EPA-SW-72-3-3
RESEARCH & STUDIES
CORROSION STUDIES IN
MUNICIPAL INCINERATORS
SOL/D l^/\S7E RESEARCH LABORATORY
OFFICE OF RESEARCH AND MONITORING
NATIONAL ENVIRONMENTAL RESEARCH CENTER
CINCINNATI, OHIO
-------
-------
SW-72-3-3
CORROSION
STUDIES
IN
MUNICIPAL
INCINERATORS
P.O. MILLER, the Principal Investigator,
with others from the staff of
Battelle's Columbus Laboratories,
performed the investigation and prepared this report
for the Solid Waste Research Laboratory of the
National Environmental Research Center, Cincinnati
under Research Grants EP-O0325 and EP-OO325-S1
U.S. Environmental Protection Agency
Office of Research and Monitoring
National Environmental Research Center, Cincinnati
1972
-------
REVIEW NOTICE
The Solid Waste Research Laboratory of the National
Environmental Research Center, Cincinnati, U.S. Environmental
Protection Agency, has reviewed this report and approved its
publication. Approval does not signify that the contents
necessarily reflect the views and policies of this laboratory
or of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
The text of this report is reproduced by the National
Environmental Research Center, Cincinnati, in the form received
from the Grantee; new preliminary pages and cover have been
supplied.
n
-------
ABSTRACT
Although the water-wall incinerator permits operation of air
pollution control devices and provides hot water or steam as a
byproduct, there are unpredictable, severe losses of metal from
the furnace tubes when these incinerators are operated.
In this combined field and laboratory study, corrosion probes
of 11 alloys were placed in two municipal incinerators, one refrac-
tory lined and the other a water wall. After the probes were
exposed to varying temperatures, times, and conditions, some of the
conclusions included: that the metal-wastage rates are temperature
dependent; that water-wall incinerators should not be used to
generate high-temperature superheated steam; that the hydrogen
chloride and sulfur dioxide in the flue gases were of sufficient
concentrations to be of concern and account for the types of
deposits found on the tubes; that the roles played by the sulfur-
and chlorine-containing compounds in the refuse are of great
importance and are closely interrelated; that zinc and lead further
complicate the corrosion process; that non-metallic materials
should be considered as coating for wet scrubbers of incinerator
gases; that corrosion of incinerator grates is not a serious
problem.
-------
-------
FOREWORD
To find, through research, the means to protect, preserve,
and improve our environment, we need a focus that accents the
interplay among the components of our physical environment - the
air, water, and land. The missions of the National Environmental
Research Centers - in Cincinnati, Ohio, Research Triangle Park,
North Carolina, and Corvallis, Oregon - provide this focus. The
research and monitoring activities at these centers reflect multi-
disciplinary approaches to environmental problems; they provide
for the study of the effects of environmental contamination on
man and the ecological cycle and for the search for systems that
prevent contamination and recover valuable resources.
Man and his supporting envelope of air, water, and land must
be protected from the multiple adverse effects of pesticides, radia-
tion, noise, and other forms of pollution as well as poor management
of solid waste. These separate pollution problems can receive inter-
related solutions through the framework of our research programs -
programs directed to one goal, a clean livable environment.
This study, published by the National Environmental Research
Center, Cincinnati, reports on metal wastage in water-well
incinerators. The increasing use of this type of incinerator makes
it advisable to anticipate problems that may occur before large-
scale use of these incinerators begins. Personnel on the combined
field and laboratory study obtained information on deposit and
corrosion problems and made recommendations on alleviating them,
and they also gathered information on corrosion problems associated
with wet scrubbers-- all necessary to better cope with the disposing
of solid wastes in our Country.
ANDREW W. BREIDENBACH, Ph.D.
Director, National Environmental
Research Center, Cincinnati
-------
-------
TABLE OF CONTENTS
Page
OBJECTIVES 1
INTRODUCTION 1
SUMMARY 2
REVIEW OF INCINERATOR PLANT OPERATION 7
Corrosion at Oceanside, Merrick, and Norfolk 7
Boiler Tube Reliability Experience at Other U. S. Incinerators .... 8
Mayson Incinerator - Atlanta, Georgia 8
Miami Incinerator - Miami, Florida 8
Lincoln and Green Bay Incinerators - Milwaukee, Wisconsin ... 8
South-west Incinerator - Chicago, Illinois 9
Corrosion at Mannheim, Stuttgart, Munich, and Rotterdam 9
FIELD STUDIES OF INCINERATOR FIRESIDE CORROSION 11
Corrosion-Probe Studies 11
Design and Construction 11
Specimens 15
Operation in Incinerators 15
Miami County, Ohio 18
Norfolk, Virginia. 18
FIELD CORROSION RESULTS 21
Procedures . 21
Area 1 at Miami County 24
Area 2 at Miami County . 27
Norfolk 30
CHEMICAL SPECIES RESPONSIBLE FOR CORROSION 32
Furnace-Gas Compositions 32
Methods of Measurement . 32
Results and Discussion 33
Relation of Solid Waste to Gas Composition 34
Analysis of Deposits 36
Miami County Incinerator Deposits 36
Norfolk Incinerator Deposits 42
Sulfide and pH Tests 47
Electron-Microprobe Analyses 49
X-Ray Diffraction Studies . 53
CORROSION FROM INCINERATOR DEPOSITS UNDER HUMID CONDITIONS . . 57
Stress-Corrosion Cracking 57
-------
TABLE OF CONTENTS
(Continued)
Page
LABORATORY CORROSION STUDIES 66
Apparatus and Procedures 66
Results 67
Gas-Phase Reactions 67
Sulfate-Chloride Mixtures at 1000 F 67
Incinerator Deposits 70
Corrosion at 800 F . 71
Corrosion at 600 F 72
DISCUSSION OF CORROSION MECHANISMS 75
Effect of Chlorides and Metal Salts 75
Effect of Sulfur Compounds . 78
CONCLUSIONS . 81
General . 81
Tube Wastage 81
Incinerator Operation 82
INCINERATOR WET SCRUBBER CORROSION . 84
Corrosion Status Survey 84
Scrubber Corrosion Studies 88
Procedures .......... 88
Results From North Montgomery County 89
Results From Columbus, Ohio 100
Fan Deposit Studies at Columbus, Ohio 100
Deposit Analyses , 104
Corrosion 104
Conclusions From Scrubber Studies 110
ACKNOWLEDGMENTS 112
PUBLICATIONS RESULTING FROM THE RESEARCH 113
REFERENCES 114
APPENDIX
Probe-Specimen Handling A-l
Experimental Results A-l
Miami County A-l
Norfolk A-l
Laboratory A-l
-------
FIRESIDE METAL WASTAGE IN MUNICIPAL INCINERATORS
(Research Grants EP-00325 and EP-00325S1)
by
P. D. Miller, W. K. Boyd, R. B. Engdahl, R. D. Fischer,
H. H. Krause, W. T. Reid, J. E. Reinoehl, E. J. Schulz,
D. A. Vaughan, P. R. Webb, and J. Zupan
OBJECTIVES
The objectives of this research program are to identify the conditions that lead to
fireside metal 'wastage in solid-waste incinerators, to determine the mechanisms by
which metal loss occurs, and to devise means of preventing corrosion in these
installations.
The objectives of the supplemental program are to assess the corrosiveness of
the environment at incinerator grates and in scrubber chambers and to evaluate mate-
rials of construction.
INTRODUCTION
It is becoming apparent that more and more attention is being given to the use of
incineration as a means of disposing of municipal refuse. This incineration must be
carried out in a manner that does not create pollution problems. The use of pollution-
control devices requires that the incinerator furnace gases be cooled to permit operation
of the flue-gas cleaning devices. An economically attractive way in which to do this is
by absorbing much of the heat by water contained in wall tubes and in convection-pass
tubes used to form the furnace. Alternate procedures are to dilute the hot gases with
cool air or with water. However, the latter practices greatly increase the gas volumes
to be handled and thus increase the size of the pollution-control equipment.
