dangerously low levels at unusual
boiler operating conditions and to determine whether some relatively easily
measured parameter would serve as an indicator of the DRE of hazardous com-
pounds .
This section consists of two parts: a discussion of kinetic theory as
it relates to DRE of hazardous compounds In boilers, and a discussion of the
results that were were obtained during the unsteady state boiler operation
4-98
-------
tests. The discussion of the kinetic considerations was first presented in
a previous EPA report (Reference 18). The second part is based on an incom-
plete analysis of the test data and interviews with the principal investiga-
tor and others who participated in the tests.
Kinetic Considerations
Currently available kinetic data suggest that theMaal oxidation can be
empirically described as a pseudo-first-order reaction:
d£ , _kc (1 )
dt
where: C = the concentration of the compound to be oxidized
k = the pseudo-first-order rate constant
t = time
Thus, the concentration of the compound at a given time, at constant tempera-
ture is:
p
In [-1] = -kt (2)
C0
where: Co * the initial (t = 0) concentration
The rate constant can be expressed in Arrhenius form as:
k - A exp[-E/RT] (3)
where: A = the apparent Arrhenius pre-exponential frequency factor
E = the apparent activation energy of reaction
R * the universal gas constant
T = absolute temperature
The pseudo-first-order rate constants for the thermal oxidation of seve-
ral organic compounds have been measured using various adaptations of a ther-
mal destruction analytical system (TDAS). The system consists of a narrow-
bore quartz tube placed in a furnace capable of attaining, in some cases,
temperatures up to 1,200*C. A gas supply (typically air) provides a contin-
uous flow of gas through the unit. A test organic compound is introduced
into the gas stream and carried into the apparatus where it is held at a
constant high temperature for a set period of time (determined by tha gas
stream flow rate). As the vapor leaves the high-temperature quarts tube,
reactions are quenched, and the product gas is carried to an analytical de-
vice such as a FID or a gas chromatograph/mass spectrometer (GC/MS). This
analysis unit determines the final concentration (or fraction remaining) of
the organic compound unde:: investigation. Such information, developed as a
function of temperature, oan be used to determine the frequency factor and
the apparent activation energy of the pseudo-first-order rate constant des-
cribing the compound's destruction.
Rate constant parameters determined in experiments using various forms
of a TDAS could, ideally, be used to predict the degree of destruction ef-
fected by a combustion device. If the temperature-residence time environ-
-------
ment presented by the combustor were known (and assuming destruction was
kinetically controlled as described by the pseudo-first-order rate constant
and not mixing, i.e., 0? availability, controlled) ORE could be predicted
by integrating Equation 1 over the temperature-residence time profile. Of
course this is not possible since the details of the temperatare-residence
time profiles or practical combustors defy description.
However, one would expect that a relative ranking of incinerability
could be established based on the magnitude of the pseudo-first-order rate
constant at temperatures of incineration (or combustion) interest. That is,
compounds with lower rate constants at incineration temperatures should be
more thermally refractory than compound;, with higher rate constants. Corres-
pondingly, given rate constant data, it is possible to calculate from Equation
2 the temperature at which a certain degree of destruction (as measured by C/
Co) is attained for a given residence time at that temperature. For example,
the temperature required to achieve 29.99% destruction (C/CO = 0.0001) in 2
seconds could be calculated. Compounds predicted to require higher tempera-
tures would be expected to be more thermally refractory than compounds recair-
ing lower temperatures.
Table 4.22 shows just such rankings. For each compound for which rate
constant data are available pseudo-first-order rate constant calculated at
2,850°F and the length of time required to attain 99.99% ORE at that temper-
ature are shown. Species are ordered in Table 4.39 in decreasing order of
difficulty of destruction. No obvious patterns that relate the predicted de-
structability to the percent chlorination, the presence of double bonds or
heat of combustion are evident. It is interesting that methane is predicted
to be more difficult to destroy than chloroform, chlorinated biphenyls, and
some other compounds that one considers to be refractory. Chloroform is not
ignitable; which means that the heat released during its combustion is in-
sufficient to support its evaporation and molecular fractionation. This is
apparently unrelated to the speed at which it fractures when exposed to an
external heat source.
Also of note are the lengths of time that are predicted to be necessary
to attain 99.99% destruction of these compounds. Even the longest are on the
order of 1% of the average residence times of gas parcels in the flame zones
of industrial boilers. The conclusion is that, even allowing for the imper-
fect predictions that often result from theory, the destruction of hazardous
compounds by industrial boilers ought to be nearly quantitative if the flame
configuration provides an adequate time/temperature contact. The mass-weighted
DREs displayed in Table 4.15 confirm that, except in cases where the flame con-
ditions were known to be poor, the ORE of hazardous compounds exceeded 99.999%.
The destruction of methane, one of the compounds predicted to be relatively
refractory, exceeded 99.999% at those sites where natural gas fuel was burned.
4-100
-------
TABLE 4.39
FIRST ORDER REACTION RATE CONSTANT AND TIME NEEDED TO
ATTAIN 99.99% DRE FOR SELECTED COMPOUNDS AT 2850°Fa
Compound
carbon tetrachloride
hexachlorobenzene
1,2,3,5 tetrachlorobenzene
hexachloroe thane
chlorome thane
ailyl chloride
1 , 2 dichlorobenzene
methane
hexachlorobutadiene
di chlorome thane
decachlorobiphenyl
ethylene
acr/lonitrile
acrolein
1,2 dichloroe thane
toluene
chloroform
vinyl chloride
1,2,4 trichlorobenzene
ethane
2, 5, 2 ',5' tetrachlorobiphenyl
2, 5, 2', 4', 5' pentachlorobiphenyl
chlorobenzene
biphenyl
propane
benzene
First Order
Reaction Rate
Constant (k)
at 2850°F
5
3
6
6
1
1
3
1
6
7
1
1
1
1
1
4
4
1
1
1
4
5
1
1
3
2
.89
.23
.43
.87
.01
.35
.03
.08
.11
.09
.02
.26
.40
.78
.83
.39
.69
.07
.75
.81
.79
.26
.05
.40
.92
.97
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
102
103
103
103
10*
104
10*
105
105
105
106
106
106
106
106
106
106
10?
10?
107
10?
107
108
108
109
1010
Time to Attain
99.99% ORE at
2850°F (seconds)
1
2
1
1
9
6
3
8
1
1
9
7
6
5
5
2
2
8
5
5
1
1
8
6
2
3
.6
.9
.4
.3
.1
.8
.0
.5
.5
.3
.0
.3
.6
.2
.0
.1
.0
.6
.3
.1
.9
.8
.5
.5
.3
.1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
10-2
10-3
10-3
10-3
10-4
10-4
10-4
10-5
10-5
10-5
10~6
10"6
10~6
10~6
10~6
10~6
10-6
10-7
10-7
10-7
10-7
10-7
10-8
10-8
10-9
10-10
Arrhenius factors and activation energies used to compute these data
are from References 19, 20, and 21.
4-101
-------
REFERENCES
1. Protocol for the Collection and Analysis of Volatile POHCs Using VOST.
U.S. Environmental Protection Agency. EPA-6GO/8-84-007. March 1984.
2. Engineering Handbook for Hazardous Waste Incineration. U.S. Environ-
mental Protection Agency. Washington, D.C. SW 889. September 1981.
3. Telecon, Paul Schuelke. Kewanee Boiler Corporation. Hinnsdale, Illi-
nois. December 8, 1983.
4. Telecon, Bill Morton. Dorr Oliver Corporation. Harrisburg, Pennsyl-
vania. December 16, 1983.
5. Engineering Analysis of the Practice of Disposing of Hazardous Wastes
in Industrial Burners. Acurex Report. January 1982.
6. Telecon, John Brannen. Union Carbide Corporation. Taft, Louisiana.
April 27, 1984.
7. Zurn Industries, Inc. Erie, Pennsylvania. Catalogue SB-72. 1980.
8. Linking Components into a Unified System. Power. December 1967.
9. Telecon, David Schnell. Aquachem, Inc. Milwaukee, Wisconsin. Decem-
ber 8, 1983.
10. Selection Guide to Boiler Instrumentation. Chow H. Cho. Hydrocarbon
Processing. August 1981.
11. Evans, R.K. Combustion Control. Power. December 1967.
12. Letter from E.B. Rashin, Radian Corporation, to Larry G. Jones, Stan-
dards Development Branch, U.S. Environmental Protection Agency. Jan-
uary 31, 1983.
13. Fluid Meters, Their Theory and Application, 5th Ed. American Society
of Mechanical Engineers: New York. 1959.
14. Spring, L.K. Principles and Practice of Flowmeter Engineering. 9th Ed.
Plimpton Press. Norwood, Massachusetts. 1967.
15. Cheremisinoff, N.P. Applied Fluid Flow Measurement: Fundamental and
Technology^ Marcel Dekker, Inc. New York. 1979.
16. Flow: Its Measurement and Control in Science and Industry. Vol. I and
II. Instrument Society of America. Research Triangle Park, North Caro-
lina. 1971, 1982.
17. Continuous Air Pollution Source Monitoring Handbook. U.S. Environmental
Protection Agency. Cincinnati, Ohio. EPA 8i5/6-79-005. June 1979.
4-102
-------
18. Waterland, L.R. Pilot Scale Investigation of Surrogate Means of Deter-
mining POHC Destruction. Aeurex Corporation. Report FR-33-135/EE to
Chemical Manufacturers Association. July 1983.
19. Lee, K-C., et al. "Revised Model foi tne Prediction of th^ Tinie-Temper-
ature Requirements for Then.al Destruction of Dilute Organic Vapors and
Its Usage for Predicting Compound Destructability." Union Carbide Corp-
oration. South Charleston, West Virginia. Presented at the 75th Annual
Meeting of the Air Pollution Control Association, New Orleans. June 1982.
20. Preliminary data. Personal communication from E. Dellinger, University
of Dayton Research Institute, Dayton, Ohio to C.D. Wolbach, Acurex Corp-
oration. January 1983.
21. Duvall, D.S. and Rubey, W.A. Laboratory Evaluation of High Temperature
Destruction of Polychlorinated Biphenyla and Related Compounds. EPA-600/
2-77-228. Municipal Environmental Research Laboratory. Cincinnati, Ohio.
December 1977.
22. Personal communication from J. Furey, Senior Heat Transfer Engineer,
Ralph M. Parsons Co. to D.A. Falgout, iSngineering-Science. January 1984.
23. Personal communication from H. Mason, Acurex Corporation to D.A. Faloout,
Engineering-Science. August 1984.
4-103
-------
SECTION 5
CHARACTERIZATION OF WASTE STREAMS
In this section, the waste streams generated and thos«j which may be
burned in boilers will be characterized. In the discussion below, sources
of data on waste stream generation wil.1 be identified and evaluated. The
estimates of the quantity of waste burned in industrial boilers will be pro-
filed.
SOURCES OF DATA
Since the adoption of the Resource Conservation and Recovery Act (RCRA),
the U.S. Environmental Protection Agency (EPA) has undertaken a variety of
studies to quantify and characterize the amount of hazardous wastes generated
in the U.S. Although some of these studies have been overlapping, the tech-
niques used, the data sources, and the quality of the data differ from one
study to the next. The major studies reviewed for th:.s regulatory impact
analysis (RIA) include the following:
o RCRA Risk-Cost Analysis Model (W-E-T Model)
o OSW Burner Questionnaire
o National Survey of Hazardous Waste Generators and Treatment,
Storage, and Disposal Facilities Regulated Under RCRA in 1981
(Rainbow Survey)
o OSW Survey of the Synthetic Organic Chemical Industry
o MITRE Study of 104 Selected Waste Streams
Each of these data bases will be discussed in detail below. Because the RCRA
Risk-Cost Analysis Model was considered to have the best data available for
the purposes of this RIA, emphasis will be placed on it, and the other data
bases will be evaluated in comparison to it. In the discussion of each data
base, the data base will be described, its limitations identified, and the
types of information included in the dcta base explained.
RCRA Risk-Cost Analysis Model
EJ?A developed the RCRA Risk-Cost Model or Waste-Environment-Technolcgy
(tf-E-T) model to support the development of regulations authorized by RCRA.
The model contains data on 154 waste streams, i.e., the combinations of sub-
stances, such as unusable byproducts or residues generated during a manufac-
turing process, that require disposal. Each waste stream is representative
of the type of waste generated by any facility or plant of the various firms
in a particular industry. The waste stream characteristics given in the data
5-1
-------
base are for a "typical" facility. Thus, simplifying assumptions were nade
about facility size and the uniformity of waste streams withir. a -jiven indus-
try. Data collection concentrated on the following industrial sectors iden-
tiried by Standard Indvstrial Classification (SIC) code:
o SIC 25 - pulp ar»d paper
o SIC 28 - chemical industry
o SIC 33 - primary metals
o SIC 29 - petroleum and coal products
o SIC 34 - fabricated metals
Since no primary data were collected, the waste profiles were developed
from the following existing sources:
o State permit information
o RCRA Part B permit applications
o Trade associations
o CWA Development Documents
o Wastewater data from the Effluent Guidelines Division
o State studies on hazardous wastes
The data base also includes the following non-Subtitle C waste streams:
o Organic waste streams which contain more than one percent by weight
of constituents of concern and have an annual generation greater than
100 metric tons. Although some of these waste streams have conven-
tionally been coasidered to be within the jurisdiction of the Clean
Water Act, their inclusion in the data base ensures flexibility in
considering regulatory options by allowing the model to cross the
boundary between hazardous and non-hazardous properties.
o A limited number of inorganic waste streams containing heavy metals
not listed in 40 CFR 261 for the characteristic of EP toxicity.
o PCB wastes.
o Production wastes from quatenary amine (disinfectants) manufacture,
distillation bottoms from linear alkyl benzene sulfonate (detergents)
production, and off-specification commercial 2,4-dichlorophenoxyacetic
acid, salts, and esters (a herbicide). These wastes pose a hazard
based on potential ecological effects rather than effects on human
health.
o High volume utility wastes including fly ash and flue gas desulfun-
zation (FGD) sludge from conventional coal combust'on.
ihe following wastes were specifically excluded from the data base:
o Hazardous wastes from federal and other government establishments
o Discarded products, off-specification products, and containers (RCRA
class P and U wastes)
o Hazardous wastes from spills and abandoned sites
5-2
-------
o State-designated hazardous wastes
o Mining wastes
o Asbestos separator wastes frorc diaphragn ceil process chlorine
production
Wastes from small generators have been parti illy included when a large volume
of a particular waste stream with well-documented characteristics are gene-
rated by a large number of sources.
Although dat-.a gaps are frequent, the following types of information are
available from this data Dase for each waste stream:
o Concentration of toxic constituents
o Portion 3f each constituent dissolved
o Quantity of waste produced per day, per facility
o Heat value
o Ash, chlorine, and non-water percentages
o Molecular weight, vapor pressure, solubility, and bio-degradation
rate for each constituent
o Waste stream pH
o Fraction of waste stream suspended
o Amount of waste generated per year
o Number of facilities producing the waste
o Four-digit SIC codes of industries producing the waste
o Applicable treataent technologies
An EPA waste code associated with the waste stream is also included in the
data ba.»e.
Because of the methods used to develop the data base, there are three
•najor liuitations in its use to characterize waste streams burned in indus-
trial boilers. First, ihe "typical" waste streams included in the data base
may not adequately represent all segments of the industry producing that waste
because industrial processes can vary greatly within a given SIC. The data
base does not contain any information on the variability of any of the waste
characteristics. Second, not all burnable waste streams are characterized.
Only those streams which are potentially toxic are included. Third, the da-
ta used to construct the waste profiles were often incomplete. If data for
a particular stream w«re missing, constituent level estimates were based on
analogies with other streams ^nd on knowledge of the waste source.
03W Burner Questionnaire
OSW recently undertook a three phase survey of burners of hazardous
wastes. In the first phase, a comprehensive questionnaire was sent to 200
facilities believed to be burning waste as fuel. These 200 facilities were
selected from a list compiled from anecdotal information and information
contained in reports. Fach facility surveyed was called, -nd answers to the
questionnaire were taken over the telephone. Responses wer^ obtained for
only 100 facilities. In the second phase, EPA attempted to develop a list of
5-3
-------
all facilities in the U.S. burning waste as fuel- EPA's mailing list, which
included 25,000 facilities was drawn from all industry types found in the U.S.
but was wei^nted towards industries using large amounts of non-electric and
non-natural gas energy sources and those SICs, e.g., the chemical industry,
likely to burn waste as fuel. The third phase of this effort will be a sam-
pling of those facilities with a positive response in the second phase using
the questionnaire tested in the first phase.