In addition to greatly decreasing the size requirements for pollution-control equip-
ment and fans, the water-wall incinerator has several other attractive features. First,
there is a gainful use of the heat energy available in the refuse. Second, the furnace
throughput can be increased because of the rapid absorption of heat possible. Third,
wall slagging problems often encountered in refractory-type construction are absent.
The technology of the water-wall incinerator was developed first in Europe where
it is in fairly extensive use. Corrosion problems have been reported, however, in some
instances. Details of instances of severe metal wastage in incinerators along -with other
data on incinerator operation can be found in References (1) to (31). The first operating
water-wall unit on this continent was at the Navy Public Works Center in Norfolk,
Virginia, beginning in 1969. Units are now operating in Montreal, Quebec, Braintree,
Massachusetts, and northwest Chicago. It is anticipated that this type of construction
will become used more and more. Thus, it is important to anticipate problems which
may arise before large-scale use begins.
-------
In March, 1969, research was started at Battelle- Columbus on a grant program
supported by the Solid Waste Management Office, EPA, which was aimed at determining
the cause and extent of fireside metal wastage in incinerators and devising methods of
alleviation. Progress on that task has been reported in detail in Annual Reports dated
April 24, 1970; April 23, 1971; and April 27, 1972.
In March, 1971, work -was started on a supplemental program, aimed at obtaining
a better understanding of incinerator-gas scrubber corrosion and also of metal wastage
of grates. That work was reported in detail in the Annual Report dated April 27, 1972.
The present report draws together the research techniques used in studying the
problems mentioned in the preceding paragraphs and also summarizes the conclusions
reached. It covers the period March 1, 1969, through February 29, 1972.
SUMMARY
One problem faced in the operation of some European incinerators using water-
cooled tubes in the walls is the unpredictably severe loss of metal from the furnace
tubes. Since usage of large incinerators is increasing in the United States, it is im-
portant that the problem be understood so that techniques for its mitigation can be
introduced. Of added importance are the facts that this type of incinerator is amenable
to pollution control and that, at the same time, it provides hot water or steam as a
by-product.
One objective of this research was to obtain information on the deposit and corro-
sion problems encountered in municipal incinerators which utilize waste-heat boilers
and then to recommend practical and economical alleviation procedures. A second
objective was to gain a better understanding of the corrosion problems associated with
wet scrubbers used in pollution control for incinerators.
The research is a combined field and laboratory study. The field -work has in-
cluded a review of the corrosion aspects of many currently operating incinerators in
this country and in Europe and, in particular, includes the use of corrosion probes in-
serted in operating units. These probes have been placed in a refractory-lined mu-
nicipal incinerator at Miami County, Ohio, and in a water-wall incinerator operated by
the Navy Public Works Center at Norfolk, Virginia. Twelve corrosion probes were
exposed at Miami County and three were exposed at Norfolk. Exposure time at tem-
perature ranged from 122 to 1318 hours. The eleven alloys listed below were used to
form the individual specimens which made up the corrosion probes. Not all alloys were
used on all probes.
-------
PERFORMANCE OF ALLOYS IN FIRESIDE AREAS'!
Resistance to Wastage
Alloy
300-600 F
Good
Good
Good
Good
Good
Good
Good
Good
Fair
Fair
Fair
600-1200 F
Fair
Fair
Fair
Fair
Fair
Fair
Poor
Poor
Fair
Poor
Poor
Moist Deposit
Good
Pits
sec**
sec
sec
sec
Pits
Pits
Pits
Fair
Fair
Incoloy 825
Type 446
Type 310
Type 316L
Type 304
Type 321
Inconel 600
Inconel 601
Type 416
A106-Grade B (Carbon Steel)
A213-Grade Til (Carbon Steel)
* Arranged in approximate decreasing order.
** Stress-corrosion cracking.
Metal wastage rates were shown to be temperature dependent; for the carbon
steels they ranged from about 10 mils per month at 325 F to about 35 mils per month
at 950 F. These values, however, should not be projected to exposures of many months
or years since rates decrease with extended exposure. It is estimated that rates for
carbon steel tubes at about 500 F might be expected to range below about 10 mils per
year for extended exposure periods.
The thermal gradients between the tube metal and the hot gases were found to be
important in addition to the metal temperature.
The tabulation shows the relative performance of the eleven materials studied.
As an approximation it can be assumed that "poor" is >20 mils per month, "fair" is
10-20 mils per month, and "good" is <10 mils per month. It can be seen that the
Incoloy 825 gave encouraging results over the entire temperature range and is much
more resistant to stress corrosion cracking than the austenitic stainless steels. The
stainless types, except Type 416, lose less weight than the carbon steels at comparable
temperatures, with Types 310 and 446 being superior. The major limitation for these
stainless alloys is stress-corrosion cracking or pitting when in contact with moist de-
posits as during downtime. The Inconel 600 and 601 materials were resistant over the
lower temperature region but were severely attacked at high temperatures.
While aluminum, chromium, and inorganic coatings furnished some initial pro-
tection to the tube surfaces they do not appear to be sufficiently durable to be con-
sidered for long-time boiler operation.
It is concluded that water-wall incinerators should not be used to generate high-
temperature superheated steam but should be operated at relatively low metal tempera-
tures, near 500 F, to minimize tube wastage. Steel tubes of compositions such as
A106-Grade B or A213-Grade Til can be considered at these temperatures. The
Incoloy 825-type alloys would be better choices at higher temperatures.
The composition of the deposits removed from the corrosion specimens was care-
fully determined to provide a clue as to possible causes of the corrosion. Significant
amounts of chloride, sulfur, lead, zinc, potassium and calcium, along -with other more
-------
inert elements, -were detected in the deposits from the two incinerators mentioned and
also in incinerators at Oceanside, Long Island, and Atlanta, Georgia. Another interest-
ing observation was that the lead content in the deposits at Norfolk and at Atlanta was
unusually high.
The proportions of the various chemical compounds found on the tube surfaces
depended on the probe-specimen temperature, and the distributions were consistent
with known volatilities of the compounds.
Composition of the flue gas was determined at the first three sites mentioned. The
HC1 and SO2 were present in sufficient concentrations (10-300 ppm) at all sites to be of
concern and to account for the types of deposits found on the tubes.
X-ray-diffraction and electron-microprobe techniques were used to examine cross
sections of corroded tubes and deposits. It was found with the microprobe that the ele-
ments chlorine, sulfur, lead, zinc, and potassium were concentrated adjacent to the
base metal. The X-ray-diffraction studies identified over 20 compounds in the scale and
deposit. Compounds of particular significance are FeCl2, FeS, KC1, NaCl, (NaK)2SC>4,
ZnSO4, PbSO4, and mixed PbO PbSO4 salts.
The sodium azide spot test also revealed that sulfide compounds were present near
the metal surface on almost all the probe specimens, particularly on those exhibiting the
most severe corrosion.
Laboratory studies were carried out at 600, 800, and 1000 F using a variety of
salt compositions and under a flue-gas atmosphere containing SO2, HC1, and, on
occasion, formic acid HCOOH. The data indicate that chloride salts added to sulfate
salts significantly enhance corrosion, particularly at 800 and 1000 F. The SO2 also is
an important constituent in this system. At 600 F, the most corrosive salts in decreas-
ing order of activity were KHSO4, K2S2Oy, and ZnCl2. At 800 and 1000 F, ZnCl2 and
PbCl2 were very corrosive.