Although the Burner Questionnaire includes a broad spectrum of informa-
tion, the following types of information are most useful for profiling burn-
able wastes streams:
o Types of wastes burned by EPA waste code
o Description of the ur.it or process producing wastes which are ourned
o Quantity of each waste burned
o Amount of waste burned as fuel which is generated off-site
o SIC code of off-site generator of wastes burned
o SIC code of facilities burning waste as fuel
o Specification limits such as minimum and maximum levels of Btu/lb,
sulfur, halogen, ash, lead, and water content for devices burning
waste as fuel
The following information was available from the second phase mailing:
o Number of facilities burning waste as fuel
o Number of facilities by SIC code burning waste as fuel
o Quantity of waste burned as fuel in 1982
o Quantity of waste burned during 1982 for each SIC code.
The results of the final Burner Questionnaire mailing as available, were used
to update or modify earlier results. However, detailed information on consti-
tuents were not available for risk analyses based on actual wastes.
Data from the initial mailing was probably biased because no effort was
made to ensure that the facilities sampled were representative of all facili-
ties burning waste as fuel. Data from the second mailing was useful primar-
ily for quantifying waste burned as fuel but not for profiling burnable waste
streams. Data from the third mailing represents a comprehensive review of
current practice.
Mail (Rainbow) Survey
Ten separate questionnaires were used to survey hazardous wast 3 manage-
ment practices. Of these questionnaires, the most useful for characterizing
burnable waste streams were the Generator Questionnaire and the Treatment,
Storage, and Disposal (TSD) General Questionnaire. A total of 11,220 Genera-
tor Questionnaires were mailed, and 9,877 responses were received. Of these,
most (7,793) were excluded from the data base for the following reasons:
o The facility had filed a Part A application for protection but did
not handle hazardous wastes.
o The facility was subject to a small quantity exemption under RCRA.
5-4
-------
o The facility was subiect to a farm exemption.
o TOO percent of the facility's hazardous w^ste was recycled, reclaimed,
or reused.
Of the 2,3-18 responses to the TSD Questionnaire, 886 were excluded from the
data base for the same reasons. Since burning waste for heat reclamation is
considered a type of reuse, all facilities surveyed which burn all of their
waste or otherwise recycle, reclaim, or reuse it were excluded. Data from
these facilities were unavailable.
For the facilities included in the data base, the following type of in-
formation is available:
o Types of waste burned by EPA waste code
o Number of facilities burning each waste type
o Quantities of wa'ste burned
o SIC codes of facilities burning waste as fuel
o Cross tabulation of any of the above information
Of course, this data base does not include information for many facilities
from which responses were received.
Major limitations to the use of this data include the following:
o Certain groups of facilities which might be burning waste as fuel
were excluded from the data base. Only those facilities which sub-
mitted either a Part A or notification form, which were not exempt
under RCRA, and which burn waste as fuel would be included. Those
generators that do not treat, store for more than 90 days, or dis-
pose of hazardous waste on-site tut do burn waste as fuel and, thus,
are exempted under Part 261.6 of RCRA were not included.
o Extrapolations from data which have be>_n highly cross-tabulated may
not be statistically sound because extrapolation factors were based
on the overall questionnaire rather than on the responses to a speci-
fic question.
o Exact quantity estimates of wastes burned as fuel cannot be obtained
because of the wording of the waste-as-fuel question.
o No data on waste stream constituents or characteristics are avail-
able.
o As discussed above, a facility was excluded if it recycled, reclaimed,
or reused (including burning) 100 percent of its hazardous wastes.
Thus, this data base was of limited usefulness to the RIA because of these
limitations.
5-5
-------
OSW Survey of the Synthetic Organic Chemical Industry
Because of riie importance of the organic chemical manufacturing indus-
try in the generation and management of hazardous wastes, OSW surveyed this
industry. This data base includes the following types of information:
o Process information including block flow diagrams
o Constituent data
o Boiler operation data if the wastes are burned in boilers
o Btu, metals, and halogen contents of wastes burned as fuel
Information in the data base is treated by ^PA as RCRA confidential because
of industry claims.
The usefulness of nhis data base was limited v.y the following factors:
o Data aggregation to avoid compromising confidentiality greatly re-
duced the specificity of the information.
o EPA waste codes were not used. Thus, comparisons with other data
bases were difficult.
o Constituent concentration data was only given by broad ranges. Sig-
nificant over- or under-estimation of POHC burning rates was possible
depending upon the assumptions used.
o The data base covered only one industry.
Because of these limitations, this data base was used solely to provide a
general framework for estimating POHC burn rates. The maximum percentage
in the range for a constituent and the average waste stream volume were used
to estimate POHC quantities. These quantities were compared with estimates
from other data bases.
MITRE Study of 104 Selected Waste Streams
A total cf 104 waste streams were identified as both incinerable and
potentially hazardous. Using OSW Background Documents and studies examining
the waste management practices of relevant industries, each of these waste
streams were characterized with the following information:
o Four-digit SIC number and EPA hazardous waste code
o Amount produced per year
o Heat content
o Listing of hazardous constituents
o Constituent levels for some of the waste streams
o Constituent level ranges for some of the waste streams
This data base was expanded later in the MITRE Incinerator Study which covered
413 hazardous waste streams incinerated at 204 facilities having operational
hazardous waste incinerators.
Major limitations on the use of this data base include the following:
5-6
-------
o Constituent data (concentrations or level ranges) are not giver for
each waste stream.
o Only 104 waste streams are characterized.
o Chlorine, ash, and water content data are usually not given.
However, this data base did provide a useful starting point for the identifi-
cation of constituents.
QUANTITY OF HAZARDOUS WASTES GENERATED
Estimates of the quantity of hazardous wastes generated in the U.S.
differ greatly. On April 27, 1984 EFA announced, as a result of a revision
to the Rainbow survey, that 264 million metric tons of hazardous wastes regu-
lated by RCRA were generated in 1981. Table 5.1 compares the waste quantity
estimates in the RCRA Risk-Cost Analysis Model with other recent surveys.
Table 5.2 summarizes the quantities of waste managed in 1981 by waste
group. The data in this table were taken from the results of the Rainbow
survey because those data represent the most recent on the quantities of
waste generated and managed in the U.S. Actually, the survey reported that
only 71.3 billion gallons were managed in 1981. The data in the table re-
flects a higher total, i.e., 82.3 billion gallons for the following reasons:
o The respondents may have interpreted the term "handled" more broadly
than intended in the survey by including wastes not managed on site,
quantities carried over from previous years as stocks in hand, or
wastes managed in RCRA exempt processes.
o Multiple hazardous waste codes were used to report some quantities.
An estimated 6.6 billion gallons reported were mixtures of hazar-
dous wastes. Solvents were sometimes reported as ignitable wastes.
In both of the cases, the waste quantities were double-counted in
Table 5.2.
CHARACTERIZATION OF WASTE STREAMS
Published data on hazardous waste str«»=, characteristics are quite lim-
ited. One of the more extensive compilations of waste characteristics i.'s
from the W-E-T Model data base and is presented in Table 5.3. This data
base and its limitations were described previously.
The roeti.l content of a hazardous waste greatly influences the risk
associated with its burning in an industrial boiler. For this reason, EPA
reviewed the previously described data bases and solicited waste analyses
results from hazardous waste burners to compile information on the metal
content of hazardous waste. The following data were reviewed and analyzed
to create Table 5.4 (References 1-11):
5-7
-------
TABLE 5.1
:OMPARI30N OF SCOPE OF RCRA RISK-COST ANALYSIS MODEL
DATA 3ASE WITH OTHER STUDIES3
Study
Risk-Cost
Analysis
Model
(1984)
Waste Quantity
(million metric
tons/year) Universe of Wastes
158 1. Potentially hazardous waste
under Subtitle C, excluding:
o Corrosive reactive and
igni table wastes
o Discarded commercial chem-
ical products
2. FGD sludge, fly ash, and
wastes selected for ecosys-
tem effects
Data Sources
EPA industry
studies, per-
mit applica-
tions, delist-
ing petitions,
state data,
trade asso-
ciations
OSW Mail
Survey
(1984)
PHB
(1980)
OTA
(1983)
Chemical
Manufacturer
Association
(CMA) (1983)
160
265
41
250
Subtitle C hazardous wastes
Subtitle C hazardous wastes
excluding commercial chemical
products discarded
1. Subtitle C hazardous waste
2. State designated hazardous
wasteb
Subtitle C hazardous wastes
in SIC 28
Responses to
Generator
Questionnaire
Responses to
TSD General
Questionnaire
EPA industry
studies
State data
Members survey
for 1981
a Taken from The RCPA Risk-Cost Analysis Model; Phase III Report submitted
to OSW by ICF, Inc., on January 13, 1984.
b Includes PCBs, waste oil, fly ash, oil field wastes, mining wastes, and
other wastes for selected states.
5-8
-------
01
I
VO
TABLE 5.2
QUANTITIES OF HAZARDOUS WASTE HANDLED tit MANAGEMENT
FACILITIES IN 1981 BY THE TYPE OF WASTE GROUPa
(billions of Gallons)
Type of Waste
Spent Halogenated and Nonhalogenated Solvents
Electroplating and Coating Wastewater Treatment Sludges
and Cyanide -Bear ing Bath Solutions and Sludges
Listed Industry Wastes from Specific Sources
Off-Specification or Discarded Commercial Chemical
Products and Manufacturing Intermediates
Acutely Hazardous Wastes
Igni table Wastes
Corrosive Wastes
Reactive Wastes
E.P. Toxic Wastes
Unspecified (Including State Regulated and Self-Defined
EPA Waste
Codes Included
in Category
F001-F005
F006-F019
K001-K106
U001-U247
P001-P123
D001
D002
D003
D004-D017
yuantity
Handled
3.2
2.6
1 3.U
2.9
0.2
1.4
33.0
3.2
11.1
11 .7
Hazardous Wastes)
TOTAL
82.3
a Taken from National Survey of Hazardous Waste Generators and Treatment, Storage, and Dis-
posal Facilities Regulated Under RCRA in 1981 submitted to OSW by Westat, Inc., in April
1984.
-------
TAM.Ii 5. 1
KCHA RISK-COST M
WASTE STREAMS UY S 1C CODES
SIC
coin i i-A
WASH IIIAIIHC
SI HI AM O.OANIIIY V At III I MAC I IK AC I I MAC I I It AC I I OH
MOMIIIH 1 10011 HI/YII) NO 01 IAC JKJ/K(;J ci ASH WAIIH &usrnii>!i>
COHlil I I III HI
COHCt NlltAI I OH
(I'l'M)
2I|91 N/A
KOIII
III .01 .il/
02. 111. III
0? Oil.III
2'j. to
I'J.'IO
!•>. in
IOIAI UUANIIIY /6.2O
2BI2 N/A (I I. III. 01 2V90
KI06 01.01.01 / 00
KO/' 01.01.22 19.00
KO/1 01.06.01 O.lll
IOIAI O.IIAHIIIY /2.2'l
u>
i 2016 N/A (11.01 tl'j /I|2.IO
O K002 KOOB 01.01.06 27.00
IOIAI 0.11 AN I I IY 769. 10
2821 N/A O2.01.02 iJ.OO
II/A 02.OI.O] '186.00
IOIAI (JUANIIIV 'J19.0O
2821 N/A III. Ol. 2(1 I6O.OO
IOIAI O.IIANHIY 160.00
2011 KOD'I Kill I KI02 Ol.lfj.OI O 4jO
IOIAI HIIANIIIY O.'jl)
20'j iOO'j 01.02.O'j 1.01
IOIAI QIIANMIY I.O1
20'jl N/A OI.OV.02 O.'jO
I,'II
/'I
I/I
6
26
II
6
II
016
616
o.o o.o o.o 0.900 ii. loom
I Oil')
OJ.OI.09
|I|U9
I Mill
I'jOOO.O O.O
I20OO.O 0.0
0.0'j O.900 O.9IHMI-OI
O.05 O.900 O.90001-01
O.O 0.0 O.O U.I'jU O.MlOOI-OI
o.o o.o o.o o.t>oo o. loom «oo
0.0 O.O O.O 0.2OO O.20OOIMMJ
u'601) 0 O.n*> O.02 O.O 0.20001-01
O.O O.O O.O O.'j?O O.I|V>OIII too
O.O O.O O.O 0.900 0.90001-01
2OOO.O O.O O..MJ 0. /'jO O.2OOOMOO
2OOO.O O.O O.O? O.90O 0.200111-01
O.O O.O 0.0 0.91>0 O.llOOOI-OI
O.O O.O O.O O./IIO 0.2900ltOO
liiooo.:; o.o o 20 o.o o.20001 ooiM).oo
20000.00
I /'
-------
TAIJLE 5. 3--Conti lined
1Mb II
SIC. SII'.tAH (JUAHIIIY
COOI 1 I'A NUIIOIKIS) NIIHIIIK (IOOO MI/Ylt) NO Ol 1 AC
1 003
1 003
206 U2HO
2061 H/A
N/A
N/A
2063 KIOJ
K iOO
N/A
N/A
H/A
N/A
N/A
H/A
N/A
KIO'I
KOI 3
KII2J
KO<>J
KO23
OJ
01
101 Al <O.
300.
303.
2IIOOO.
I'lOIIO.
1 7OOO.
?(l',dll.
o
0
o
0
o
o
o
o
.11
0
0
o
0
.0
0
0
o
o
o
o
IHACI
Cl
0.0
0.0
0.26
O.O
o.o
o.o
o.o
O.02
0.0
o.o
o.o
0.0
o.o
0.0
o.o
o.o
O.'l'l
o.o
0.0
00
1 KAC I
AMI
O
0
o.
o.
o.
o.
0
o.
o
o
II
o
0
o.
o
0
o.
II
II
o.
.02
.20
.02
.1)2
.02
02
02
,02
.()>
. OL'
C2
.02
.02
.02
.02
02
10
. 10
.1)2
.1)2
IKACI
UA 1 1 K
(I
0
O.
O.
0
O
O.
O
O
O
II
0
0
0
0
0
O.
O
O
0
. o
.0
.0
930
. 930
.930
9. '9
99 7
. 996
.009
963
^920
.923
.977
. 700
.989
. O
.0
. O
. O
IHAC: ion
SIJSI'I HOI O
O.2DOOI -Ol
O. 20001 MM)
O. IOOOI MM
O.'lllOOl -Ol
O.'lllOOl -01
O.'lOOOl -Ol
o.o
o.o
o.o
o.o
o.o
o.o
0.0
o.o
o.o
o.o
O.2OOOI -Ol
O.2OOOI -Ol
O.2OOOI -Ol
0.20001 -01
COHbl 1 Illl NI
HI IIIYI 1 IIIYI Kl IOIII
HI IIIVI 1 IIIYI Kl IOHI
2.«t-l»
CIIKOMIOM (VI I
/i HI:
CIIKdHIIIH ( VI I
corn it
NICMI
CllliOMIOM (VI )
HI IKOIII H/I HI
III H/I HI
AN II 'NI
III H?l HI
Clll OKOIil It/I HI
1.2-OICIII 01(0111 H/I NI
i.it-tHcwoiumi ii/i m
III H/I HI
HAI 1 1C AHHYDKIOI
HAI 1 1C ANIIYDKIO!
I'liniAi M: Aim YOU ioi
AH II IHI
I'll! HOI
10'lHAI DIIIYOI
I'll) HOI
I'll! HOI
NI IKOIII N/l HI
DIN/ INI
AH II INI
1.2.
-------
TABLE 5. J- -Con tint led
WASH iiiAiiNc
SIC Slid AH (JIIAHIIIY VAI III
COIII 1 I'A HIIHOIK(S» NOHItl II (IOOO HI/YK) NO OI 1 AC (KJ/KC)
KOtt-f
K022
K001
N/A
N/A
N/A
KOI 5
K09II
K02f
KO2'I
2«69 K009
KOI 1
K026
N/A
N/A
H/A
N/A
N/A
01
01
01
01
01
01
Oi
01
01
01
IOIAI 1)0 AN II
02
KOI1 KOMI O2
(12
02
0?
(12
(12
()2
.on
. 0 1100
(I
0
O
0
o
o
o
o
o
0
(I
(1
o
o
o
o
o
o
IKACI
O . 56
0.0
0.0
o.o
0. 1 1
o.o
o. 'i5
o.o
(I.O'I
o.o
O.OI
o.o
o.o
o.o
O.OI
o.o
o.o
o.o
1 KAC 1
ASH
0. IO
o . 05
0 . 02
ii. 05
O.02
O . O5
O. IO
O. IO
0.02
0. IO
0.02
11.02
O.O2
O . O2
0 . 02
0.02
0 . 02
(I 02
IKACI
UA 1 1 K
O.O
O.O
O.OOI
o.o
0.0
o.o
o.o
0.0
O.O 10
o.o
0.9/0
0.900
O.5OO
0.969
0.701
().6ll5
o . 900
0 . 90O
IKACI ION
JillSI'l NOI I)
O
O
O
O
O
O
O
o.
0
o.
0
(I
o.
o.
o.
o.
o.
o.
. 5(1(1(11 -OI
. IOOOI 100
.20001 -01
.2OOOI -O)
.2OOOI -OI
.20001 -OI
.2OOOI -OI
. IOOOI Mil
.9OOOI MIO
. IOOOI Mil
.11
.()
O
, O
.O
.2OOOI -OI
. O
0
I
(.ON SI 1 1(11 HI
HI XACIU 01(0111 H/t HI
III N/0| AJANdlltACI HI
IIIN/0(A)I'YKIHI
UIKYSltU
riii NOI
Nl 11(0111 H/l Nl
IHN/lrtl
AND (HI
I'll) HOI
CAKIIOH II IKACOI OKI 01
IOI III III
(01 01 Hi
m H/I HI
111 H/OIKICHI OKIDI
I.2.H- IIIICIIIOKOIII Htl
III XACIU 01(0111 H/l HI
IOI III HI
IIIH/YI CHI OK III!
roiHAi ic AHHYDKim
tot m ta -2.ii-oiAMifii
IOI OI HI -2.'l-l>l 1 SOCYA
I'imiAl 1C AHHYIIKIIX
t .ll-NAHIMMUnilHONt
CHI OKOIOKM
IOKHAI 1)1 IIYOI
ACHY) OHI1KII 1
CYAN III!
ACI IOHI IK II 1
I'AKAI 1)1 IIYOI
I'YKIDIHI
IOI III HI
1.2-1)1 CHI OHOI (HANI
VIHYI CHI OK IOC
HI IIIYI CHI Oil! Dl
CHI OKOI OKM
CAKItoH II IKAUHOKim
ACI )AI IXIIYDI
1 IIIYI 1 Nl OXiOl
IOI 01 HI
I'll! HOI
III H/l HI
IOI Ollll
CHI HOI
UIN/I M
JOHCI HI KAI ION
(ITHI
looooo.
IOOO.
IOOO.
IOOO
9(1(1(1
22OOO.
',(1(1(1(1
90/00(1.
M/000
IOOOOO
500000
looooo
looooo
900(11)
(,OOO
6OOO
IOOO
1 1OOO
'ill ',11110
1
IO
1600IIO.
16OOOO.
100(1
IOOO
'(00
IOOO
61KH1
165OOO.
11)000.
500.
H69O.
2 /!IO.
?000.
IOOO.
/B'> .
15(1000.
19/000.
2OOO.
IIOOH.
aooo.
nooo.
1000
5OOO.
.00
oo
oo
.00
.00
.00
00
.00
.00
.00
.00
.00
.00
(1C
. oo
.00
.00
oo
.00
.00
.00
oo
. or.
.00
.00
oo
00
.00
oo
00
.00
(10
00
oo
00
oo
oo
oo
oo
00
00
oo
00
oo
-------
TAIJ1.K 5. 3--Continned
I
u>
HASH
SIC Slid AM O.UANIIIY
COOl 1 I'A NI'HIII !<(!>) NOMIIIK (IOOO HI/YK( NO OI I AC
N/A
N/A
N/A
N/A
N/A
N/A
KOI/
KOI a
KOI 9
KO20
H/A
N/A
KOIO
KOIO
KO29
H/A
N/A
N/A
N/A
M/A
N/A
H/A
02
02
02
O2
02
01
01
01
OI
01
01
01
01
01
OI
01
01
01
01
01
Oi
01
.02
.02
.02
.02
.02
.01
.on
.on
.on
.on
.on
.on
.on
.on
.on
.on
.on
.on
.on
.on
.on
.on
. in
.16
19
.21
.22
.02
.02
.01
.on
.00
.on
. to
. 11
. in
. 16
. i/
. 19
. i'2
.21
. 2I|
.20
.26
I19d
29 1
no
1
2
2
6
10
UO
02
'j
O
no
26
1
1
2
III!
(i
22
1
n
10
00
no
61
f2
60
16
no
10
nt
UO
96
no
10
10
60
/o
/o
2(1
/O
no
00
n
10
10
6
i
i
i
•>
if
10
1
9
10
1
1
0
7
1 1
1
n
i
i
III Al IN<
VAI 01
(KJ/KC
016
noo
o
looo
fOO
2 fOO
a/oo
6
-------
TAHI.K 5. 3--Coni intied
u»
I
WASH HIAIIN'
SIC SI 1(1 AH IJIIANIIIY VAI Ul
COIU 1 I'A NOMIIIK(S) NOHIIIII ( IOOO MI/YK) NO Ol 1 AC (K.I/KC
N/A
N/A
N/A
N/A
N/A
KOI 6
Koy>
K096
N/A
N/A
K02I
KO28
IOIAI
28/9 KO'll K098
N/A
N/A
roo2
IOIAI
2892 N/A
KIMl'l KO'l6
IOIAI
2U91 KOH6
0]
01
01
01
01
01
01
01
01
01
O'j
00
qiiAN 1 1
02
02
02
01
qUANI 1
01
01
qiiAN 1 1
01
.(Hi
. O'l
.IHl
.on
.IHl
.O'J
.O'j
.O'J
.O'j
.00
.02
.02
IY
.01
.01
.0)
.01
IY
.02
.O'j
IY
.O'J
.27
.28
.29
. 11
. ,'2
.Ol
01
.O'l
.07
.00
.01
.02
.01
.90
.91
.O/
.01
.Oil
.01
1
o
1
0
190
1
1'J
1
9
111
O
O
19826
•j
'jOO
Ooo
1
1006
'10
.112
.60
. (I'l
.60
.60
-OO
.20
. 10
. 60
.27
.61
.an
.00
.00
.00
.21
.21
;• IIDOO.OO
1
2'|OII 1
10
. 70
. 70
.OO
2 I'j'jOO
2 21OCO
1 IKIIIO
2(1 J'lOOll
1 1 7000
IO '1/00
l| 6 1OO
1 OOOO
2 '.'100
I|B 2 '7000
2 0
l| 600O
2 IOOOO
'119 1'jOOO
l|19 IOOOO
101 IIOM)
lilt O
29 '100
I|'>1 O
1
O
o
o
o
II
0
o
0
o
0
o
o
o
o
o
0
o
0
0
IKACI
O.O2
11.02
O.O
O.O1
O.O)
0.87
O . II 1
0. 77
0.2*1
O.O
O.O
0.81
O.O
O.M
O.O
O. 16
O.O
O.O
O.O
IKACI
ASH
O . OO
II. OO
O.20
O.O2
O.O'.
O. IO
O. IO
O. 10
O . OO
0.02
O.O
O.2II
o . 20
0 . 20
0 . 20
0 . 02
O.O
0 . O'j
O.O'j
IKACI
WAN ft
O.IHIO
O . Of.ll
0 . 00 1
O.O
0.0
o.o
o.o
o.o
o.o
0.0
o.o
o.o
O. 7'jO
0. /OO
o. /oo
0.0
0.980
0.9'lO
0.910
IKACI ION
SUSI'I NOI 1)
0.
O.
O.
O.
O.
o.
o.
0.
o.
0.
o.
o.
0.
0.
0.
0.
0.
0.
o.
2OOOI -01
2OOOI -Ol
20001 -ill
IOOOI -III
0
IOOOI «OI
IOOOI Mil
10001 nil
IOOOI IOI
IOOOI 401
99901 400
IOOOI Mil
22OOI «OO
22001 MX)
2OOOI MIO
2OOOI -Ol
IOOOI -O2
00001 -01
60001 -Ol
COIICI NIKAI ION
CON'^M III! Ill (I'I'M)
ACKOI 1 1 N
1. 1- 01 CHI 01(01 IHl Nl
2-cni onoriii NOI
ACKOI 1 1 II
Ml III VI 1 IIIYI Kl IOIII
Al 1 Yl Al COHOI
IIIN/YI (.III OK IHl
OIMI IIIYI Al KYI AM INI
DIH/AI Mill HIM CHI OK II)
ICOO
HI XACIII 01(0111 N/l Nl
HI XACIII 01(0110 IAD II III
HI XACIII OKOI IIIANI
It IKACIII OKOI III) Ul
. . 1 .2- II IKACIII OKOI
. .2.2-11 IKACIII OKOI
,2-OICIII OKOI IIIANI
. . 1- IKK III 01(01 IIIAN
. .2- IK 1 CHI OKOI IIIAN
. .1.2-11 IKACIII OKOI
. .2.2-11 IKACIII OKOI
III XACIII 01(01 IIIANI
1. 1-1)1 CHI OKOI'KOI'AN- 1
NAIMIIHAI 1 Nl
ACI NAI'llllll III
I AS
ANI IMONY
CAKHON II IKACIII OK IOI
I.I. 1- IK K.HI OKOI IIIAN
VINYI CHI OK IOI
IOXAI-IH HI
PAKAIIIION
CHI OKOANI
CHI 01(01)1 N/l Nl
1 IAD
HAD
CHKOMIOM (VI |
II AH
IOI HI Nl
2IHIOOO
2'|OOO
liOOOO
IOOO
looo
IOOOO
IOOOO
OOOOO
KM IOOO
1 70
IIOOIK)
IIOHOII
2OOOOO
2OOOOO
I'lioooo
2/OODO
I67OOO
2/ IOOO
27 1000
70000
7OOOO
I'lOOOO
2/OOOO
IHIOOO
'IOOOO
IOOOOO
999000
looo
IOOOO
IOOOO
IOOOO
IOOOO
loooo
OOOOOO
100
jOOO
100
760
IOO
oo
oo
oo
oo
oo
01)
oo
oo
"O
00
00
00
oo
•iil
oo
00
oo
oo
oo
oo
oo
00
00
oo
oo
oo
00
00
oo
oo
oo
Ol)
00
oo
oo
oo
oo
00
00
101 Al O.IIANHIY
IO.OO
-------
TAHI.K 5. 3--Continued
bic
COOt I I'A NOHIII II (
29 1 1 N/A
N/A
KO'lO
KO'19
hOMI
KO'» 1
Kll'j2
N/A
N/A
V M/A
«*
l/i
1 O 1 Al
2992 N/A
N/A
N/A
1 0 1 Al
1000 N/A
IOIAI
1112 N/A
WASH
Slid AH
b ) NOHIII II (
Ol .(II .O'l
III .O'l.OI
O'l. (II .Ol
O'l.OI .11.'
O'l.OI .111
O'l.OI .O'l
lUl.OI . O'j
O'l.OI .(!(>
O'l.OI.O/
o'l.oi.oa
QOAN 1 (1 Y
O'l . 112 . 0 1
O'l. 112. O2
O'l. O2. (11
00AM I I I Y
OV01.02
ytlANI 1 IY
01.01 . 12
I1OANI 1 IY
IOOO HI/YK) 1,0 01 IAC
20.60 22!>
261.91) I'lll
29 / 60 160
V< . '10 1 / 1
1. /O 106
I'/6.2O 2o'»
1 . IO III1)
IMl. /'I I'll)
I.OO I'jl
'III. Ill I'll)
IO/'j.M>
69.OO 6
12.OO 6
9.OO ?
1 10. OO
•j2.OO l'>6
V.OO
ill.OO III
III Al INI.
VAI III
(H.I/Mi)
2IIO.U
looo.o
'< too.o
2OOOO.O
2OIIO.O
(looo.o
1MIO.O
1 looo.o
iMioo.o
2OOO.O
/OOil.O
.'OOOO.O
2 lOOO.O
0.0
0.0
1 KAC 1
< 1
o.o
o.o
o.o
o.o
o.o
o.o
o.o
o.o
o.o
o.o
o.o
o.o
o.o
o.o
o.o
IIIACI
Abll
0
o
o
o
o
o
o
o
(I
o
o
o
o
o
o
20
II)
10
2ii
10
211
Ml
Ml
211
M)
/()
.•o
20
o
o
IKACI
WAI in
0. /Ml
0 . IIM)
0. /90
0. 1UO
O.6IO
O.6 ill
O.'l.'O
0. I'll!
O.'llO
O.'llO
O.O
O. IOO
O.60O
o.o
0 . 900
IKACI ION
bllbl'l NDI 1) CONblllOINI
0.
1).
O.
O.
O.
O.
o.
0.
o.
o.
o.
o.
o.
o.
o.
2'lOOt 1(10 ( IIHOH HIM (VI )
IOOOI till)
loom too
2/OOI too
J'lOOl too
moot too
MIOOI IOO
II'IOIM too
21OOI tOO
'j2OOl IOO
IAD
.IIIIOH 1 OH (VI)
(All
.IIKOMIOH ( VI )
1 Al)
:IIKOHIOM ( vi )
1 Al)
IIKOMIOM ( VI )
1 Al)
IICOMKIM (VI |
IAD
•.IIKOMIIIM ( VI )
IAD
.IIKOMIOH ( VI )
1 All
IIIUHIIOM (VII
IAD
:ill(OMIOH (VI |
IAD
/OOOI toil I 1 Al)
' III N/0(A)ANIIIKACINI
1)1 N/0( AJI'YItl HI
IOOOI Mill 1 ( Al)
I)IN/0(A>ANIIIHA<;|N|
III N/0( A jl'YKI Nl
IOOOI IOO 1 1 Al)
III N/0( A)ANIIII0(.. -01 ( IWOMIOM (VI )
1 1 Al)
<.:;-j«;i NIKAI u
( ri'M)
1 1.00
9 .110
'( Oil
1 .00
0 IK,
1 .00
6.00
1 .00
10.00
.'(1 .00
l| . OO
/•j.OII
1 .11(1
6 .IO.OO
:• .00
>l .00
1 .00
1 .(Mi
20.00
•jj .00
'lOIMIO.OO
loooo. oo
lOOOO. 00
loooo. oo
10000.00
loooo. oo
20000.00
IIIKKKI. no
loooo. oo
HUH) 111)
too oo
100.00
IMKOO
10000.00
IMl. 00
28. OO
69 . 00
-------
TAfll.E 5. 3--Conti uuo.i
WASH III A Mill.
SIC b I Id AM OOANIIIY VAI 1)1 I HAM I MAC I IHACI IKACIIIIN
(.01)1 I TA NUIIIIl I((S| HUHIIIK ( IOOO MI/Ylt) NOOI IA<; (KI/KI.) i:l A:!ll WA 11 M SUSI'I HIM II
CONS I I III! Ill
N/A
KH6I
K(i<)/
KObl
III .III . I J Ml. 11(1
oi.ai.^i in. oo
o I. n. '.o i ,Mr.n no
oi.oi.oi /<• oo
01 IK, (!.• j'j.iio
O'j.Ol.OI »|61.<>O
IOIAI OUAMIIIY }6aV'IO
ui
i Jill N/A 01.01. I/ 1? Mil
N/A 0->.lll Bl 60
N/A (l'> 01 111 ;?..'(!