Correlation of the results of the different studies made on the deposits and cor-
roded surfaces combined with the laboratory results provide at least a tentative explana-
tion of the corrosion mechanism. It is proposed that HC1 and C1-, released adjacent to
the tube surfaces are important factors. In addition, SO2 and SO^ gases, along with
sulfur-containing compounds, cause additional corrosion.
The roles played by the sulfur- and chlorine-containing compounds in the refuse
are of great importance and are closely interrelated.
Chloride salts reach the tube surface by direct volatilization in the flame and by
reaction of the HC1 formed during burning with volatilized K2O and Na2O. The chloride
salts deposited on metal surfaces then react with SO2 and oxygen near the tube to evolve
high concentrations of HC1 directly adjacent to the metal so that FeCl2 is formed.
Additional FeCl2 is formed by reaction of iron and chlorine. The chlorine is formed
when portions of the HC1 are oxidized to C12 is a catalyzed reaction with oxygen. This
reaction is confined to areas near the metal -where Fe2C>3 is present to act as catalyst.
The iron chloride in turn can decompose to form Fe2C"3 an<^ additional C12. The iron
oxide scale helps confine the system near the tube. Thus, conditions exist in which
cycling between formation and decomposition of FeCl2 can take place.
-------
The other role played by sulfur is that of forming low-melting pyrosulfates or
bisulfates by reaction of sulfates with the 803 formed near the oxide scale. These low-
melting salts react directly with the steel to form FeS and additional sulfate and iron
oxides.
The corrosion process is further complicated by the presence of zinc and lead
salts which serve to lower the melting points of the mixtures on the metal surface.
Field studies aimed at the second objective, scrubber corrosion, have been car-
ried out at the Montgomery County North Incinerator. Three types of stress specimens,
some of them welded, were exposed in the hot water exiting from the scrubber at the
incinerator. A duplicate set was exposed to the same water at a lower pH at Columbus,
Ohio, under stagnant but aerated conditions. Candidate alloys for these studies were
chosen primarily from those known to be fairly resistant to stress-corrosion cracking.
Pitting attack and selective attack at crevice areas constituted the major form of
metal deterioration at the incinerator scrubber. At Columbus in the low pH solutions
(2.0) stress-corrosion cracking -was observed on Types 304 and 316L stainless steel
specimens. Specimens of Inconel 600 and 601 exhibited what is termed stress-
accelerated corrosion at North Montgomery County.
The alloys most resistant to all types of corrosion of those evaluated were
Ti6Al-4V, Hastelloy C, Inconel 625, Hastelloy G, Hastelloy C-276, Hastelloy F,
Ti75A and Type S-816.
The less-resistant alloys were Carpenter 20, Incoloy 825, Types 316L, 310 stain-
less steels, Inconel 600 and 601, Armco 22-13-5, USS 18-18-2, and Type 304. The
following tabulation lists these 18 alloys and their general resistivity to corrosion in the
scrubber solutions. The Incoloy 800 was not exposed in the scrubber water.
Evaluation of Alloys for Incinerator Scrubbers
Alloy
Ti6Al-4V
Hastelloy C
Inconel 625
Hastelloy F
Hastelloy C-276
Hastelloy G
Ti75A
S-816
Carpenter 20
Incoloy 825
Incoloy 800
316L
310
446
Inconel 600
Inconel 601
Armco 22-13-5
USS 18-18-2
Type 304
Corrosion
Scrubber Solutions
Good resistance
Good
Good
Good
Good
Good
Good
Good
Pitted
Pitted
Pitted, SCC
Pitted
Pitted
Trenches
Trenches
Pitted
Pitted
Pitted, SCC
Results
Fan Deposits
Good resistance
Good
Good
--
Pitted
Pitted, SCC
Pitted
Pitted
Pitted, SCC
--
--
--
Pitted, SCC
Pitted, SCC
Pitted, SCC
-------
The last column in this tabulation shows eleven alloys exposed to damp deposits
from the induced draft fan from N-Montgomery County. Stainless steels Types 304,
316L, USS 18-18-2, Armco 2Z-13-5, and Carpenter 20 were cracked to varying de-
grees. Pitting was noted on the alloys just mentioned as well as on Incoloy 825,
Incoloy 800, and Type S-816. The most resistant alloys studied were Ti6Al-4V,
Inconel 625, and Hastelloy C.
It is believed that there is merit in considering nonmetallic materials as coatings
for construction of wet scrubbers for incinerator gases.
It was concluded that corrosion of incinerator grates was not a serious problem.
-------
7
REVIEW OF INCINERATOR PLANT OPERATION
At the beginning of the research program in 1969, a survey was made to obtain
information on the deposit and corrosion problems encountered in municipal incinera-
tors that utilize waste-heat boilers. Installations operated by the township of Hemp-
stead, Long Island, New York, at Oceanside and at Merrick; and by the Naval Public
Works Center at Norfolk, Virginia, were visited.
The status of other incinerators in this country was determined through telephone
conversations and/or correspondence. In addition, four large municipal incinerators
in Europe were visited.
Corrosion at Oceanside, Merrick, and Norfolk
Severe boiler-tube corrosion was experienced at the Oceanside plant requiring
replacement of many of the tubes in a 2-year period. At the Merrick plant less severe
corrosion has occurred; there, the boiler has been retubed twice in its 20-year history.
No corrosion problems have been encountered during the 5-year operating experience
at the Norfolk plant.
Specimens of boiler-tube deposits and sections of corroded boiler tubes were ob-
tained from the Oceanside incinerator. The deposits were found to contain from 5
to 15 weight percent lead and about 5 percent zinc in addition to the sodium, potassium,
and iron ordinarily found in deposits of boilers fueled with coal or oil. From. 1 to
3. 5 weight percent chloride was found in different portions of the deposits, together
with the usual 20 to 30 weight percent of sulfate. The presence of lead, zinc, and
chloride in the incinerator deposits is the most significant difference noted to date be-
tween these deposits and those commonly encountered in coal-or oil-fired boilers.
Similar observations have been reported by Bryers and Kerekes. (32) Also, it was
found that hygroscopic salts were present on the Oceanside tubes; X-ray diffraction
identified ferrous chloride as a major constituent of these areas.
A metallographic examination was made of corroded sections of two steel boiler
tubes and one stainless steel hanger from the Oceanside incinerator. The wall thickness
in some areas had been reduced from about 125 mils to 34 mils in the tube removed in
1968 and to 83 mils in the tube removed in 1969. There was nothing unusual in the
microstructure of these tubes and no localized attack was evident, except the pitting as
noted. Thus, the examination revealed that the tube wastage is the result of uniform
corrosion. It should be noted that, while the attack is uniform in nature as opposed to
selective, the wastage is more severe in some areas than in others. For example, the
4 and 8 o'clock positions generally suffered the most severe wastage. In fact, some
tubes were actually perforated in other specific areas.
The stainless steel hanger was also examined at the thinnest area (0. 5 in. ) Gen-
eral surface roughening, but no selective attack, that was observed indicated that the
metal wastage was the result of general attack on the hanger surface. The microstruc-
ture of this material was typical of that for an austenitic alloy. That there was no
evidence of sensitization (chromium carbide precipitation at the grain boundaries) in-
dicated that the hanger had not been above 900 F for any appreciable length of time.
-------
Boiler Tube Reliability Experience
at Other U. S. Incinerators
Only a small percentage of incinerators in the United States utilize some of their
waste heat in the form of steam or hot water. Operators of the following steam gen-
erating plants were contacted by phone or in person to ascertain their experience, if
any, with tube corrosion:
Mayson Incinerator, Atlanta, Georgia
Miami Incinerator, Miami, Florida
Green Bay and Lincoln Incinerators, Milwaukee, Wisconsin
Southwest Incinerator, Chicago, Illinois.