N/A 05.01 IMl 8'j.l|0
IOIAI quAHlllY ?/ii.ou
IIL'O N/A 0'. (II ()•> ;MI'J.I(I
N/A 0*>.OI.06 16'JI'J 10
IOIAI cjoAiiiilv I'jo'jii.;-o
1111 N/A Ol .Ol . Ml 6'IO.IMI
IOIAI UUANIIIY 6*10.00
jj)? N/A Ol.Ol.li!! 'I.' (Ill
IOIAI OIIANIIIY i.'.oo
1111 N/A 01 (II (IV I I 00
N/A 01.01 10 Id 00
(>l
16
u,
/O
II
^000
.'0(10
I/
O.O O.O
O.O O.O
lll'IKI (I O.O
O.O O.O
O.O (I (I
O.O O.O
0.0 0 (I
O.O
0.0
O. 9 I'I O. JOOOI -Ol
I!. 000 O. ISOOI «00
o.o o.o it it o.aoo o.i'oooi-o;*
O.O O.O 0.0 O.'JVtO O.liOOOI-OI
o.;'o o.o
O O O.O
o y/iiui «oo
o. 10001 MII
(I.O O.O O.O O./MI (I .'HIIOI I'M)
o.o
41. O
0.0
o.o
o.o
o.o
O. IdlKI! MII
O. Hindi Mil
O. IOIIi.1 MII
O.O O.O O.O O.O O. Illlllll MII
O.O O.O O.O O.O O. U COl M.M
O.O O.O 0.0 II.BOO 0. I'XKII MIO
O.O O.O O.O O.BOO O. IDdiM MIO
o.o o.o o.oo.oo
moo. oo
t"j;'o.i»o
IO.OO
ioo.no
/'mo oo
IIHIII . oo
iloo.oo
'lOiiuo oo
I /(IO.OO
looo.oo
'jllO.OO
IMIIKI. OO
IIIIKI IIO
Ml .OO
MIO.OO
,' Kid on
K>oo.oo
6'tO.OO
ion.oo
loom) on
60.00
•j/O (10
HOOII (III
ll'l .00
M.iHlll. OO
':;*oo.oo
I'll .00
.'(tOOO. 'Id
-------
TAHI.I-: b. 3--riiiiliniic ii. linn o. loooi MID CYAinui
O.O O.O O.O ll.9>lll O. ',11(1111 -III
0.0 0.0 0.0 O.VVj 0.^(1(1(11-0,'
o.o o.o o.o o.yyo o.i'oooi-o;*
(OHI I NIK A I ION
|C CM)
o.o o.o o.o
o.o
COCCI K
ClllidMIIIM (VI I
CADMIUM
I IAD
NM Kt I
CAIIMIIIM
COCCI II
CYAN I III
CIIIIOMItIM ( VI )
NICKI I
COCCI K
CYAN I ID
I I OOKIIII ( I I IHIKIDI S|
NICKI I
COCCI K
CYAN ID)
CMKOMIIIM ( VI )
NICKI I
/INC
Mid. 00
III/Mill. OO
ll'i.OO
/ . OO
I 1000 (Id
VIId. (Id
I1,II.Ill)
l.'dddd 00
C'OOO.OO
IKIOOO.OO
;'iioo.oo
joiioo. oo
MIIIOII .OO
moo oo
611(1.00
JMI.Od
'ioo.no
IOO (III
I'll).OO
(').(IU
i /d.mi
I Mid. (Id
.''1.00
6 .oil
9.00
i'l.OO
6.00
•)(l. 00
I /.OO
;• .110
i .00
IOIAI d'lAHIIIY
IVIIO'J III
-------
TAUI.K 5.3--CoriliMOo
IOIAI QUAN 1 1
01
01
IOIAI (JOAN II
Ol
01
01
01
01
01
01
01
01
01
01
01
01
.III
IV
.01
IV
.02.
.02
IV
.01
(II
.01
.01
.01.
.01
.01.
Ol
0?
. 02
.01.
.01.
. 06 .
. II
.01
.III
.O'l
.(12 2
(II
02
01
O'l
,0'j
.06
00
(12
01
01
90
01
IMI
I'jll
1
1
1(1
9
ill
I|'|OO
2V
'l*>
10.
19
1
20
O
?
9.
29.
/I.
"1
.00
.OO
. 10
. 10
li-'i
.'t\
.V,
.00
. VI
2/l(i
ll | ')ft
/222O
196
6161
III Al INI
VAI ill
II
0
I/ 100
a2oo
0.
11200.
1 1200
12100.
I2600.
6i)OO .
'jlOO,
«*'l /OO.
•J600.
'|2OO.
2'jlOO.
IIODO.
21OOO.
. 0
. o
.11
.0
. o
.11
0
.11
0
.0
.0
, 0
(I
o
o
o
U
IKACI
Cl
o.o
o.o
O . '> 1
0.21
0.0
(I. 'i9
O.'IH
O.'jO
ll. 5 1
O.H8
0 . 6 /
o.o
0. 16
O.I/
0.0
o.o
O . l>6
IKACI
ASH
O.O
O.O
(I. •'!!
0.60
0.0
O. II)
O. Ill
0 III
O III
O.O2
o.o;'
0.02
0 . 20
0.20
0 . 20
O. 2(1
O.O2
IKACI
MAII II
O.U90
OO
O.O
O.O
O.'jIlO
O. 1110
O. 100
0. 100
0. IOO
0.0
0.0
0.0
0.0
0.0
O.IVjO
0 . 6(10
o.o
IKACI ION
MISH NO) 0
O
O
(I
O
O
O
O
O.
0.
0
O.
0.
O.
o.
o.
o.
0.
. mom
. loom
.20001
. loom
. Vjiiol
. IOIIOI
. loom
, loool
10001
.20001
.20001
.2OOIII
.20001
.20001
20001
2OOIII
0
Mill
Mil
MIO
Mil
MIO
Mill
MIO
MIO
MIO
-Ol
-01
-01
"OO
<00
MIO
MIO
CONbl 1 Mil Nl
1 1 All
CADMIUM
1 IAD
Ml HI, IIHV
II IKACIII OKOI 101
II IKA.UIIOUO! till
AIC.I NIC
IIAKMIM
(.ADM Mill
1 IAO
IK II. Ill OKOI III! Nl
Nl
III
I.I.I- III ICIII OKOI IKAN
O ICIII OKOMI MIAN)
II IKACIII OIIOI III!
Nl
1.1. 1 -III ICIII 01(01 1 UAH
DM III OKOMI IIIANI
Nl IKOIIIN/IHI
I.I. 1- IK ICIII OKOI MIAN
IHCIII OKOMI IIIANI
101 01 Nl
Ml IIIVl 1 MIVI M
( IIKOMHIM (VI)
1 IAO
HI KCDKY
CIIKOMMIH (VI )
101 01 Nl
IT.H- r.»j'l
IONI
l.2,l|- 11)11. Ill OIIOIII H/l
N/A
Wj
.Ol.
01 'II
II2'UI.
(10
HO'IHO
0.
o
O.O
O.O
0.0
0.
10001
Mil
Af(SI NIC
IIAKMIM
CIIKOHIIJM ( VI )
1 IAO
(01(1,1 II IK A I ION
I'looo.oo
IOOOO.OO
IOOOO.OO
IOOOO.00
(lOOIIOO. OO
2SOO(iO.O()
MI.OO
260.OO
I 1.00
MI'), on
60OOOO.OO
6(10000.00
6OOOOO.OO
600000.00
ftOOOIIO.OO
liooooo. no
'JIIOIIOO.OO
2OOOOO.OO
:'ooooo.oo
loooo.00
loooo.oo
')<, .(III
l|(). 01)
12.00
lilOO.OO
loooo.00
•jOOOOO.OI)
'XIOIIOO.OO
'IO.(K)
IIIOO.O')
I'll).(10
90.00
IOIAI 1J1IANIIIV
-------
TABLE 5.4
METAL LEVELS IN HAZARDOUS WASTE FUELS IN PPM (References 1-11)
Ash As Ba Cd Cr Pb Ni Hg
No. of Samples 40 186a 159s 191a 198 199 169 175
90th Percentile 20 18a 2513 1Oa 296 572 25 <1
50th Percentile 2.82 .5 <5 <.5 <5 <8 <2 <.06
10th Percentile .05 .02 <.2 <.1 <.2 <.5 <.2 <.01
Note: "Less than" values were included at the detection limit when
determining means and percentiles. Where "less than" values
contribute significantly to the sum of all values, the means
and percentiles are presented as "less than" values.
a Some used oil data was included in the data base for 90% worst
As, Ba, and Cd.
5-19
-------
o Thirty samples from Reference 1 (those wastes with heating values
qreater than 5,000 Btu/lb).
o Cix samples from kiln test burn reports, Sites B-G (References 2-7).
o Twenty-two data points presented in the Mitre reports for spent
flammable solvents and chlorinated solvents (References 8-9). Given
that a low/high range was presented for metals levels in many waste
streams, such wastes were included in the data bz.se as two wastes,
one with the low metals level and one with the high metals level;
thus, the 22 data points were derived from 13 waste streams as charac-
terized by Mitre.
o One hundred forty-two samples from a company participating in the
Keystone workshop (Reference 10), burning 100% waste in lightweight
aggregate kilns.
o Five samples from ICF (W-E-T Model) data base of wastes as generated
(Reference 11).
o Other waste streams characterized for metals by Mitre and ICF (W-E-T
Model) either had low heating value (less than 5,000 Btu/lb) and are
not currently burned as fuels, or were waste streams known not to be
burned as fuels (e.g., petroleum refinery wastes typically sent to
petroleum cokers or land treatment).
Some data on arsenic, barium, and cadmium levels in waste oil were included
in the data base used to compute the 90th percentile values.
QUANTITY OF HAZARDOUS WASTE BURNED IN INDUSTRIAL BOILERS
Basic data for quantifying the amount of hazardous waste burned in indus-
trial boilers were available from the OSW Burner Questionnaire. The limita-
tions of this data base v1 xch were discussed above must be remembered. Table
5.5 shows the breakdown of hazardous wastes as to the amount burned versus the
total generated. Table 5.6 provides estimates of distribution waste burned by
SIC code. These data represent the latest and best estimate of the quantity
of hazardous waste being burned.~~~*
5-20
-------
TABLE 5.5
BURNABLE HAZARDOUS WASTE UNIVERSE
Item
Waste Quantity (million
metric tons per year)
Total Hazardous Waste Generated
Total Burnable Hazardous Waste
Amount Burned:
Incinerators
Boilers <50 Million Btu/hr
Boilers 50-100 Million Btu/hr
Boilers >100 Million Btu/hr
Kilns, Etc.
Total
160
25
2 -
1
2
0.35
6
5-21
-------
TABLE 5.6
ESTIMATES OF WASTE-DERIVED FUEL BY SICa
SIC Code
20
22
24
25
26
27
28
29
30
31
32
33
34
35
36
37
39
40
42
49
51
72
75
TOTAL
Industry
Food Products
Textile Mill Production
Lumber and Wood
?urniture and Fixtures
Paper and Allied Production
Printing and Publishing
Chemical and Allied
Petroleum and Related
Rubber and Plastics
Leather and Leather Production
Stone, Clay, Glass, Concrete
Primary Metals
Fabricated Metals
Non-Electric Machinery
Electrical Equipment
Transportation Equipment
Miscellaneous Manufacturing
Railroad Transportation
Motor Freight Transport
Electric, Gas, Sanitation Serv.
Wholesale Trade-Nondur .
Personal Services
Auto Repair and Services
Quantity
(million
gal/yr;
3.4
0.2
100.9
113.1
213.9
173.2
375.8
84.9
30.9
0.03
9.0
73.4
1 .1
0.4
0.01
0.6
0.2
0.004
0.02
0.07
0.1
0.3
0.03
1009.4
Number of
Facilities
104
135
273
118
1 17
118
320
78
76
2
26
145
47
162
13
131
27
7
14
43
7
477
198
2639
Data from EPA's Burner Questionnaire.
5-22
-------
REFERF.NCES
1. Hazardous Waste and Virgin Oil Assessment of Baseline Metal Content.
Versar Inc. April 1986. Draft Report.
2. Evaluation of Waste Combustion in Cement Kilns at General Portland, Inc.,
Paulding, Ohio. Research Triangle Institute and Engineering-Science, Inc.
March 1984.
3. Day, D. R. and Cox, L. A. (Monsanto Research Corporation). Evaluation of
Hazardous Waste Incineration in an Aggregate Kiln, Florida Solite Corporation.
May 30, 1984.
4. Day, D. R. and Cox, L. A. (Monsanto Research Corporation). Evaluation of
Hazardous Waste Incineration in Lime Kilns at Rockwell Lime Company.
October 1983.
5. Smith, G. E. and Rom, J. J. (Systech Corporation). Hazardous Waste Combus-
tion in a Dry Process Cement Kiln. September 1982.
6. Evaluation of Haste Combustion in a Dry Process Cement Kiln at Lone Star
Industries, Oglesby, Illinois. Research Triangle Institute and Engineering-
Science, Inc. April 1984.
.'. Peters, J. A. et al. (Monsanto Research Corporation). Evaluation of
Hazardous Haste Incineration in Cement Kilns at San Juan Cement Company.
August 1983.
8. Composition of Selected Hazardous Haste Streams. Mitre Corporation.
November 1981.
9. Composition of Hazardous Haste Streams Currently Incinerated. Mitre
Corporation. 1983.
10. Burning of Hazardous Waste in Industrial Boilers and Furnaces. A Keystone
Center Workshop. Washington, D.C. February 11, 1986.
11. RCRA Risk - Cost Analysis Model Haste Stream Data Base. SCS Engineers.
July 1984.
5-23
-------
SECTION 6
COST ELEMENTS FOR REGULATORY ANALYSIS
A key element of the RIA is an assessment of the potential economic
impact of the various regulatory options being considered. This section
identifies the major cost factors being considered in the assessment and
provides the cost data that are not being obtained from other sources.
Regulating the burning of hazardous waste in industrial boilers could
potentially have an economic impact on those being regulated as well as the
regulatory agencies. This section is limited to the costs associated with
burning the wastes. Potential cost to regulatory agencies are addressed in
a separate document (Reference 1).
The economic impact on those burning hazardous waste ir. industrial toil-
ers is being analyzed by EPA in terms of how the net fuel and waste disposal
savings is altered by the various regulatory options. This net savings is
the difference in: (1) the credits associated with the fuel replacement and
the elimination of alternative waste disposal waste; and (2) the increased
capital and operating costs associated with firing waste above that for fir-
ing conventional fuels. Therefore, to analyze the economic impact of a re-
gulatory option, cost data for the significant elements of these credits and
incremental costs are needed. The choice of cost elements will depend, to a
large extent, on the level of detail specified in the analysis. These data
are being obtained from a number of sources. The items being provided in this
document fall into three major categories: (1) conventional fuel prices, (2)
costs to modify the boiler system to fire the waste, and (3) the major operat-
ing and maintenance costs associated with burning wastes.
CONVENTIONAL FUEL PRICES
The prices of conventional fuels used by industrial boilers are used in
the analysis of fuel replacement credits. Two sets of prices are presented
here. The first set is for 1982, which is the year covered by the OSVJ Burn-
er Questionnaire survey of waste fuel users. The results of this survey are
also being used in the economic impact anlaysis. Having the fuel prices in-
dex for the period covered by the survey facilitates the economic impact anal-
ysis. The 1982 prices are:
6-1
-------
o Natural gas - S3.63 per million Btu
o Distillate Oil - $7.24 per million Btu
o Residual Oil - $4.62 per million Btu
o Coal - $1 .09 per million Btu
The above prices were developed from data taken from publications by
the Energy Information Administraiton which is the branch of the Department
of Energy responsible for collecting, compiling, and disseminating data on
U.S. energy cost and usage. Natural gas and oil prices were developed from
data provided in Reference 2 while coal prices are based on data from Refer-
ence 3. These prices are representative of the national average values paid
by industrial users in 1982.
One aspect of the RIA is to estimate the impact of the regulation on
future operations of the waste burner. Projected fuel prices for the period
covered by the analysis are therefore needed. Projected fuel prices have al-
ready been generated for use in developing New Source Performance Standards
(NSPS) for industrial boilers. Fuel prices projected for the period 1985
through 2010, are listed in Tables 6.1 through 6.4. These projections were
taken from Reference 4 and are currently being used by the EPA Office of Air
Quality Planning and Standards (OAQPS) Branch for its economic modeling.
Updates to these costs are made periodically. Mr. Robert Short is the EPA
Project Engineer responsible for the updates. He is located at Research
Triangle Park, North Carolina.
BOILER SYSTEM MODIFICATION COSTS
When fossil fuel-fired boilers are used to burn hazardous wastes, capital
expenditures may be required for a number of system modifications, including:
waste pretreataient, storage and handling facilities; boiler modifications;
and combustion control instrumentation modification. Additional expenditures
may be incurred as a result of regulatory requirements. For instance, re-
strictions on emissions of particulate matter, metals, and hydrochloric acid
may require additional air control equipment or modifications to existing
devices. Further capital outlays will result if waste feed metering and
monitoring of the boiler temperature, combustion gas 02 and CO levels are
required.
Cost data for the above elements have been aggregated into three groups:
o Waste pretreatment costs
o Boiler modifications costs
o Air pollution controj. device costs
The cost data presented below were developed from published data, vendor
information and engineering judgement. The following discusses the various
elements comprising these three groups with one exception. Cost data for
waste storage and handling facilities have been generated by the EPA for its
economic modeling efforts and are not presented here. However, these data,
which are documented in Reference 5, are being used in impact analysis.
6-2
-------
TABLE 6.1
PROJECTIONS OF REGIONAL INDUSTRIAL NA7'JRAL GAS PRICES*
(1982 $ per million Btu)
Demand Region 1985 1990 1995 2000 2005 2010
1. New England 6.09 7.16 8.50 11.47 14.67 16.50
2. New York/New Jersey 4.90 5.41 6.42 8.67 11.09 12.47
3. Middle Atlantic 4.29 4.59 5.45 7.35 9.40 10.57
4. South Atlantic 4.82 5.70 6.76 9.13 11.67 13.13
5. Midwest 4.14 4.90 5.81 7.85 10.03 11.29
6. Southwest 4.27 4.63 5.50 7.42 9.49 10.67
7. Central 3.77 4.55 5.40 7.30 9.33 10.49
8. North Central 4.22 4.79 5.68 7.67 9.81 11.04
9. West 4.69 5.44 6.45 8.71 11.14 12.53
10. Northwest 4.90 5.08 6.03 8.15 10.41 11.72
a Taken from Reference 4.
6-3
-------
TABLE 6.2
PROJECTIONS OF REGIONAL INDUSTRIAL RESIDUAL FUEL OIL PRICESa'b
(1982 S per mil.Uon Btu)
1 .
2.
T ^
4.
5.
6.
7.
8.
9.
10.
Demand Region
New England
New York/New Jersey
Middle Atlantic
South Atlantic
Midwest
Southwest
Central
North Central
West
Northwest
1985
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
09
02
01
20
34
43
40
25
49
31
1990
5.
5.
5.
5.
5.
5.
5.
5.
5.
5.
15
09
09
32
44
63
51
40
71
49
1995
7.32
7.23
7.22
7.56
7.73
8.00
7.83
7.67
8.10
7.79
2000
8.
8.
8.
9.
9.
9.
9.
9.
9.
9.
92
82
81
21
42
75
54
35
88
50
200D
1 1
10
10
11
11
12
11
11
12
11
.11
.98
.97
.48
.73
.14
.89
.65
.31
.84
2010
12.78
12.63
12.62
13.20
13.50
13.97
13.67
13.40
14.16
13.62
a 1.6% sulfur.
0.3% sulfur » 1.6% sulfur + 50.68/MMBtu.
0.8% sulfur - 1.6% sulfur + $0.35/MMBtu.
3.0% sulfur - 1.6% sulfur - $0.45/MMBtu.
b Taken from Reference 4.
6-4
-------
TABLE 6.3
PROJECTIONS OF REGIONAL INDUSTRIAL DISTILLATE FUEL OIL PRICES3
(1982 $ per million Btu)
Demand Region 1985 1990 1995 2000 2005 2010
1. New England 6.33 7.28 10.02 12.07 14.84 16.96
2. New York/New Jersey 6.27 7.22 9.93 11.97 14.71 16.82
3. Middle Atlantic 6.24 7.18 9.88 11.90 14.63 16.72
4. South Atlantic 6.08 7.03 9.68 11.66 14.34 16.39
5. Midwest 6.20 7.13 9.82 11.83 14.54 16.62
6. Southwest 6.08 7.06 9.72 11.71 14.40 16.45
7. Central 6.15 7.08 9.74 11.73 14.42 16.49
8. North Central 5.98 6.94 9.55 11.50 14.14 16.16
9. West 6.04 7.01 9.65 11.63 14.29 16.34
10. Northwest 6.04 7.01 9.65 11.63 14.29 16.34
a Taken from Reference 4.
6-5
-------
TABLE 6.4
DELIVERED INDUSTRIAL COAL PRICE FCRECA£Ta'b
(January 1983 5 per million Btu)
Demand Region Coal Type
1 . New England Bituminous
2. New York /New Jersey Bituminous
3. Middle Atlantic Bituminous
4. South Atlantic Bituminous
5. Midwest Bituminous
Sub-
Bituminous
Sulfur Content
( Ib SO?/MMBtu )
<0.80
0.80 - 1 .08
1 .08 - 1 .67
1 .67 - 2.50
2.50 - 3.33
3.33 - 5.00
>5.00
<0.80
0.80 - 1 .08
1 .08 - 1 .67
1 .67 - 2.50
2.50 - 3.33
3.33 - 5.00
>5.00
<0.80
0.80 - 1 .08
1 .08 - 1 .67
1.67 - 2.50
2.50 - 3.33
3.33 - 5.00
>5.00
<0.80
0.80 - 1 .08
1 .08 - 1 .67
1.67 - 2.50
2.50 - 3.33
3.33 - 5.00
>5.00
<0.80
0.80 - 1 .08
1 .08 - 1 .67
1.67 - 2.50
2.50 - 3.33
3.33 - 5.00
>5.CO
<0.80
0.80 - 1 .08
1985
3.53
3.42
3.30
3.25
3.19
2.68
2.94
3.34
3.20
3.10
2.94
2.87
2.39
2.60
2.93
2.77
2.60
2.40
2.41
1.98
1.77
2.92
2.74
2.55
2.30
2.64
2.09
2.52
3.13
2.94
3.00
2.70
2.59
2.18
2.23
2.63
2.63
1990
3.77
3.67
3.67
3.68
3.50
3.98
3.21
3.52
3.41
3.42
3.22
3.14
2.71
2.83
3.20
3.05
2.96
2.74
2.70
2.50
2.14
3.32
3.12
2.60
2.80
2.53
2.69
2.64
3.39
3.22
3.14
2.97
2.91
2.46
2.42
2.84
2.84
1995
3.93
3.80
3.88
3.35
3.64
3.29
3.41
3.62
3.51
3.54
3.35
3.22
2.90
2.97
3.34
3.17
3.11
2.88
2.82
2.81
2.32
3.47
3.26
3.03
3.06
2.55
2.81
2.71
3.53
3.37
3.30
3.08
3.02
2.78
2.52
2.84
2.84
2000
4.01
3.92
4.03
3.96
3.68
3.48
3.56
3.74
3.63
3.66
3.47
3.31
3.14
3.13
3.45
3.30
3.25
3.00
3.04
2.96
2.54
3.66
3.42
3.31
3.17
2.70
2.96
2.87
3.66
3.48
3.45
3.21
3.14
2.91
2.67
2.92
2.92
6-6
-------
TABLE 6.-I—Continued
Sul
Demand Region Coal Ty?e ( Ib
6. Southwest Bituminous
0.
1 .
1 .
2.
3.
7. Central Bituminous
0.
1 .
1 .
2.
3.
Sub-Bi tumi nou s
0.
1 .
1 .
8. North Central Bituminous
0.
t,
1 .
Sub- Bi tumi nou s
0.
1 ,
1 .
9. West Bituminous
0.
1 .
1 .
Sub- Bituminous 1 .
10. Northwest Bituminous
0.
1 .
1 .
Sub-Bituminous
0.
1 .
1 .
fur
Conr.ent
S02/MMfctu) 1985
<0
80
08
67
50
33
>5
<0
80
08
67
50
33
>5
<0
30
08
67
<0
8C
08
67
<0
80
08
67
<0
80
08
67
67
<0
80
08
67
•
-
-
-
-
-
•
m
-
-
-
-
-
•
•
-
-
-
*
-
-
-
*
-
-
-
«
-
-
-
-
•
-
-
-
<0.
80
08
67
-
-
80
1
1
2
3
5
00
80
1
1
2
3
5
CO
80
1
1
2
80
1
1
2
80
1
1
2
80
1
1
2
2
80
1
1
2
80
1
1
- 2
.08
.67
.50
.33
.00
.08
.67
.50
.33
.00
.08
.67
.50
.08
.67
.50
.08
.67
.50
.08
.67
.50
.50
.08
.67
.50
.08
.67
.50
a By ICF Inc., 1850 K Street, N.W., Washington,
b Taken from Reference 4.
2
2
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
0
0
2
2
2
2
2
3
3
2
2
2
2
2
2
D.C.
.94
.89
.03
.71
.57
.95
.82
.77
.76
.95
.95
.94
.69
.34
.>:•
.53
.37
.43
.64
.47
.29
.91
.52
.36
.86
.80
.69
.63
.26
.53
.34
.13
.10
.17
.30
.06
.06
.05
.05
1
3
3
2
3
2
3
2
2
2
3
3
3
2
2
2
2
2
2
1
1
1
1
1
1
0
0
2
2
2
2
2
3
3
2
2
2
2
2
2
990
.31
.26
.65
.09
.79
.17
.84
.97
.95
.23
.22
.15
.61
.47
.60
.60
.44
.48
.87
.66
.34
.40
.62
.48
.92
.90
.87
.76
.49
.78
.42
.37
.24
.27
.81
.14
.14
.10
.05
1995
3
3
3
3
2
3
2
3
3
3
3
3
2
•"1
4.
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
3
3
3
2
2
2
2
2
.58
.51
,C2
.19
.94
.39
.97
.09
.C7
.34
.18
.27
.58
.52
.72
.72
.52
.63
.92
.74
.35
.50
.39
.47
.01
.04
.91
.82
.85
.60
.49
.42
.30
.29
.78
.29
.29
.29
.05
2000
3.77
3.69
3.16
3.21
3.08
3.46
3.07
3.14
3.09
3.46
3.22
3.40
2.66
2.58
2.81
2.81
2.81
2.54
2.01
1 .76
1 .38
i . 54
1 .75
1 .59
1 .12
1 -06
3.15
2.98
2.91
2.6?
2.59
3.58
3.42
3.42
2.80
2.36
2.36
2.36
2.10
, 20006.
6-7
-------
Waste Pretreatment Costs
The cost, to pretreat a hazardous waste stream to make it suitable for
firing in an industrial boiler offsets the money saved in fuel replacement
and alternate disposal costs. Estimating the magnitude of this offsetting
cost requires cost data for the typical pretreatment systems being employed.
SPA collected preliminary data concerning facilities pretreating wastes used
as boiler fuel in an OSW Burner Questionnaire Survey. This was followed up
by a telephone survey of 11 facilities pretreating wastes.
Results of these initial efforts indicate that four basic types of
pretreatsnent are practiced: blending, solids removal, water removal, and
heating the waste to adjust the viscosity. Blending of a waste stream with
either a conventional fuel or another compatible waste stream is used to up-
grade it to the quality required by the particular boiler design. Th^s is
commonly done to reduce the concentration of metals, solids, water, or halo-
gens, and/or to increase the heating value. Solids are being removed to pre-
vent plugging of the burner gun and to meet particulate emission standards
if they are inorgai.ic. Water reduces the caloric value of the waste anc can
cause burner flame instability. Viscous wastes must be heated to make them
pumpable and to permit proper atomization necessary ijr high combustion ef-
ficiency.
Wastes requiring more extensive pretreatment than the four basic types
listed above are generally disposed of oy alternate means. Many wastes sim-
ply cannot economically be made suitable for firing a boiler, i.e., inorganic
wastes.
The type ~f equipment used for blending, solids and water removal, and
thermal pretreatment is generally extremely simple. This equipment was des-
cribed in Section 4. Costs follow for a typical equipment arrangement for
each of the four pretreatment processes. These costs are developed for the
range of waste flow rates considered representative of those being fired in
industrial boilers. The flow rates given in the developed cost curves below
are the rates of treated waste leaving the pretreatment process.
The following includes a subsection on the desired waste fuel charac-
teristics for effective operation and discussions of the four pretreatment
costs identified Above.
Waste Fuel Characteristics for Effective Boiler Operation
Fuel characteristics can have significant effects on boiler operation,
such as:
o Deterioration of boiler tubes and refractories
o Excessive emission levels
o Reduced combustion efficiency
o Increased maintenance requirements
6-8
-------
However, the range of boiler sizes and types, combined witn the range of
available waste fuels, generally make it impossible to set. specific limits on
fuel paraaenters. Each case must be based on a separate economic evaluation,
considering cost of various levels of fuel pretreatment; of boiler operating
cost, including possible reduction in steam production; and of possible flue
gas pollution control tecnnxques.
The effects of several fuel characteristics on boiler operation will be
discussed below.
Chlorides and Halides. Several sources (References 6,7) have indicated
that they prefer to keep the concentrations of chlorides and halides below
0.5 - 1.5*. However, other sources (References 3,9,10,11) have indicated
that the actual corrosive effect may depend on boiler operating techniques
and the final form which the halides will take. If the ash contains enough
free alkali, the resulting chloride salts may cause severe fluxing of the
boilar refractories and deterioration of some types of alloy boiler tubes'.
The corrosive effect also depends on boiler design parameters which govern
the probability of ash impingement on boiler surfaces while the ash is still
hot enough to be soft and adhere to the surfaces.
High levels of chloride salt, may also form a mist of fine particles,
contributing to stack opacity. Dilution of the waste fuel is the only way
to reduce chloride concentrations.
The primary effect of releasing halides as acids will be the possibil-
ity of increased corrosion of ductwork and equipment handling stack gases if
the gas temperature falls below the dew point. This can be avoided by first
warming the boiler, using a non-corrosive fuel, then introducing the waste
solvent, and monitoring the stack to assure that the temperature does not
fall below the dew point.
Sulfur. Although excess sulfur in the waste fuel may not be found very
often, if burned it will cause corrosion in ductwork and equipment handling
combustion gases if the gas temperature falls below the dew point. This can
be avoided if the boiler is first heated using a suitable fuel, and if the
stack gas temperature is monitored (Reference 9). Alternatively, high sul-
fur content waste may be co-fired with low sulfur waste or fuel to reduce
the sulfur concentrations of the boiler feed (Reference 6).
High sulfur levels in the waste solvent will also cause excessive levels
of 3C>2 in the boiler emissions.
Sediment and Particulate Matter. High levels of sediment and particulate
matter will i- crease apparent viscosi v of the waste solvent, cause blockage
of filters or burner nozzles, settle in waste solvent lines, and may (if not
incinerated) cause increased opacity of the stack gas,. If particle size is
below 100 mesh, and piping is properly designed, wastes containing 5% or more
solids may be red to the boiler. Effects on emissions depend on boiler de-
sign and stack gas treatment, as well as the contributing effects of ash in
the waste. This will be discussed in the next section.
6-9
-------
Ash. Effects of ash, combined with uncombusted particules, include
fouling the tubes (Reference 7), slagging the furnace walls (References 8,
9), and increasing opacity of the emissions. These effects can vary widely,
depending on actual composition o: the ash, design of the Boiler, and stack
gas treatment. Boilers designed for natural gas or distillate fuels :say be
significantly affected because they aie no; equipped with soot blowing pro-
visions, but there may be little effect in boilers designed for residual or
solid fuels (Reference 9). Although not definitive, ash fusion-temperature
tests may permit some prediction of boiler effects, but test firing in the
actual boiler, or in a similar boiler at the manufaturer's facility, is the
best way to determine the effects of a given fuel on a given boiler.
Moisture. It is of primary importance that the boiler fuel be a single
phase or a stable emulsion. Feeding a two-phase mixture might result in ra-
pidly alternating feed compositions. The results can vary from flame failure
to rapid, almost explosive, heat release.
Even when the water and organic components of the waste are miscible,
there are two other effects which must be watched: decrease in flue gas dew
point, and decrease in boiler efficiency. Moisture in the fuel will appear
in the stack, and the higher the moisture level, the lower the dew point of
the stack gas. Unless care is taken, there might be a sndden increase of
corrosion in stack ducting and equipment.
Very high moisture levels also lower boiler efficiency. Heat required
to vaporize the moisture is carried to the stack, and, as a result, furnace
temperature is lowered, resulting in reduced heat transfer through the boil-
er tubes. These effects can be minimized by limiting the waste fuel rate to
about 5% of the total fuel rate. With this proviso, there is no limit to the
moisture content of the waste fuel; the primary fuel will produce enough heat
to vaporize and incinerate the waste while meeting the steam demand.
Viscosity
If viscosity of the waste fuel is too high, it will not be possible to
atomize the liquid into droplets small enough to oxidize completely. Good
acomization can usually be achieved if waste fuel at the burner is less than
750 SSU (Reference 12). Heaters are available which can heat the fuel as
high as 500°F (260°C) to reduce viscosity.
Reasonable pump designs and piping pressure drops set limitations of
waste fuel viscosity at about 10,000 SSU. It is practical to maintain stor-
age temperature at 200°F (95*C).
Thus, the limitations on viscosity are:
o 750 SSU at 500°F
o 10,000 SSU at 200°P
Fuel Composition
Smaller boilers are usually equipped with simple controls which link
the combustion-air dampers with the fuel control valve (Reference 6). Sig-
nificant changes in fuel composition will cause fluctuation in the required
6-10
-------
fuel.-air ratio. If too little air flows through the boiler, CO concentration
in the stack gas will increase, and combustion efficiency will decrease. One
technique to avoid this problem is to set waste fuel feed at a constant flow
rate, to supply no more than ?0% of the expected heat load. Primary fuel
flow is varied to supply the required total hea ; load, and combustion air
flow is set to assure that there will be adequate air throughout the boiler
firing range, and for all expected waste fuel compositions.