None of the above incinerator installations have had significant corrosion in their
boiler units. All of them produce relatively low-pressure steam (200-300 psi) and the
steam temperatures are in the range 300-400 F. In addition, the boilers are far enough
from the incinerator combustion chamber so that flue-gas temperatures probably do
not exceed 1000 F, as compared with temperatures up to 1900 F at Oceanside. This
combination of low boiler-tube and flue-gas temperatures may account for the lack oi
corrosion problems at these sites.
Mayson Incinerator - Atlanta, Georgia
This waste-heat boiler following a Volund-type rotary-kiln incinerator has gen-
erated steam for outside use steadily since 1951 with no problem from corrosion.
Annual revenue from steam and metal scraps is $200, 000. An earlier similar plant
built elsewhere in Atlanta in 1940 had given no problems. Two additional units were
started up at the old plant in 1963 on the basis of the generally good experience. Steam
temperature and pressure are moderate in all of these units. A later section of the
present report discusses the chemical composition of the deposits from this boiler.
Miami Incinerator - Miami, Florida
Generally trouble-free experience in generating 1, 500,000 pounds per day of low-
pressure steam for a nearby hospital has encouraged the inclusion of waste-heat boilers
for the new Miami incinerator. If there have been any corrosion failures, they have
been minimal.
Lincoln and Green Bay Incinerators -
Milwaukee, Wisconsin
Generation of building heat and hot water in waste-heat boilers at the Lincoln and
Green Bay incinerators has been free of corrosion difficulties.
-------
Southwest Incinerator - Chicago, Illinois
The waste-heat boilers following the rotary-kiln-type incinerators at this plant
have been generating steam for outside use since 1963 without serious difficulty from
tube corrosion.
Corrosion at Mannheim, Stuttgart, Munich, and Rotterdam
Travel to England in November, 1969, for Battelle on other matters afforded an
opportunity to visit four large municipal incinerators in Germany and Holland. The
four installations visited were selected because all had reported corrosion problems.
The four installations were remarkably similar on many counts; however, comments
from the operators about possible causes of metal wastage differed considerably.
Prior to these plant visits, a day was spent at Battelle's Frankfurt Laboratories
reviewing the corrosion problem in incinerators. That discussion pointed out the fact
that no one in Europe really understands the cause of the corrosion. The explanations
advanced are based on surmise rather than on facts, and many analyses have been
made of the deposits in corroded areas.
Conclusions from these European visits are as follows:
Corrosion occurs to some degree in all large municipal heat-recovery
incinerators, to an extent and under conditions still generally
unknown.
Attempts in the past to describe the mechanism of the attack have been
based mainly on supposition, using basic premises and similarity with
other systems. More recent studies have aided greatly in providing
a better understanding of the processes involved.
Probes have been placed in some incinerators, but no worthwhile data
have been obtained from them, apparently because no serious attempts
were made to interpret data from the deposits formed on the probes.
The evidence regarding the role of lead and zinc in the corrosion is
sparse.
Erosion is blamed for much loss of metal, generally on the supposi-
tion that products of corrosion removed from the metal surface by
erosion expose fresh metal to oxidation.
Temperature is considered an important factor, with some agreement
that a maximum metal temperature of about 600 F should not be ex-
ceeded. Nevertheless, steam temperature as high as 1000 F is not
uncommon.
Corrosion is often attributed to "flames", based on the observation
that corrosion is found in wall areas where flames are present.
-------
10
Alternate oxidation and reduction caused by intermittent flame con-
tact is the cause of corrosion, according to some workers, on the
basis that Fe2C>3 reacts with HC1 under reducing conditions to form
FeCL,, which then converts to the oxides under oxidizing conditions.
An alternative proposal is that iron is converted to FeCL?, which
changes to FeClo at slightly higher temperatures.
Chlorine from plastics such as polyvinylchloride is blamed only in
part for corrosion, in the belief that much larger quantities than
2 percent of the refuse will be necessary before PVC becomes
troublesome.
Refuse is used to produce electricity, but the quantity is small, and
the efficiency of conversion is low. Such plants are mostly used in
conjunction with an existing power plant of much greater capacity.
Municipal refuse can be burned in combination with other fuels,
such as pulverized coal or residual fuel oil, in a variety of fur-
naces. Many different configurations have operated successfully.
Large amounts of sulfates are found in most deposits in corroded
areas, but the pH of a water solution of such deposits generally
is not low. Apparently, the sulfates could be connected with
fillers used in paper, such as CaSO^.
Combustion conditions are unbelievably bad in large municipal in-
cinerators burning raw garbage. Martin grates, Dusseldorf
grates, and conventional chain-grate stokers all have large masses
of cold refuse in proximity to burning zones. That the final material
dumped from the end of the combustor contains so little unburned
material is remarkable. Agitation within the fuel bed in a Martin
stoker is generally considered beneficial to complete combustion.
Relatively large amounts of material burn in suspension (mainly
bits of paper), because of the very high velocity of the primary air
through holes in the fuel bed where combustion is occurring at a
high rate.
-------
11
FIELD STUDIES OF INCINERATOR FIRESIDE CORROSION
Corrosion-Probe Studies
In order to obtain realistic estimates of tube wastage and deposit formation, it was
necessary to perate probes in the field over extended periods.
The field studies conducted on this program have included:
(1) Installation of corrosion probes in the municipal incinerator operated
by Miami County, Ohio and in the Salvage Fuel Boiler operated by the
Navy at Norfolk, Virginia.
(2) Sampling of flue gas at the Miami County, Ohio, the Oceanside,
Long Island, and the Norfolk incinerators.
(3) Analysis of the boiler-tube deposits from the Oceanside incinerator,
the Atlanta Mayson incinerator, and those built up on the corrosion
probes in the Miami County, Ohio, and the Norfolk, Virginia, incin-
erators.
Design and Construction
The probe was designed to include 34 cylindrical specimens nested together end to
end and then inserted into the incinerator through a side wall. The section of the probe
extending through the wall was water cooled. The specimens exposed within the furnace
were cooled by air flowing inside the tubular specimens. A computer analysis was used
[to ascertain the geometry of the internal support tube required to give the most linear
specimen-temperature variation over the range of about 350 to 1100 F for a probe with
34 specimens. Each specimen was about 1. 25 inches in OD, 1. 00 inch in ID, and 1. 5
inches long. Figure 1 shows an assembled probe positioned with the hot specimen end
down and the water-cooled end up. The corrosion specimens begin at the transition
point just above the left hand of the man holding the probe. The probe is placed in a
horizontal position in the incinerator.
The appearance of individual specimens is illustrated in Figure 2, which shows a
probe being assembled. The four holes visible on the hot-end closure section are pass-
ageways for the exit of the cooling air. Figure 3 is a schematic of the final exposure-
probe apparatus. The specimens are nested together with lap joints as shown in Detail A
and are retained axially at the cooling-air-outlet end by a retainer which is fixed to the
internal support tube with webs as shown in Section A-A. The axial restraining force in
the internal support tube is obtained by compressing the spring on the air-inlet end of the
probe at assembly. Also, the spring compensates for diffential thermal expansion between
the specimens and the internal support tube.
Specimen temperatures are measured at four stations with Type K thermocouples
either welded into the wall of the specimens or inserted into recesses drilled lengthwise
into one end of the appropriate specimen.
-------
FIGURE 1. CORROSION PROBE AFTER REMOVAL OF DEPOSIT
Cold end at top. Specimens begin just above the left
hand of the operator and extend downward.