Larger boilers are usually equipped with combustion controls which can
permit firvng 100% waste fuel if the heat of combustion is high enough to sus-
tain a good flame, if fire box temperatures are high enough to insure adequate
combustion, and if heat realease is adequate to supply steam demand.
Pretreatment Cost
Blending Costs. A blending system consists of three tanks, each equipped
with an agitator. Tank size is a function of waste fuel throughput and ex-
pected blending time. Because one tank would be required as a dry tank, ad-
ditional capital costs for blending are based on two storage tanks and three
agitators. These costs are shown below.
Capital Costs - Blending
Installed3
Capital Cost
(S)
1,000 12,000
2,000 20,000
5,000 30,000
10,000 51,000
20,000 80,000
a Two tanks and three
agitators.
Thermal Treatment. Although there is an infinite range of possible li-
quid wastes which might be burned, the techniques for introducing the wastes
into the boiler are similar, and are based largely on techniques already de-
veloped for burning conventional fuels: The liquid must be sufficiently
atomized to pernu t fairly rapid combustion. The resulting flame must be as
long as possible to minimize flame temperature and to minimize formation of
NOX, but the flame must not touch the boiler refractories, walls, or tubes
unless the boiler is specifically designed for such contact. Although good
atomization may be achieved with fluid vicosities as high as 5,000 SSU (Re-
ference 12), it is common practice with standard fuel oils to limit viscosity
to 750 SSU at the burner nozzle. Pumping problems increase with liquid vis-
ccsity, ani common practice with fuel oils is to maintain viscosity below
10,000 SSU for storage and transfer, with a-iditional heat supplied just be-
fore the burner to further lower the viscosity.
For purposes of developing thermal pretreatment costs, liquid wastes
have been categorized into three groups according to their viscosity:
6-11
-------
1. Viscosity at ambient temperature _<_750 SSU.
2. Viscosity at ambient temperature is between 750 and 10,000 SSU.
3. Vii-cosity at ambient temperature 2.10,000 SS, but <10,000 SSU, at
200°F.
If the viscosity is less than or equal to 750 SSU, the waste is both
pumpable and capable of being properly atomized. No thermal treatment is
needed for wastes in this category. Wastes of viscosity beteen 750 and
10,000 SSU are pumpable but must be heated to achieve satisfactory atomiza-
tion. A fuel heater (shell and tube heat exchanger) may be used to raise
the liquid temperature sufficiently to lower the viscosity to below the lim-
it for gooJ atomization (750 SSU). An equipment arrangement for this purpose
is shown schematically in Figure 6.1. In this scheme, the waste is heated
(by steam) as it pases through the heat exchanger before entering the boil-
er. The only thermal treatment equipment items needed are the heater and
piping. The installed costs of this equipment are plotted as a function of
waste feed rate in Figure 6.2 as curve 1.
Wastes of greater than 10,000 SSU viscosity at ambient temperature must
not only be treated to insure good atomization, but must also be kept hot
enough to prevent them from setting up in the storage tank and piping. Equip-
ment requirements for waste in this category include the heat exchanger for
heating the waste as it is pumped to the boiler, provisions to recirculate
the waste back to the storage tank, as well as items to keep the liquid warm:
o Storage tank insulation
o Storage tank heaters
o Pipe tracing and insulation
o Addition of a spare circulating pump
Installed costs of this equipment, excluding the storage tank insulation and
heaters, are plotted as a function of waste throughput in Figure 6.2 as curve
2. Because ES is not supplying waste storage costs, the tank heater and ir
sulation costs are being provided as separate cost items. These costs are
presented in Table 6.5. The storage tank heater for which costs are presented
here is the immersion steam coil type. Insulation costs are for three inches
of fiberglass insulation.
Solids Removal Cost. As discussed in Section 4. sludges and slurries
are generally not pretreated to reduce the solids content to levels where the
liquid can be burned in an industrial boile- but, rather, are mixed with solid
fuels (coal or non-fossil fuels) before firing in a solid fuel unit. For the
solids removal system priced below, an upper limit on solids content of 5% by
•weight is used. Wastes of higher solids content will probably be disposed of
by alternate means.
A typical solids removal system will include either both settling and
screening or settling and filtration. A typical system is illustrated by Fi-
gure 6.3. In this scheme, the waste is first pumped from a processing unit
into one of three tanks. The three tanks are alternately used in either a
receiving mode, a holding mode to provide a quiescent environment for set-
tling, or a transfer mode from which the supernatant waste is pumped to fi-
nal solids removal by either screening or filtering. A duplex basket type
6-12
-------
FIGURE
BLOCK DIAGRAM, THERMAL. TREATMENT TOR
IN THE RANCH OF 750 - 10,000 !_.S" *T AMBIENT TEMPERATURE
RECYCLE
10 STORAGE
FROM
SIORAGF
o
BACK-PRESSURE
REGULATOR
DUPLEX
STRAINER
STEAM
O
o
'1
f PUEL~|
1 HEATER |
CONDENSATE
ROIIER
-------
FIGURE 6.2
THERMAL TREATMENT EQUIPMENT COST
AS A FUNCTION OF WASTE FLOW RATE
a
u
-------
TABLE 6.5
STORAGE TANK HEATER AND INSULATION INSTALLED COSTS
Tank Size Heater Costs Insulation Costs
(gal) (jjj (_£)
500 11,000 430
1,000 12,000 660
2,000 14,000 960
5,000 17,000 1,310
10,000 19,000 2,850
20,000 38,000 3,640
6-15
-------
FIGURE 6.3
TYPICAL SOLIDS REMOVAL SYSTEM
FROM PROCESS UNIT
SLURRY TO
DISPOSAL •
•TO BOILER
DUPLEX
STRAINER
OR
FILTER
Sett!ing
Tan I;
6-16
-------
strainer (100 mesh) is commonly used when the polishing stage is achieved by
screening. A dual cartridge filter is generally the choice if a filter is
ussd. Either of these devices permits cleaning without process shutdown.
The cost for this type system is shown in Figure 6.4 as a function of the
throughput of treated waste. In developing this curve, it was assumed that
v»aste entering the system contained 5% solids by weight as mentioned above.
The capital costs should be essentially independent of the solids content for
levels below 5%. The system shown in Figure 6.3 is capable of reducing the
solids content down to levels where no plugging of the burner gun nozzle oc-
curs, wnich is probably between 0.5 to 1%.
The equipment items included in the cost development are the tanks, pump,
piping, and a cartridge type filter. Since a duplex screen is generally less
expensive than a filter, a filter was assumed in developing the cost curve.
The tanks are above ground, vertical, conical roof of carbon steel construc-
tion with concrete support pad. They were sized to provide a detention time
of approximately 10 hours which should provide sufficient settling time for
most nonviscous wastes. The costs estimate includes only two of the three
tanks shown in Figure 6.3 because a waste holding tank is generally used
when no pretreatment is required. A Moyno® pump is assumed for removal of
the sludge or slurry formed by the settled solids. It is assumed that this
sludge/slurry is either pumped into a tank truck, drums, or a cart for trans-
port to safe disposal.
water Removal costs
Decanting was the only means of removing water from wastes fired in
boilers reported by those contacted during the telephone survey. This pro-
cess only removes the water in excess of the solubility limit in the organic
fraction of the waste. Standard above ground cylindrical storage tanks are
used as decanting devices. A typical arrangement is depicted schematically
in Figure 6.5 and is very similiar to that used for removing solids.
The total installed capital coats for this type of system were developed
as a function of the dewatered waste flow rate. These costs are presented
in Figure 6.6. Costs are presented for three different assumed inlet water
contents: 10, 50, and 75% by weight. It was assumed that the water content
of the waste is reduced to 5% by weight. This level of water should not
cause any significant adverse effects on the boiler.
Equipment items included in the costs are two tanks, a pump for trans-
ferring the decanted water, and piling. Because a holding tank is often used
when no pretreatment is required, one of the three tanks shown in Figure 6.5
is not included in the cost estimates. Similarly, the duplex strainer and
pump for transferring the treated waste to -he boiler are required if there
is no dewatering and, therefore, are also not included. The tanks are cylin-
drical, above ground, conical rocf design of carbon steel construction with
a concrete support pad. They are sized to provide approximately 10 hours
detection time which should be adequate for dewatering most wastes where de-
canting is practical. A gear pump of carbon steel construction was assumed
for removing the water from the tank. It was sized to remove the water layer
in less than one hour.
6-17
-------
FIGURE 6.4
SOLIDS REMOVAL EQUIPMENT COST AS
A FUNCTION OF WASTE FLOW RATE
e
O
C
10OO
Wast* Row, S
6-18
-------
FIGURE 6.5
TYPICAL WATER REMOVAL SYSTEM
rn^.., r^ri^v_=.as u.\n
f
\
WATER TO v,
^_ ' \
^-
r
< (
i
i
, i
i
j
J
Decani
Tank
- i
\
j
i
i
i
t
i i
j
Decant
Tank
'
1 !
i
— )
w^^K
^--DUPLEX
STRAINER
S — W .— H .. ~ < J
V-— '' Decant
Tank
TO BOILER
6-19
-------
FIGURE 6.6
WATER REMOVAL EQUIPMENT COST
A3 A FUNCTION OF WASTE FLOW RATE
Wast* FLOW, G
6-20
-------
Boiler Modification Costs
Burn-?- Modification Costs
The cost for modifying an existing burner system to provide for firing
hazardous waste is extremely site specific and depends on the existing burn-
er type and capacity, number of burners employed, type of conventional fuel
fired, properties of the waste, and quantity of waste to be firei. Generally,
the least expansive approach is taken. Some boilers were originally designed
to fira hazardous waste as either the primary or as the supplemental fuel.
Others require only that a burner gun be replumbed to fire the waste. This
would not require significant capitial expenditure. In many other instances,
the hazardous waste is blended with the conventional fuel and fired with no
modifications being made to the burner. This is the basic approach used
when burning solid wastes in coal-fired boilers.
Costs given here are for providing the necessary burner components to
fire a gaseous or liquid hazardous waste in natural gas, oil, coal, and com-
bination fossi-1 fuel-fired boilers. Solid hazardous waste firing is not ex-
tensively practiced and is generally limited to coal-fired units where it is
blended with the coal before being fed to the boiler. All costs given below
assume that the waste is piped to the burner, i.e., costs do not include a
fuel handling train. They are based on prices obtained from boiler burner
vendors and therefore may be higher chan actual costs as many large indus-
trial boiler owners fabricate their own waste burners.
There are two basic alternatives commonly practiced for burner system
retrofit to accomodate waste firing that require significant captial expendi-
tures. The first alternative is to install additional burner guns for waste
firing in the air register of the existing burner(o). Waste is simply atom-
ized into the primary fuel (natural gas or oil) flam* envelope. In boilers
equipped only with ring burners for natural gas, liquid waste co-firing can
be effected by plumbing a liquid waste nozzle into the burner centerline in
much the same fashion that these burners are retrofitted for fuel oil firing.
For safety reasons, the waste burner gun must be tied into the flame safe-
guard syacem to shut off the waste flow in case of flame out. A good qual-
ity air or steam atomizing burner gun capable of handling the range of waste
flows typically fired in this type arr&ngement can be purchased for under
$5,000. The total installed cost, including plumbing and slectrical to tie
into the flame safeguard system, can run from $10,000 to $15,000.
In some burner retrofits, replacement of the entire burner assembly may
be required. A complete burner assembly includes air register, burner gui**,
ignitor, flame safeguards and windbcx. The number of new burner assemblies
installed as part of the retrofit depends on the quantity of waste to be
burned and the boiler size. The cost of a single burner assembly depends on
a number of factors including the number 01 fuels/wastes fired (which deter-
mines the number of guns), the properties of the fuels/wastes, and the fuel/
waste quantities. The small gas and oil fired packaged boilers selected for
the economic analysis (15 million Btu/hr heat input) are single burner units.
A complete burner assembly for one of these boilers that is capable of firing
both oil (or gas) and a liquid hazardous waste typically costs $25,000 to
6-21
-------
530,000 installed. A 1 5i_ million Bru/hr heat input oil or gas fired ooiler
generally has from one tc six burners. Therefore, the Burners for tnose
units vary from 25 to 150 million Btu/hr heat ir.--'-.~. Installed cost of tnese
burner assemclies typically runs from $35,000 to $40,000 for the 25 million
Btu/hr unit to $73,000 to $90,000 for the * 50 million Btu/hr unit. A large
field erected oil fired ooiler of 400 million Btu/hr heat input generally nas
four or more burners. Therefore, costs for the individual burner assemblies
should be about the same as those for a 150 million Btu/hr neat input boiler.
Replacement of a complete assembly for a pulverized coal-fired boiler
tc provide for hazardous waste burning is unlikely because it is almost al-
ways possiole to insert some type of waste gun in the air register of one
of tnese units. Furtnermore, many of these units also have oil or gas guns
that can be replumoed for firing the waste. For the size pulverized coal
boilers selected for the economic analysis, the burners are most liXely be-
tween 35 to 75 million Btu/hr capacity. Installed cost of a complete assem-
bly capaole of co-firing a liquid or gaseous waste witn the coal typically
runs from $75,000 to $100,000. Installation typically is 50% of the equip-
ment costs.
Stoker boilers are purchased with or without auxiliary burners to fire
oil or gas. When the boiler does not have a bur.ier that can be retrofittt.
to burn the waste, a complete burner assembly is necessary. To install a
complete burnar assembly in the furnace well of a stoker not having an exist-
ing opening to accept it would b« extremely expensive because a section of
the wall cubes must be removed. This generally requires an expensive engi-
neering study and costs for complete retrofit are estimated to run between
$200,000 to $300,000 depending on the boiler and burner designs. If an ex-
isting opening can be used, the installed cost would be considerably less per
burner assembly. As with the other boiler types described above, the cost
would depend on the size of burner assembly installed. If the waste is to
be co-fired with the coal, the burner would likely be designed to provide
approximately 25% or less of the total ruel requirement. For the 30 million
Btu/hr stoker, the maximum capacity burner assembly would be slightly less
than 10 million Btu/hr and would typically cost $20,000 to $25,000 installed.
For the 75 and 150 million Btu/hr stoker models, one or two burners m'.y be
used. A single 25 million Btu/hr burner costing approximately $35,COO to
$40,000 installed would be representative for the 75 million Btu/hr boiler.
For the 150 million Btu/hr stoker boiler, the burn assembly retrofit is
likely to cost $60,000 to $70,000 regardless jf whether c.ie or two units
are used to supply up to 25% of the total heat input.
Oxygen and Carbon Monoxide Monitoring
Continuous monitoring of 02 anc* CO concentration in the exhaust gases of
a boiler burning a hazardous waste may be necessary to insure that good com-
bustion conditions are being maintained. Many boilers are already equipped
with 03 and/or CO monitoring as components of automatic excess air trim sys-
tems. These systems were purchased primarily to save fuel costs through in-
creased combustion efficiency. Automatic excess air systems are becoming in-
creasingly common on new boilers of all types (Reference 3). Furthermore, one
of the requirements of the new source performance standards for industrial
boilers being considered is that either 02 or C02 monitoring be installed to
measure the amount of diluent air leaking into the stack gases (Reference 14).
6-22
-------
Tne cost of ooth 0~ and CO continuous emissions monitoring systems vary
depending or. vendor and type (in-situ or extractive). This variation is in-
significant, however, for the Ifel of accuracy needed for the economic anal-
ysis. Therefore, a single cost figure is giver, for each monitoring system.
The cost data presented here were developed using engineering judgment and
prices obtained from vendors, and reference costs were for tne 1983 period.
Typically, an installed oxygen monitoring system costs approximately
515,000 including analyzer, sample transport and conditioning system (if
an extractive type), strip chart recorder, and installation. This cost is
essentially independent of boiler size and type.
Carbon monoxide monitoring systems are more expensive than C<2 systems
oecause a more sophisticated analyzer is required. A representative cost
for an installed CO monitoring system is $20,000 including the CO analyzer,
a sample transport and conditioning system (for an extractive type), a strip
chc.rt recorder, and installation. As with the 03 system, this cost should
be independent of boiler size and type.
The above costs assume that the collected strip chart data will be re-
duced manually. For an additional $20,000 an automatic data reduction system
can be provided (Reference 14).
Furnace Temperature Monitoring
One basic strateoy for regulating the combustive destruction of hazard-
ous waste in boilers is to set performance goals thet must be met Dy the
equipment. The ability of the combustor 10 generate a specific environment
(time, temperature, turbulence) is one of several performance goals being
considered as part of an approach for regulating the burning of hazardous
waste in boilers. If this performance goal is adopted, some means of moni-
toring the boiler furnace temperature may be required.
One or more thermocouples will be needed to monitor the temperature in
the hot environment of a boiler furnace. A pryometer is not capable of suf-
ficient accuracy. The thermocouple must be capable of measuring temperatures
in the range of 2000 to 3000°F and, therefore, should be radiation shielded.
If more than one thermocouple is required, a scanner is recommended to alter-
nately switch to each thermocouple. A simple strip chart recorder is adequate
for providing a permanent record of thermocouple ouput. A linearizer with
reference junction compensation is necessary. This can be either a separate
device or built into the strip chart recorder. Sufficient thermocouple wir-
ing is needed to link the various components. Total cost of this type of
monitoring system is typically $4,000. These costs are based on prices ob-
tained from Reference 15. Installation should be less than $1,000. These
costs should be relatively independent of boiler size and type.
Waste Feed Metering. If limitations on the hazardous waste feed rate
are adopted, capital outlays for flow metering will be necessary. Liquid,
gaseous, or both types of flow meters will be needed, depending on the waste
fired. Two types of liquid flow measurement devices well-suited for this
application are the positive displacement meter and the orifice meter. These
6-23
-------
units art relatively inexpensive and are applicable to a wiae range of wastes.
Either device can be ootainea for approximately 52,000 to meter flows up tc
2,400 gallons per hour, wnich is tne maximum flow rate one wo^-ld ai.ti npate
for waste feed to an industrial size boiler. (A 400 million Btu per hour heat
input boiler fires spproxiir.ately 2,-100 gallons per hour of residual fuel oil
at full load.) Some type of recording is needed to provide a permanent record
of the waste feed rate. A simple $2,000 strip chart would serve this purpose
adequately. Installation is likely to cost an additional 51,000. Thus, the
total installed cost of a liquid waste feed metering system will De approxi-
mately $5,000.
Three types of meters which should cover the range of gaseous waste
applications are the turoine meter, the vortex shedding meter, and the ori-
fice meter. Orifice meters result in a large permanent pressure drop and,
therefore, are limited to applications where such large pressure drops can
be tolerated. Both the turbine and vortex shedding meters are low pressure
drop devices. A turbine meter or vortex shedding meter capable of measuring
gas flow rates up to 30,000 standard cubic feet per hour costs approximately
$6,000. Most hazardous waste streams fall within this.flow range. An ori-
fice meter capable of measuring flows up to 30,000 standard cubic feet per
hour costs approximately $3,000 including all the ancillary items (pressure
transducers, etc.). A recording device for any of these devices can be pur-
chased for approximately $2,COO. Installation should cost less than $1,000.
Therefore, the total installed cost of a turbine or vortex shedding metering
system might typically cost $9,000 and an orifice meter system might cost
$6,000.
Process Control Instrumentation
As with burner gun assemblies, the costs to retrofit the combustion
control instrumentation to permit the firing of hazardous waste is very site-
specific and generally che most economical approach is taken. Because the
cost is so site-specific, some generalizations are necessary to tailor them
for use in analyzing the economics of incinerating wastes in boilers. For
example, the costs presented here include only components necessitated by
the addition of waste firing.
For many retrofits of boilers to fire hazardous waste, no significant
change to the combustion control instrumentation is required. For instance
when a waste is co-fired with a conventional fuel at a rate of less than 25%
of the total fuel demand, the genc-ral practice is to base-load with waste at
a steady feed rate and modulate the conventional fuel only. In this case,
the combustion control instrumentation is essentially unaltered. Similarly,
when the waste is blended with the conventional fuel before it is fired,
there is no combustion control instrumentation modifications required. In
situations where the total hear input is provided by waste(s) that can be
fired in the same burner(s; already installed on the boiler, there, also,
are no significant changes in the combustion controls required.
Installation of a completely new combustion control system is rarely
needed to fire hazardous waste. When meiulation of the waste feed rate is
6-24
-------
required, it is almost always .achieved by integrating some additional compo-
nents into the existing system. A possible exception is when tne existing
system is so antiquated tnat it is more economical t? install a new one than
to upgrade it to handle tne waste.
The majority of oil, gas, or combination gas/oil boilers of 30 million
3tu per hour heat input or less have direct-positioning (jackshart) type
combustion control. Because of its simplicity, this type of control system
can easily be modified to integrate modulation of the waste feed. This can
easily be done in most applications tor less than $5,000 including hardware
and installation. Hardware includes waste flow valve, additional control
linkage for the jackshaft, and some minor plumbing items.
All other boilers have some form of a metering type combustion control.
Hardware necessary to control the combustion of the waste can generally oe
integrated into this type control system for $2C;,000 to $30,000, including
installation. Hardware includes a setpoint controller, flow control valve,
flow metering device, miscellaneous piping, and electrical wiring. The flame
safeguard is not considered as part of the control system here because it was
included as a burner assembly component.
Air Pollution Control Devices Costs
If EPA promulgates emission standards for boilers firing hazardous waste,
similar to those adopted for incinerators it is likely that so-ne boilers may
either require the addition of removal equipment for particulate matter and
HC1 or the reduction of the ash and/or halogen content of the waste to be
fired. Air pollution control devices may also be needed to limit the emis-
sions of toxic metals wnich may be emitted as either fine particles or as
vapors. Due to the large capital and operating expense associated with the
application of air pollution control devices, it is not likely that burning
hazardous wastes will prove economically attractive should these devices be
required. Boiler operators are more apt to meet emission standards by either
blending the waste to reduce the concentration of the contaminant of concern
or find alternate means of disposal. It is impossible, however, to predict
with certainty the industry trends relative to the additional air pollution
control devices. Therefore, the costs presented in this section are for the
control devices that are most likely to be applied should performance stan-
dards for particulate matter, metals and/or HC1 be adopted.
If control devices are chosen as the means to meet emission standards,
then one of three basic systems is likely to be used. For particulate mat-
ter control (including metal particles), either an eloctrostatic precipita-
tor or a baghouse is the likely candidate, "he high energy costs associated
with scrubbers will limit their use when only particulate matter removal is
required. However, a scrubber system is a logical choice for the removal of
HC1 or vaporous metals. When more than one of the four pollutants (particu-
late matter, HC1, metal particles, metal vapors) must be removed, a scrubbing
system becomes even more attractive.
Capital costs are presented below for each of the three types of air pol-
lution control devices (APCD) as a function of the volume of gas treated. The
costs presented here were developed from information obtained from Reference
16 and 17.
6-25
-------
The purchased equipment costs provide the oasis for estimating the cap-
ital APOD costs. Factors used to calculate tne capital costs are listed ir.
Table 6.6. These factors represent the inaividua- capital cost components
expressed as a fraction of the purchased equipment costs. Direct capital
costs consist of tne basic and auxiliary equipment costs in add.tion to the
laoor a.nd material required to install the equipment. Indirect costs are
tnose costs not attributable to specific equxptr.ent it^as. Each of tne con-
trol system coat curves presented below include tne cost of auxiliary equip-
ment normally associated witri such a system.
Electrostatic Precipitator Costs
The electrostatic precipitator is likely to be a primary candidate should
a. boi.er operator only need to install an APCD to meet a particulate matter
emission standard. These dev^c-es typically have a high particulate matter
collection efficiency. Furthern:ore, they are low pressure drop devices which
make them especially well-suited for retrofit applications such as industrial
boilers co-firing wastes. Another advantage of being a low pressure drop device
is that new fans or other gas handling equipment are not needed. Consequently,
the costs presented below include no provisions for a new fan or other gas
handling equipment.
Capital costs of electrostatic precipitators are plotted in Figure 6.7
as a function of the exhaust gas treated. These costs were developed using
the Deutsch equation:
A = -Q In (1 - n)/w
where: n is the design efficiency specified
w is drift velocity, ft/s
Q is the exhaust flow rate, cfs
A is the net plate area, ft2
The drift velocity quantifies the electrical characteristics of the dust and,
therefore, has a large effect on the collection efficiency and plate area.
For these reasons, costs given for drift velocities of 0.11 b ft/sec are re-
presentative cf high-resistivity dust and of 0.282 ft/sec are representative
of low-resistivity dusts. High resistivity dust cost data should be used for
application to low sulfur containing fuels. Conversely, cost data for low
resistivity should be applied for high sulfur containing fuels.
Since they are also a function of the efficiency, ESP costs are reported
for two different levels, 99.5 and 99.9%. It is unlikely cnat efficiencies
in excess of 99.9% vill be required for this application as it is difficult
to routinely achieve higher levels.
The ESP cost curves are for an insulated carbon steel unit since the
wastes or waste/fuel mixtures burned may possibly contain significant quanti-
ties of water, sulfur, phosphorous, or halogen compounds. The presence of
significant quantities of these materials in the fired waste can result in
the formation of sulfuric, hydrochloric, or phosphoric acids in the combus-
tion gases. Insulation will keep the temperature of the combustion gases
above the dew point of these acids and thur, prevent corrosion of the ESP.
6-26
-------
TABLE 6.6
AVERAGE COST FACTORS FOR ESTIMATING CAPITAL COSTS
0
1
2
Cost Factors
IRSCT COSTS
Purchased equipment caste
a, Control devica and
auxiliary equipment
b) Instruments .1 controls
c } Taxes
A) Freight
Subtotal
. Installation direct costs
a) Foundations & support
b) Erection & handling
c) Electrical
d) Piping
e) Insulation
f) Paintiny
Subtotal
F3P
0.82
0.10
0.03
0.05
1 .00
0.04
0.50
0.08
0.01
0.02
0.02
1 .67
Wet
Scrubber
0.82
0.10
0.03
0.05
1 .00
0.06
0.40
0.01
0.05
0.03
0.01
1 .56
Fabric
Filter
0.82
0.10
0.03
C.05
1 .00
0.04
0.50
0.08
0.01
0.07
0.02
1.72
INDIRECT COSTS
3
Installation indirect costs
a) Engineering & supervision
b) Construction & field expense
c) Construction fee
d) Startup
e) Performance test
f) Model study
g) Contingencies
TOTAL
0.20
0.20
0.10
0.01
0.01
0.02
0.03
2.24
0.10
0.10
0.10
0.01
0.01
0.0
0.03
1 .91
0.10
0.20
0.10
0.01
0.01
0.0
0.03
2.17
6-27
-------
FIGJRE 6.7
:APITAL COSTS OF ELECTROSTATIC
NOTE
NOTE S
n • COLLECTION -FIC
NCTc A - FOR 3JST HAVING HrCH 3E5IST!Vrrv
N07I 3 • FOR OUST HAVING MCGcSATtTO LCW SSSISTIVITY
100'
10. '000
100,000
EXHAUST GAS 3ATI, sc:m
I.:CQ ooc
6-28
-------
baghouse Costs
Baghouses have increasingly b
-------
FIGURE 6.3
CAPITAL COSTS OF FAdR/.C FILTERS (REVERSED AIR),
CARBON STEEL CON?Tm'CTICK>
—4-53-
1QOCO
100,000
CHAUS7 GAS SATS. 3cr
i.ooeccc
6-30
-------
FIGURE ts.i
V*
f-,
ICC'—
10:—
10 000
1QO.XG
EXHAUST GAS RATE, atitr,
i.'OXJ.COC
6-31
-------
impacts on boiler subsystems tnat can result from Durning was-ces containing
these species are discussed in detail in Referer.ee 19. There also may be
increased boiler 0/M costs as a result of increasing the number of burner
guns and using more complex combustion control instrumentation. These costs
are not expectad to be significant, however.
Equipment required to meet adopted regulatory requirements will likely
farther increase tne O/'M costs associated with firing hazardous waste over
those associated with firing conventional fuels. Air pollution control ae-
vices addea to meet emission limitations will increase the 0/M cost burden.
Further costs will be incurred to maintain Oo and/or CO monitoring, waste
metering, and furnace temperature monitoring systems. O/M costs assocaited
with waste metering are, however, expected to be insignificant. The coats
of maintaining a furnace temperature monitoring system is difficult to esti-
mate because of the uncertainty in predicting the frequency of thermocouple
replacement.
This section presents 0/M costs for the following of the aforementioned
items:
o Waste pretreatment
o Boiler modification costs
o Air pollution control devices
Increased 0/M costs associated with waste storage and handling ether than
pretreatme.it are provided by others.. Boiler maintenance costs stemming from
corrosion and fouiing-related problems resulting from burning hazardous
wastes are not given as there is not sufficient information available to
develop them.
Pretreatment 0/M Costs
The annualized 0/M costs associated with the four pretreatment processes
(blending, thermal treatment, water removal, and solids removal) for which
capital costs w«re given in this section are presented here. Certain ele-
ments of these costs are generic to all four processes. These are listed
in Table 6.7.
As shown in Table 6.7, the 0/M costs include both direct and indirect
components. The direct components include operating labor, maintenance,
utilities, and residue disposal. Operating labor is taken at S9.75/marhour.
Estimated labor requirements are one manhour per shift for solids removal,
water removal, and blending. For thermal treatment, it is assumed that no
operating labor is needed. Supervision is estimated at 15* of the total op-
erating labor costs (Reference 20). Maintenance requirements are difficult
to predict accurately for these types of operations. For such actuations,
maintenance is generally tafcon as 2-6% of the capital costs (Reference 21).
Five percent was used in these estimates.
The only utilities required for these processes are electricity to drive
the pumps and steam for the thermal pratreatment. Pump power consumption is
estimated from the following equation:
6-32
-------
TABLE 6.7
:OMPONE>TS Ur ANNUALIZEU C'JST:.
Direct
Operating Costs
Cost Factora
Operating labor
Operator
Supervisor
Maintenance
Utilities
Electricity
Steam
Residue disposal
Wastewater
Sludges
59.75/manhour
15% of operator
5% of capital costs
S0.05/fcwh
So.00/1000 IDS.
$2.00/1,000 gal.
SlOO/ton
Indirect
Operating Costs
Overhead
Property tax
Insurance
Administration
Capital recovery cost
80% of operating labor and
maintenance labor
1% of capital costs
1% of capital costs
2% of capital costs
0.132 (using i = 10% and
an equipment life of 15
years )
a All costs are in 1984 dollars.
6-33
-------
kwh = 0'746(GPM)(hd)(SG)H
3960n
•nere: 3PM = flow rate, U.S. gpm
nd = nead of fluid, ft.
33 = specific gravity relative to water 8 60°F, 29.92 jnches mercury
n = overall pump/motor efficiency = 40%
H = hoars of ooeration
An electricity cost of 50.05 per kilowatt hour (Reference 22) was used in the
power costs determination.
Steam is needed for the thermal pretreatment. An average cost of 56.OO/
1000 pounds of steam at 100 psi was usf-d.
Total residue disposal costs include the costs for the sludge generated
by the solids removal pretreatment and wastewater formed during the water re-
noval pretreataent. Disposal of the sludge is by landfilling in a secure,
hazardous waste landfill. The cost of this type disposal depends on location
of the landfill. A representative cost is §100/ton (Reference 22). Water re-
moved from wastes that are fired in boilers may be sent either to a municipal
sewer (more typical of smaller facilities) or to the facilities' wastewater
treatment plant. Municipal sewer charges are quite variable, but $2.00/1000
gallons is a reasonably representative charge. If the wastes are sent to an
industrial wastewater treatment plant, the cost of disposing of the wastewa-
ter should be less. The $2.00/1000 gallon figure was assumed in calculating
the annualized 0/M costs. Residue disposal is increasingly the dominant 0/M
cost for solids removal pretreatment as the quantity of waste being treated
increases. At 50 gal/hr , it is slightly over one-half the total 0/M cost,
while for a waste stream flow of 1000 gal/hr, it would b<= slightly over 90%.
The indirect operating costs include the costs of overhead, taxes, in-
surance, administration expenses, and capital charges. Taxes, insurance,
and administration can collectively be estimated at 4% of the capital costs,
while overhead charges can be considered at <30% of the labor charges for
both 0/M. The annualized capital charges refj.ec>. the costs associated with
capital recovery over the depreciable life of the system and can oe deter-
mined as follows:
Capital Recovery Cost * (capital costs) x
where: i » annual interest rate
n = capital recovery period
For these estimates, a useful life of 15 years and an average annual interest
rate of 10% were assumed.
6-34
-------
Blending O/M Costs
Size and number of blending tan,:s will vary with waste J'jel flow, and
the time required to blend the mixture. This section is bas^d on a three
tank system: one filling, one blending, and one feeding the boiler. However,
because one tank would be required as a dry tank, the costs are based on two
storage tanks, agitators for ail tnree tanks, and energy for all tnree tanks.