-------
in
ii
r*
m
W
CQ
O
ft
O
O
U
O
W
CP
CO
W
U
W
CO
-------
t/3
P
H
w
CQ
O
rt
PH
I
w
g
X
w
u
O
u
w
w
U
CO
ro
H
P
O
-------
15
The thermocouple lead wires are brought out of the probe through the center of the
internal support tube, so that temperatures can be recorded continuously on a strip-
chart potentiometer recorder. Since the computer results indicated that the temperature
variation is linear for regions with a constant gap between the internal support tube and
the specimens, the four temperatures accurately define the specimen temperatures.
The specimen temperatures are controlled by regulating the amount of cooling air
admitted to the probe. The output from a control thermocouple, which is attached to the
specimen at the same axial location as the Thermocouple 3, is monitored by a propor-
tional temperature controller. At the start of a test, the controller setpoint is adjusted
to maintain either the lowest temperature at Thermocouple 1 or the highest temperature
at Thermocouple 4, depending on the test objective. The controller maintains this tem-
perature by varying the amount of cooling air bypassing the probe through a motorized
butterfly valve located between the blower and the probe. A Roots-blower air pump
delivering up to about 34 cfm was used with a 5-hp motor as a drive.
Specimens
Eleven different metals and alloys and five different types of coating were evalu-
ated in the field probe studies. Carbon steels Al06-Grade B and AZ13-T11, a low-alloy
Cr-Mo steel, were used in the lower temperature regions, i. e. , not over 1000 F. Types
304, 310, 316, 321, 416, and 446 stainless steels were exposed primarily at the higher
temperature regions but some were exposed at lower temperatures to provide a direct
comparision with the carbon-steels. The nickel-base alloys, Inconel 600, Inconel 601,
and Incoloy 825 were also used to form specimens for use at several locations over the
length of the probe.
The nominal compositions of these alloys is given in Table 1. This table also in-
cludes the composition of additional alloys which were evaluated in the wet scrubber as
discussed in later sections of this report. As can be seen in the table, some alloys were
evaluated in both fireside areas and in the scrubber.
Specimens protected with several types of coatings were included in the fireside
probe studies. One set was coated by Super chrome Plating and Engineering Company,
Inc. , with about 5 mils of hard chromium. Such coatings had provided protection to
superheater tube areas in oil-fired power stations. A second set was coated by Armco
Steel Corporation with aluminized Type 1 coating to a thickness of about 2 mils. Such
aluminized coatings have proved useful in flue gas or exhaust gas atmospheres. Three
types of inorganic coatings (cermets) were also evaluated. The inorganic coatings
(W, 570W, and 570+W) were applied by the Sermetel Division of Teleflex, Inc. , and
represented types used to protect turbine blades during high-temperature operation.
These coatings were applied to Al06 and T-ll steel specimens.
Operation in Incinerators
The helpful and cooperative managements of the incinerator at Miami County, Ohio,
and at the Navy Public Works Center at Norfolk, Virginia, permitted insertion of the
probes in two routinely operating incinerators. The program at Miami County began
early in December, 1969. Since two self-sufficient probe assemblies were constructed,
it was possible to operate at Norfolk, Virginia, concurrent with operation at Miami
County. The work at Norfolk began in August, 1970. Table 2 shows the exposure
-------
16
CO
5
CO
Da
i
u
CO
Q
2
<
S
CO
£a
Uj
2
Q
CO
D
CO
O
j
<
O
z
o
CO
0
a.
0
U
j
2
S
0
2
T-H
CQ
U
5
^
o
u
<
u
S
.,
CO
1)
I
£
U
^
S
J2
.0
a
U
CO
0)
TD CD
'w -^
<" 2
.£ Cu
ft-i
rj
"^
a
LO LO . LO ^
w. -; °. ! °. i
ooo o U
LO O IO LO LO
rH CO O rH rH
i 1 1 £ i
1 ' i-H ' '
LO
, , (N ( 0
1 ' 0 ' (N
, , , ° I
1 ' ' rH '
">< 'x'
ctf nj
§ £
LO IO IO ,
C- C- (N 0 J
O O O i I
rH O O
CTj ctf -^ i I O
CQ CQ rH (N CO
, , , O LO
1 ' ' CD CD
">?
cd
E
LO IO LO
o o o LO 10
LO
CO LO CO CN (N
00 (N CM CN -i >H >-« >H >H
u> en co
CO HJ CD O O
£>H J>H ^ 2 2^
^ u, O
CO O CD >-. >>
^ rH ( O O
ex ex o ^ K
^ >^ y rt ed
H H - ffi ffi
X
cfl
C, nJ CT>
CO LO C1 LO CO Q IO
OOOj^CQO^O
tM 0* 0* U CO* 0* H o
LO LO
O CJ LO t-
IO O O O
II II
11 II
1 0 00
LO
ii (N rf
11 CN 0*
11 II
II II
LO LO
O CJ (N LO
r-I O* O O
r 1 LO O O
nj .
CQ CN O CO
O O O t
O3 O^ CO '
O LO
CO iH CN H >H ^ ^H
CO
O O CO O
2 Z >- 2
LO
1
CO
11 IO LO O
' CN CN O
CN CO OO OO
^ t t>^ "^
8 § II
C o O O
-
i o o ;
CM' CM"
i i i
LO
i i i i
i i i i
CO ' ' '
1 1 1
1? s?
§ £
LO IO
o c- c- ,
1
rH o' o' '
CO , 1 r-.
co CQ en co
1 1
O LO
Oi CTJ r-i O
tN rH (N
0 01 05 0^
en OT «)
>-< >-* 2 >-<
Ul V3
O <1) 1) O
py ^H [a-i Z
o
£: o CN
s ^ "
^ & &s
, CO CO . CO LO LO
O O J O O O O
1 ! <=> LO ^ co ; |
r-1 rH O rH O O
"><"
nj
i i E t i i i i t
i t "-^ i i i i i i
CN (N
i 1 1 i 1 i i 1 1 i
1 1 i 1 1 1 1 1 1 1
CO
1 1 1 1 0' 1 1 1 1 1
II "^ 1 1 1 I 1 1
1 1 ^ ^ 1 1 1 1 1 1
">?
£
LQ ^ LO LO
C- ? CN C- i-H ( (
CN o" rH _] O* rH o' O ' '
r-( r I IO LO rH r-I r-H ( ,
CQCOLOLOC30CQCQCQ1
. « II10-!!!
rH rH
" " II,.,
oo CM' CD CD* CD ' ' ' ' '
rH rH LO LO IT-
LO
LO LO CM
CO c- 10 10 10 CO rH j ] j
SSSSSoooSS
>-H>M>H2Z2>-<>H
OS "> tyi VI
OcOOOOJCDUCDOO
2^-<22r~*r>~'^lr~l22
r- rH ^
CM , LO <
'rHOOrtT-H'->-gg
""^^ c^co'c'c
l^lllfi^ln
0
1
i-l
°^
c
M-(
c
u
E
o
"S
u
G
rt
W)
1
£
rt
«_i
a
O
tt
0,
.3
^
.S
13
1)
3
M
03
00
'rt
r-I
s
rt
E
>+-«
O
00
O
C
a)
H
-------
17
TABLE 2. CORROSION-PROBE EXPOSURE SCHEDULE
Probe
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Site
Miami County
Miami County
Miami County
Miami County
Miami County
Miami County
Miami County
Norfolk
Miami County
Norfolk
Norfolk
Miami County
Miami County
Miami County
Miami County
Gas Temp,
F
1200-1500
1200-1500
1200-1500
1200-1500
1800-2200
1200-1500
1800-2200
1000-1200
1200-1500
900-1100
1000-1200
1200-1500
1200-1500
1200-2000
1200-1700
Time Interval
Dec. 2 - Dec. 23, 1969
Jan. 21 - March 2, 1970
March 2 - March 30, 1970
April 21 - April 30, 1970
May 15 - June 1, 1970
July 7 - July 24, 1970
July 20 -July 28, 1970
Aug. 11 - Aug. 28, 1970
Oct. 21 - Jan. 4, 1971
Sept. 14 - Oct. 2, 1970
Dec. 14 - March, 19, 1971
June 6 - July 10, 1971
Aug. 24 - Sept. 8, 1971
Oct. 12 - Nov. 14, 1971
Dec. 12, 1971 - Jan. 4, 1972
Total Time
in Furnace,
hr
506
893
666
219
411
412
188
240
1820
227
2304
816
356
792
576
Time at
Temp, hr
312
389
323
122
190
218
101
180
828
188
1318
492
206
420
291
Heated During
Downtime
No
Yes
No
Yes
Yes
No
Yes
No
No
No
No
No
No
No
No
-------
18
schedule for each of these incinerators. It will be noted that the total time in the furnace
was usually much longer than the time at temperature. Operation at Miami County usu-
ally was for 4 to 5 days a week; at Norfolk, the two furnaces were operated alternately
1 week at a time.
As can be seen in Figure 3, the probe assembly was designed to provide heating to
prevent condensation from the atmosphere during downtime. Table 2 indicates the runs
where this equipment was used (Runs 2, 4, 5, and 7).
Miami County, Ohio. The corrosion probes for Runs 1, 2, 3, 4, 6, 9, 12, 13, 14,
and 15 were installed in the Miami County Incinerator's furnace-outlet sampling port
located over the last half of the bridge wall, as shown in Figure 4, Location 1. This
sketch shows the incinerator arrangement relative to the sampling port. In this location,
the corrosion specimens are exposed to radiation from the flame and they in turn radiate
to the cooler downstream scrubber. Seldom, if ever, did flame from burning refuse
reach the probe; most of the heat transfer to the tube probably was by convection. In
some cases, pieces of burning refuse carried by the flue-gas streams and burning in sus-
pension may deposit on the probe and cause local high temperatures.
No special firing of the incinerator was done for this study. The 150-ton-per-24-
hour-day incinerator operated around the clock on a 4 to 5-day-week schedule. Burning
rates were governed by such factors as the type of refuse received and the moisture con-
tent of the refuse. While the incinerator was being fired, the flue-gas temperature at
the location of the corrosion probe normally ranged between 1200 F and 1500 F. During
furnace-shutdown periods over weekends or during breakdowns, the temperature in the
incinerator dropped below 200 F. Because moisture is usually present in the scrubber
system, humidity in the incinerator during shutdown could be quite high. As mentioned
before, provision was made in some tests to keep the probe warm during this period.
The corrosion probes for Runs 5 and 7 were installed at Location 2 in the same
incinerator. This is directly in the flame area, so that temperatures on occasion reached
2000 F.
Norfolk, Virginia. Corrosion probes for Runs 8, 10, and 11 were installed in one
of the two units of the Salvage Fuel Boiler (incinerator) operated by the Navy Public Works
Center at Norfolk, Virginia. Each furnace can burn 150 tons of refuse per day. The
walls of each furnace, including the main uptake to the convection boiler, are water cooled.
The water walls radiantly heated are 2-1/2-inch diameter on 3-1/4-inch centers. The
upper convection section of the boiler consists of 2- 1/2-inch-diameter tubes on 7-inch
centers in a conventional two-drum design.
Figure 5, taken from a paper by Moore(9), shows the layout of this unit and the
location of the corrosion probes. Probes 8 and 11 were located at the lower position
shown in the convection passes of the boiler. Probe 10 was located about 8 feet higher
up in the same general area. Gas temperatures in these areas were in the range 1000
to 1200 F at the lower location and in the range 900 to 1100 F at the upper location.
The main difference in the type of refuse fired at Norfolk as compared with that at
Miami County was a greater content of wood and smaller amounts of tin cans. However,
the refuse does contain rather large pieces of scrap metal. The refuse at Norfolk, which
is collected from Naval activities and ships, include paper products, crates and packing
materials, plastics, and garbage, as well as the noncombustible materials such as metal
-------
o
H
w
£
t t
u
o
o
A
o
a
r-4
I
00
u
H v
W rt
£4 o
CO CO
w
rt
p
o
I(
h
-------
WATER. COOLED
FEED CHUTE.-*
FIGURE 5. SALVAGE FUEL, BOILER PLANT(9), U. S. NAVAL BASE,
NORFOLK, VIRGINIA, BOILER SECTIONAL ELEVATION
-------
21
and glass. Observations also suggest that the amount of PVC plastic in the Norfolk
refuse is less than that at a municipal incinerator such as the one at Miami County.
FIELD CORROSION RESULTS
Procedures
The corrosion probes as removed from the incinerators are covered with scale and
deposits. More of these deposits were found at Miami County than at Norfolk. The de-
posit built up into a V-shaped layer of varying depth lengthwise along the probe, with
the apex or point of the V projected into the oncoming gas stream. The appearance of
Probe 9 after removal from the incinerator at Miami County is shown in Figure 6. The
high-temperature end of the probe is to the right of the photograph and is being raised
by the man shown in the picture.
It can be seen that the deposit in this area is about 4 inches thick. Another even
thicker deposit is evident on the water-cooled section of the probe as shown to the left
in the photograph.
These deposits were removed from all probes in a numbered sequence. Thus, the
variations in composition as a function of temperature could be determined when the
deposits were analyzed.
Frequently, and particularly on the carbon steel, an "egg-shell" scale was formed
directly adjacent to the base metal and below the deposit. This scale, which was mag-
netic, was sometimes 10 to 20 mils thick and was discolored blue-black. The appearance
of this scale and deposit can be seen in Figure 7. Note that the deposit builds up in con-
toured striated layers. The deposit, which was about 3/4 inch thick in this case, came
from Probe 4 near Specimens 16 and 17. The corrosion specimens shown in this photo-
graph have been descaled chemically.
Individual specimens were separated and examined. They were tested first for the
presence of sulfide by placing a drop of sodium azide solution (33) On the surface and
viewing the reagent under a binocular microscope. Evolution of nitrogen, gas bubbles
gave a positive identification of sulfide. Another area was then checked for pH by plac-
ing a moistened strip of Universal pH paper over the surface. Thus, the hydrolytic
properties of salts present on the surfaces could be determined.
Selected specimens were retained intact for special examination by X-ray, electron-
microprobe, and metallographic procedures.
The other pieces were then descaled by methods which prevented attack of the base
metal. Details of the stripping procedures are given in the Appendix. The amount of
metal wastage was determined by weight-loss measurements.
On the basis of OD and ID micrometer measurements, the overall weight loss was
adjusted to reflect the proportion of loss which occurred on the outside surface. This
was of significance only in the cases where appreciable oxidation had occurred on the
inner surfaces. The wastage rates for the specimens were calculated and expressed as
-------
CO
O
ft ft
W n)
Q *<
§ °
Z o
o
O
>< o
2 °
7 -H
t-\ -M
I-H O
U 4)
U
I
W
PQ
sO
W
O
ii
PH
-------
CO
*
w
PQ
O
O
r^
W
M 2
s:
o
-------
24
"mils per month" penetration. It must be noted, however, that such a projection as-
sumes a linear corrosion rate. As will be pointed out later, corrosion in incinerators
probably follows a parabolic-shaped curve and the rates decrease as the exposure time
increases.
Area 1 at Miami County
As was indicated in Table 2, ten corrosion probes have been run in the hot-gas
section of the incinerator at Miami County. The metal-wastage rates, expressed in mils
per month for comparison purposes, are summarized for the individual experiments in
Figures 8 and 9. The figures also show the temperatures reached by each specimen and
the total time that the specimen was at temperature. Details of the individual weight
losses, along with the penetration rates, are given in the Appendix in Table A-l. The
graphs in each figure have been arranged with those representing the longest exposure
period located at the top of the page and the shortest at the bottom.