Waste materials would be pumped from storage into one tank, the volume
transferred based on flow meters or tank level measurements. Several waste
streams could flow to the blending tank at one time. As soon as the agitator
propeller is covered, the agitator can be started. Required agitation tine
will vary widely, depending primarily on the viscosity of the waste material.
Hazardous waste flowing to the tank should be shut off when the blending
tank is filled. Filling may take one shift and blending may take one shift,
after which the contents of the tank are fed to the boiler during one shift.
Operating and maintenance costs for this operation are shown in the fol-
lowing table.
Operating and Maintenance Cost - Blending
1,000 21,500
2,000 23,600
5,000 26,300
10,000 31,700
20,000 39,400
One hour per shift (3 shifts/day) plus
maintenance (5% of capital costs).
Thermal Treatment O/M Costs
Figure 6.10 presents the annualized thermal treatment O/M costs as a
function of waste feed rate for the two categories of wastes for which capi-
tal costs were given earlier in this section:
Curve 1 - Wastes with viscosities greater than 750 SSU but less than
10,000 SS'J at ambient, temperature.
Curve 2 - Wastes with viscosities greater than 10,000 SSU at ambient
temperatures.
As with the capital costs, provisions for heating and insulating the storage
tanks are not included in the O/M costs for wastes with viscosities greater
than 10,000 SSU but are reported separately in the following table. Wastes
with viscosities between 750 and 10,000 SSU at ambient temperature do not
require storage tank insulating and heating.
6-35
-------
FIGURE 6.10
U/m CusiTa rOiv Thi.Ki-i«L TREATMENT
I
0
u
Curve
10
100
Waste Flow, GPH
100C
lurve ' - Pumo&ble at ambient temperatures but must be heated for gooc
atoioizatior.
2 - Must be heated to Xeep pumpable and to obtain good atociizatior;,
6-36
-------
Operating and Maintenance^ Cost - Thermal Treatment
Tank
SiZi
(gal)
500 470
1,000 890
2,000 1,710
5,000 2,190
10,000 3,900
20,000 5,380
Water and Solids Removal 0/M Costs
The 0/M costs for removing water from waste streams containing 10, 50,
and 75% water are given in Figure 6.11. These cos. ;s assume that the water
content is reduced to 5% betore the waste is pumped to the boiler. The re-
moved water is assumed to be sent either to the plant's wastewater treatment
unit or to the sewer.
The solids removal costs are given in Figure 6.12. Disposal of tne
collected slurry is by far the dominant c/M cost as discussed above.
Boiler Modification 0/M Costs
Burner Modification 0/M Costs
The burner gun assembly and the associated controls to fire waste fuel
should require little operator attention beyond that required for the boiler
firing conventional fuels. Maintenance cost will be about 5% of capital
cost, capital recovery 13.2%, and taxes and insurance 2%. Therefore, total
annual O/M cost will be about 20.2% of the original burner gun assembly cap-
ital cost.
Processs Control Instrumentation
..o i 'ditional operating labor will be required to operate the process
control _ crumentation. Maintenance and fixed costs are shown below.
Boiler Heat Capital 0/M
Input Cost Cost
106Btu/hr $ $/yr
< 30 5,000 1,000
> 30 25,000 5,000
6-37
-------
FIGURE 6.11
ANNUALIZED 0/M COSTS FOR WATER REMOVAL
1000 ^?
Q
r 10C
100
Waste Plow, GPH
1000
6-38
-------
FIGURE 6.12
ANNUALIZED 0/M COSTS 5X)R SOLIDS REMOVAL
1 000
x
*«
IS
X
O
V
ti
TOO
10
10
100
Waste Flow, GPH
1000
6-39
-------
Oxygen and Carbon Monoxide Monitoring O&M Costs
Annual oparatino costs fcr severa.'. continuous monitoring systems hove
oeen eati-nated for evaluating the economic impact of NSPS for industrial boil-
ers (Raference 1-4). These costs should also be applicaole to "clean" exhaust
gas applications. An annual 0/M cost of $18,^00 in 1982 dollars we.s estimated
for an oxygen monitorin9 system. Costs for a carbon monoxide system should oe
about the same. 7h-j major ite.,ns n.cluded in this estimate are tne maintenance
and performance certification. One-half raanhour per day was assumed to be re-
quired for -he maintenance at a rate, of 535.81/manhour, including supervisor
and cverhead. One certification test per year, costing $1i,900 was assumed.
Requiring 02 and CO monitors on industrial boilers burning hazardous
waste could result in a fuel cost savings to the operators of these devices.
This potential saving would rasult if the operators used the CO/02 monitor3
to maintain low excess air (LEA) combustion of the fuels. With LEA combus-
tion, less fuel is required because less heat is lost out the stack with tne
combustion gases.
The aagnitude of the potential fuel savings that can be obtained by LEA
combustion must be determined individually for each boiler because it depends
on many factors. The major factors influencing the potential savings include:
o toiler type and condition
o Burner type and condition
o Combustion control type and condition
o Operating load level
The boiler type and condition have a large impact on the amount of fuel
saving that may be achieved through LEA combustion, ooirs types have design
characteristics that limit the range of LEA operation. Also, the flue gas
exit temperature for one type boiler can b« significantly different from
those of another type. Since the fuel savings for a given excess air reduc-
tion is temperature dependent, boilers with higher exit flue gas temperatures
should be capable of achieving a higher fuel savings per unit excess air po-
tential of LEA. The condition of the boiler also impacts the fuel savings
potential of the LEA. A boiler that has significant air in-leakage is more
difficult to operate at low excess air levels because the air infiltration
may distort the 02 reading drastically.
The type and condition of the burner(s) installed in the boiler also
greatly influence the fuel savings potential of LEA operation. A burner is
designed to operate efficiently over a specific excess air range. If oper-
ated at an excess air range lower than the design level, proper mixing of
the fuel and combustion air cannot be achieved. Poor air and fuel mixing
would likely result in incomplete combustion of the fuel and higher fuel
consumption. Gas burners generally operate at lower levels than coal burn-
ers. There is also a wide variation in the excess air level operations
capability of burners for a given fuel. The condition of the burner also
affects the potential fuel savings because the fuel flow through a dirty
or damaged burner is difficult to control.
6-40
-------
Another important factor determining the potential of LEA combustion and
hence fuel savings is the type and condition of the combustion controls. Com-
bustion controls vary widely in comp'^xity from the simple single-point posi-
tioning units typically found on smaller units to the metering s"stem of a com-
plex, computerised process control system. Interfacing the CO/Oo monitor to
these systems has limitations that are unique to each type of control system.
The level of LEA achievable is limited to how well the CO/C>2 monitoring is
used oy the control system. Also, the condition of control system mechanical
components also impacts the fuel savings potential. Damper linkage may flex
slightly, and bearings may ••'ear over time. Even metering systems are suscept-
ible to some shortcomings, since their flow transmitters are operated at tem-
peratures and pressures that vary significantly from those at which the trans-
mitters were initially calibrated.
More excess 02 is needed at low loads because of poorer mixing of the
fuel and air. Consequently, the operating load level also impacts the fuel
savings of LEA combustion.
Because of the influence of the factors discussed above, a detailed break-
dowr of potential tuel savings by boiler type and fuel is deemed unjustified.
For the purpose cf determining the economic impact of requiring Oj/CO monitor-
ing, a 2% savings is estimated as a typical average value. This estimate is
based on discussions with combustion control equipment suppliers, information
foi.nd in the literature, and data on oxygen levels and boiler exit flue gas
temperatures. These savings are also based on an estimated 0.5% increase in
combustion efficiency per 1% reduction in the 02 level in the flue gas.
Net all boiler operators would receive the fuel saving estimated for LEA
operation as a consequence of requiring Oj/CO monitoring. Some boilers are
already equipped with 0% trim or O2/CO-trim. In fact, those boilers equipped
with CO-trim systems may actually be required to operate at a higher LEA level
than they are currently operating at, depending on the level of CO limits im-
posed, and thus would consume more fuel. Boilers with CO-trim systems typicil-
ly are operated with a CO setpoint of from 200 to 400 ppm. A CO limit lower
than this range would require that they operate at a hicher LEA level than
their current setpoint level. Also, some boilers are equipped with 02-trim
systems. Thus, estimating the fuel savings potential cf requiring O2/CO mon-
itors on a particular population of boilers requires a knowledge of how many
units in the population already are operating at LEA levels (i.e., how may
are using 02 or CO monitors to achieve LEA combustion.) Once the fraction
of boilers currently employing LEA controls is determined, the potential fuel
savings for the entire population may be estimated from the potential savings
of a single unit.
The savings of a single boilei: can be estimated by multiplying the total
annual fuel cost by 0.02 or 0.04 depending on the typt fuel burning device be-
ing considered. The annual fuel c^st is estimated by multiplying the design
heat input by the unit fuel cost presented in the preceeding paragraphs. For
example, the maximum potential annual fuel savings of a 150 x 106 Btu/hr, re-
sidual oil-fired boiler would be:
.02 x 150 x TO6 Btu/hr x 8760 hrs/yr x S4.62/106 Btu * $121,400
6-41
-------
This assumes that the boiler operates 24 hours per day, 365 days a year.
This savings can be adjusted to .aatch different assumptions .'egard:.ng load
factor.
A comprehensive survey of boilers was not conducted to determine the
fraction of units already equipped with an LEA capacity.
Air Pollution Control Devices 0/M Costs
The annualized 0/M costs for e^ectrcstactic precipitatcrs, baghouses,
and combined venturi/gas absorption systems presented j.n Figures 6.13 through
6.15 are based on 8700 hr/yr operating time and the cost factors presented in
Table 6.3. The annualized costs given in Figure 6.15 are tor a combination
venturi/gas absorption sy.stem. When no venturi is required, t.he 0/M costs
are approximately 85% of that given by Figure 6.15. The 0/M oosts include
direct costs such as operating labor and materials, maintenance, r-,.iacement
parts, utilities, and collected particulate disposal. AA<3O included are in
direct costs such as overhead, insurance, taxes, and capital recovery. Cost
factors presented in Table 6.8 were estimated from information contained in
Reference 16 and represent 1977 dollars. Methods for updating these costs
to the year finally selected for indexing the economic impact analysis are
detailed in Reference 16, The annualized 0/M costs for the three types of
control devices presented in Figures 6.13 through 6.15 are in 1980 dollars
(Reference 23).
Estimated operating labor requirements for APCD systums (Reference 16)
are 0.5 to 2 manhours per shift for electrostatic precipitators, 2 to 4 man-
hours per shift for baghouses, and 2 to 8 manhours per shift for wet scrub-
bing systems. The only utility requirement for baghouses and electrostatic
precipitators is electricity. For baghouses, the power requirement is ay-
proximate iy 0.2 kwh per 1000 ft2 for the reverse air far. motor. The power
requirement of energizing the plates of a precipitator :.s approximately 1.5
watts per square foot of collection area deference 20). For the scrubber
system, the ID fan is the major electricity consuming item. The following
formula was used for calculating the electric requirement of the fan (Refer-
eace 16):
fcwh - 0.746(CFM)(6PJ(SG)H
6356n
where: kwh » icilowact-hour^
CFM * actual volumetric flow rate, acfm
jP » pressure loss, inches WG
n * efficiency, usually 60%
H « hours of operation
The scrubber also requires water which must be treated to remove solids and
neutralize its collected HC1.
6-42
-------
FIGURE 6.13
ANNUALIZED 0/M COST OF ELECTROSTATIC t-FECIPlTATORS,
CONSTRUCTION
<
O
>:
<
! NOT! A
I MOT! :
n • COLLECTION' EFFICIENCY
*OTt t. • FOR DUST HAVING HIGh R[
NOTE B • FOR DUST HAVING MODERATE TC < CV. R£S!STIVIT>
100
1000C
100 OX
EXHAUST GAS RATE, acfrt
6-43
-------
•ICURE 6.14
ANNUALIZED O/M COST OF REVERSE AIR FABRIC FILTERS,
CARBON STbt_ CUNbTRUCTION
O
O
10. oo:
100.000
EXHAUST GAS RA.lt. acfrr
6-44
-------
FIGURE 6.15
ANNUALIZED 0/M COST OF A COMBINATION
VttviVRI/GAS ABSORPTION SYSTEM
I I L 1 ! .
10 OK-
100000
EXHAUST GAS RATE, acfir
."I":'
6-45
-------
TABLE 6.3
COMPONENTS OF ANNUALIZFD COSTS
(Reference 16)
Direct Operating Costs
Cost Factor3
Operating labor
Operator
Supervisor
iMaintenance
I,abor
Material
Utilities
Electricity
Water treatment and cooling water
Waste disposal
$7.87/manhour
15% of operator
38.66/manhour
100 % of maintenance labor
$0.0432 kwh
$0.2500/1000 gallons
$10.00/ton
Indirect Operating Cost
Overhead
Property tax
Insurance
Administration
Capital recovery cost
80% of operating labor and main-
tenance labor
1% of capital costs
1% of capital costs
2% of capital costs
0.16275 (as an example of xf%
and equipment life of xf years)
a All costs are in December 1977 dollars.
6-46
-------
REFERENCES
1. Schoel, C.L., and Hammaker, G.S. (Development Planning and Research
Associates, Inc.) Administrative Compliance Cost Elements and Unit
Costs for Potential Hazardous Waste Fuel Regulations. Prepared for the
U.S. Environmental Protection Agency. EPA Contract 6&-01-6621. March
1984.
2. Energy Information Administration. Monthly Energy Review. Washington,
D.C., Publication DOE/EIA-0035 (83/1213]). December 19C3.
3. Energy Information Administration. Coal Production - 1982. Washington,
D.C., Publication DOE/EIA-0118 (82). September 1983.
4. Memorandum from T. Hrgan, Energy and Environmental Analysis, Inc., to R.
Short. EPA:EAB. June 22, 1983.
5. Reserved
6. Telecon - David Schnell, AquaChem, Inc., Dec. 8, 1983.
7. Telecon - Joe Burkehart, Dedert Corporation, Dec. 7, 1983.
8. Telecon - Paul Schuelke, Kewanee Boiler Co., Dec. 8, 1983.
9. Meeting - William H. Axtman, Russell N. Mosher, American Boiler Manufac-
turers Association, Arlington, Virginia, January 25, 1984.
10. Telecon - John Kirkland, Heat Combustion Engineers, Inc., December 7,
1983.
11. Telecon - Edward Sabol, North American Manufacturing Co., December 15,
1983.
12. Engineering Handbook for Hazardous Waste Incineration - Cincinnati, Ohio.
U.S. Environmental Protection Agency, EPA SW-889. 1981.
13. Memorandum from E.B. Rashin to Larry G. Jones. Availability and Use of
Automatic Excess Air Trim Systems. Standards Development Branch, U.S.
Environmental Protection Agency. January 31, 1983.
14. Dickerman, J.C. and Kelly, M.E. "Issue Paper- Compliance Monitoring
Costs." Radiar Corporation. Durham, North Carolina. September 25,
1980.
15. 1984 Temperature Measurement Handbook. Omega Engineering, Inc. Stam-
ford, Connecticut. 1984.
16. Neverill, R.B, Capital and Operating Costs of Selected Air Pollution
Control Systems. SARD, Inc. Miles, Illinois. EPA-450/5-80-002. De-
cember 1978.
6-47
-------
17. McConnick, R.J., and DeRosier, R.J. (Acurex Corporation). Capital and
O&M Cost Relationships for Hazardous Waste Incineration. Prepared for
the U.S. Environmental 5'rotection Agency. EPA Contract 68-02-3176 and
66-03-3043. July 1983.
18. Technology Assessment Report for Industrial Boiler Applications: Parti-
culate Collection. Prepared oy GCA Corporation for tne Industrial En-
virormental Research Lab, U.S. Environmental Protection Agency. Report
No. EPA 600/7-79-178h. December 1979.
19. McCormick, R., et al. Engineering Analysis of the Practice of Disposing
of Hazardous Waste in Industrial Boilers. Contract No. 68-03-3043, U.S.
Environmental Protection Agency, Cincinnati, Ohio. January 1982.
20. Chilton, C.Ii. Cost Engineering i . the Process Industries. McGraw-Hill,
New York, 1980.
21. Peters, ".S., and Timmerhause, K.D. Plant Design and Economics for
Chemical Engineers. McGraw-Hill, New York, 1980.
22. Comparative Evalution of Incinerators and Landfills for Hazardous Waste
Management. Prepared by Engineering-Science for the Chemical Manufac-
turers Association, May 1982.
23. Control Techniques for Particulate Emissions from Stationary Sources,
Volume I. Prepared for Emissions Standards and Engineering Division,
U.S. Environmental Protection Agency. Report No. EPA-4SO/3-81-005a.
6-48
------- |