The data are fairly consistent considering that there are wide variations in oper-
ating conditions and "fuel". Some overheating was experienced in several cases, par-
ticularly in those shown in Figure 9, because of blower outages and because of power
failures. However, meaningful comparisons between materials were obtained in all
cases.
Several trends of importance can be seen in the ten bar graphs. First, for the
A106 and Til carbon steels,
The rate of corrosion increases as the temperature increases and can
range from 5 to 15 mils per month at 325 F to about 20 to 50 mils per
month at 950 F.
The two steels are comparable in corrosion resistance with the A106
being slightly superior to the Til grade.
Second, for the stainless steels,
The rate of corrosion increases for Types 304 and 321 as the tem-
perature increases and ranges from about 10 mils per month at
625 F to about 40 mils per month at 1200 F.
Type 304 steel is somewhat more resistant then Type 321.
Types 446 and 310 are somewhat superior to Types 304, 321,
and 316.
Type 416 is less resistant than the other grades of stainless
steels evaluated.
Third, for the nickel-base alloys,
Incoloy 825 shows a resistance equivalent to that for Type 304
over the temperature range studied.
-------
Probe 9
Spec
- 8
- w
n n n n n n_
9 10 M 12 13 14 15 16 17 16
Average temp, F, 828 hr 290 320 355 390 425 470 500 530 570 600 640 680 725 760 80 660 900 940 1000 1010 1020 1030 1040 1060 1070 1080 1090 1100 1115 1130 1110 1150 1160 1175
Pro b " 12
O m
s e
234567
28 29 30 31 32 33 34
Averoge remp, F, 492 hr 305 350 390 435 480 525 570 615 660 700 735 470 8OO 835 870 9OO 935 970 OOO 1015 1025 I04O 10551070 JO80 1095 1110 1120 1135 1150 1160 1175 1180 1205
Probe 3
Probe 6
Average te-np, F 218 hr 310 360 400 450 490 540 580 625 670 710 750 780 625 860 900 940 970 IOIO 1050 10601080 1090 1110 II3O 1140 1160 1180 1190 1210 1230 1240 1260 12701280
Probe
FIGURE 8. TUBE WASTAGE RATES FOR PROBES EXPOSED AT MIAMI COUNTY
-------
Probe
_D_D_
23 2* 27 28 29 30 SI 32 33 M
) 390 440 480 530 5TO 610 660 TOO 720 5
Probe 2
Mo» temp ,F,20 fir 500 !
Average temp F,369hr 3CO ;
Probe I
-
-
-8 >- 3 =
< n " ^
nDnn
s
$
n
j
i
S
>L
i
1
!
8
II
s
8
1
i
s
n
*
s
§
s
5
1
J
* S
5
1
Probe 15
Specim
E»posure-ti
Probe 13
2 40
i
1
1 50
S
E
8-
f 40
i
5O
10
0
J!
e
~
i
I
t
n
II
D
1
J
2
1
3
i
,
1
1
1
5
0 n
II II II
i
5
J
i
i
2
1
3
J
'
f
2
1
3
J
'
,
1
U,
J
1
|
-
I
j a
5 2
1
|
S
fl
II
:
i
J
3
s
'
!
|
.
i
«
1
1
i
1
|
i
°
4 Z
E
\
t
5
E
(-
5
I
6 'I
7 2
i
C
- ;
'
f
n
II
0 31 J
1
I
1
,
*
_
-
s -
n n
s
° .
23 24 23 26 Z7 2B 29 3O 31 JZ S3 54
Eiposur< timt 206 hr
FIGURE 9. TUBE WASTAGE RATES FOR PROBES EXPOSED AT MIAMI COUNTY
-------
27
Inconel 600 and 601 have poor resistance in the high-temperature
regions, i. e. , above about 800 F.
Fourth, for protective coatings,
Metallic coatings of aluminum and chromium furnish short-time
protection but are not durable during long exposures.
Inorganic ceramic type coatings do not show long-time protective
action.
Fifth, heating during downtime showed no significant effect (see Probes 2 and 4). As will
be discussed later, however, the downtime conditions may lead to stress-corrosion
cracking of the stainless steels.
Comparison of the results from six probes when arranged according to the time of
exposure shows that the initial corrosion rates are high and that they taper off with time.
Such a comparison for the A106 carbon steels is illustrated in Figure 10, where the spec-
imen numbers are plotted as a function of wastage and exposure time. The approximate
temperatures are indicated by a lower band near 400 F extending to an upper band of
850 F. It can be seen that the wastage rates range from about 3 to 12 mils per month at
828 hours to about 30 to 55 mils per month at 120 hours.
The results suggest that the scale formed during corrosion is protective to some
extent and that the rate of attack will decrease appreciably as the exposure time is in-
creased. German experience as reported by Eberhardt and Mayer(l6) agrees with this
observation. Hilsheimer(2) also points out that cleaning the tubes increases the rate
of corrosion. Some cleaning, most likely with soot blowers, will be required to main-
tain good heat transfer, but the optimum frequency of cleaning is not known at present.
The present data are not precise enough to permit an exact extrapolation of wast-
age rates to exposures of many months or years, but some idea of the magnitude can be
obtained from Figure 10.
Metallographic studies of sections from specimens of tne A106 and Til carbon-
steel alloys indicated that the attack was uniform. Stainless steel specimens on the other
hand showed some structural changes and varying degrees of intergranular attack.
General observations on all specimens from all probes indicate that somewhat
greater wastage occurs in the area away from the point of maximum thickness of the
deposit. If the maximum deposit and wall thickness are considered to be at the 12 o'clock
position, the minimum wall thickness often occurs around the 6 o'clock position.
Area 2 at Miami County
Two corrosion probes were exposed in the flame area in the incinerator at Miami
County. Because of the high temperature here, the hot end of the probe reached about
1500 F even after the probe had been shortened to include only 19 specimens as com-
pared with the 34 normally used. As can be seen in Figure 11 for Probes 5 and 7, re-
spectively, the lowest specimen temperature reached was about 460 F. This figure also
shows the wastage rates experienced for the two probes. Detailed results are given in
-------
I
2
w
Pi
w
O
O
h-l
H
U
W
O
O
ii
H
O
pi
fH
O
U
fi
O
en
O
TJ
0)
CO
O
(U
0)
^
(U
TO
-------
Probe 5
100
90
80
70
-C
1 60
OJ
Q.
« 50
E
to
o>
5 40
o
30
20
10
n
_
_
-
-
_ h
O
1
w
o -
107 ^ 1.8
-
h
ID
O
h-
tD
O
<
-
<
O
fO
CM
ro
,
10
5
o
ro
i
^
C
r<
3
") C
1 "*
J
">
O CM
ro ro
"
CM
ro
3-
fr O
O ro
ro p
Specimen 23456
Average temp, F, 190 hr 475 520 575 625 670 720
\\J\J
90
80
70
£Z
| 60
!_
° 40
o
5
30
20
10
O
i
_ 12
CD
Q
-
i_
tX>
O
ro
g
7 8 9 10 II 12 13 14 15 16 17 18 19
770 825 875 925 990 1050 1115 1180 1280 1310 370 1430 1500
to
Q
1- <
1167 '
CD
I
9
f
0
ID Q
PO ^- "~>
o
ro
1 °
ro
(D
vj-
(D
ro
"
Q
ro
ro
ID
sj-
Specimen " I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19
Average temp, F.IOI hr 460 500 550 600 640 690 740 790 830 880 940 1010 1070 1140 1200 1260 1330 1390 1450
FIGURE 11. TUBE WASTAGE RATES FROM EXPOSURE IN FLAME AREA
-------
30
Table A-1 in the Appendix. Comparison with the results presented earlier for Area 1
shows that the corrosion for identical specimen temperature is much greater for the two
probes exposed in the flame area. The higher temperature of the flame would increase
the deposit temperature and accelerate reactions just above the cooler metal surface.
As can be seen in the upper bar graph in Figure 11, the wastage rates for the
carbon steels increased rapidly as the metal temperatures increased.
The stainless steels were severely attacked at temperatures above about 900 F.
As can be seen in the lower bar graph, the Type 446 material was somewhat superior
to the others examined. It may be significant that the attack on the stainless steels ap-
pears to go through a maximum over the temperature range 1000 to 1200 F, which
corresponds to the range where enhanced sensitization of austenitic alloys occurs.
Norfolk
In general, the three corrosion probes (Runs 8, 10, and 11) exposed at Norfolk suf-
fered much less corrosion than those exposed at Miami County. The most noticeable
exception was the low-temperature end of Probe 8, which corroded much like those of
Miami County. Details of wastage rates for the Norfolk probes are presented in Fig-
ure 12.
It will be noted that the Norfolk probes did not show the marked rise in corrosion
rate with an increase in specimen temperature as was found at Miami County. It is
believed that the lower ambient gas temperatures at Norfolk are related to this effect.
Norfolk Probe 11 was in the furnace 2304 hours and during this time was at tem-
perature 1318 hours. A comparison of corrosion rates for this run with the other two
at Norfolk shows that the corrosion rate decreased as the exposure time increased. This
decrease is believed to be associated with the protective action of the scale found on the
pipe wall.
Very little difference can be seen between results for A106 as compared with the
Til carbon steel on these probes. Of the stainless steels, Types 446 and 310 seemed
slightly superior to Types 304, 321, and 316.
The corrosion results for Probe 11 show that both aluminum and chromium coatings
furnished protection. The appearance of the specimens, however, indicated that this
protection would be only of relatively short duration since the coating had been corroded
away on about half the area on many pieces.
-------
9I£ CZ
DIE C
9t>t>
bO£ 1
IZE d
9IE I
DIE [
9bt>
t?0? d
i3£ c:
9I£ 1
DIE C
9t/fr C
K>£ C
O-90IV C
90iv i ;
IV- 901V C
*OE C
901V CZ
in czz
901 V CZ
iv-gav [
9I£ CI
Ol£ C
*O£ C
O-9OI V [
9bb [
IV-90IV [
90i v n ""
IV-9OI V [
II 1 '
901 V C
111!
? S 8 8 9 c
* i
8 S
oj Q
CJ1
fO gJ
s 1
a §
OJ CT>
is §
N 00
S S
00
^* 00
a s
w 8i
o
00
s §
* g
» S
* g
n
^ |
10 |
_ Q
* S
- S
fr C
lllCZ
901 V d W
1 1 .1- 1
901 V 1
II 1 d W
901 V 1
1 1 1
S §
rO °
E S
OJ «">
ro co
_ O
03
S §
"' 03
a §
s «
& 8
CM r^-
u> S
00 (^
S 8
s E
S g
a s
PJ [£
S g
a, S
_
* §
to o
ID
» §
~ ID
~ S
~ U>
to
9 §
» i
- §
* g
o
in **
t
» 8
^~
o
(VI ^J"
rO
O
cn
£VJ
qlc,
Ol£{
9frt> C
V0£ 1
laci
9I£C==
Ol£(
9frt- C
fro£cr
i^gi
9IE '
OI£C
gfrtc
t>0£ 1
ISEC:
901 V 1
niC=
fro£ czz
IZ£ CZI
901 VI
111 1
9I£ |
DIE C
frOE 1
IZE 1
9bb '
II | |
901 tf 1
II 1 1
II 1 1
1 1 1 1
S §
S 1
W 0
10 01
JO ^
o g
ro at
01 (£
s 8 a
& 1 0
in h
UJ w j
~ m oi
MH
« « O
~ " Z
» ^
s § H
S 1 <
o Q
~ S8 w
51 § 0
8§ S
- s W
» g W
pq
- g O
j
o Dj
u> S tr*
~ " PH
in
m m t-7-
^ f^
* S 0
~~ r^
= o S
H
CJ QO .
- » C/3
<
^ ^
10 W
CQ
>-i
P
H
O .
* CM
0 i-H
cvj r-
- s S
D
rT
ooooo oooooo v_^
Jtfl-Kl
-------
32
CHEMICAL SPECIES RESPONSIBLE FOR CORROSION
Furnace-Gas Compositions
Understanding of the mechanism by which deposits form and corrosion occurs on
the metal surfaces of the incinerator, among other things, requires a knowledge of the
furnace-gas composition. Sampling of these gases has been done at the Miami County,
Ohio, the Oceanside, New York, and the Norfolk, Virginia, sites.
Methods of Measurement
Ordinary combustion-gas measurements to determine carbon dioxide, oxygen, and
carbon monoxide were made by standard Orsat methods, Fyrite analyzers, and the
Bailey Heat Prover. The water content was measured by condensing the moisture and
determining its volume.
Sulfur oxides were sampled by bubbling flue gas through 1 percent hydrogen per-
oxide solution and analyzing the resulting sulfuric acid by the barium chloranilate
method.
Chloride analysis was made on this same solution by decomposing the H;p>O2 with
MnO£ and titrating the chloride with silver nitrate.
Fluoride analysis was made with an ion electrode (Orion). Gaseous HF was con-
verted to SiF^ by reaction with glass and absorbed in a buffered test solution. The
chloride content was verified by using a chloride-ion electrode on this same solution.
Total acidity in the flue gases also was monitored using a modified SO^ analyzer
(Instrument Development Company), which is based on the measurement of the electri-
cal conductivity of an absorbing solution of hydrogen peroxide.
Nitrogen oxides were sampled in 3-liter evacuated flasks. Analyses were made
by the phenol disulfonic acid method recommended by the Los Angeles County Air Pol-
lution Control District.
Aldehydes also were sampled in 3-liter evacuated flasks and were analyzed by the
sodium bisulfite method as recommended by the Los Angeles County Air Pollution Con-
trol District.
A check was made for unburned hydrocarbon gases by collecting a 3 -liter sample
for mass-spectrometer analysis.
In one run an analysis was made for SOo by condensing it from the gas stream and
determining sulfate. Normal boiler practice in coal-fired installation results in a flue
gas in which the SO^ levels are about 1 percent of the SO2- In this case, the SO^ con-
centration was 0. 3 percent of the SO£ found.
-------
33
Results and Discussion
As would be expected from the variable nature of the solid waste burned in munic-
ipal incinerators, a -wide range of values for the different gases was found. The data
are summarized in Table 3.
TABLE 3. FURNACE-GAS COMPOSITIONS
Concentration Ranges
SO2, ppm
HC1, ppm
HF, ppm
Nitrogen Oxides, ppm
Organic Acids, ppm
Oxygen, vol %
CO 2, vol °7o
CO, vol %
Miami County
2-303
5-115
Traces -0.6
10-138
35-136
15.0-9.5
6.0-10.8
Traces-0.2
Oceanside
0 -100
1-330
Traces- 1. 1
107(a)
Not measured
17.0-8.0
3.5-12
0.005^
Norfolk
0-10
14-75
Traces-6
4-7
150-340
16.5-15.
3.9-5.6
0.1-0.2
.0
0
(a) Single measurement.
Two gases of primary concern from the corrosion standpoint are SOz and HC1.
The amounts of these gases found in each of the three incinerators varied greatly.
The maximum values are not much over 100 ppm, except for one measurement for HC1
of 330 ppm at the Oceanside incinerator and 300 ppm SO^ at Miami County. In general,
the SC>2 and HC1 concentrations at Norfolk were less than those at the other sites. This
could have some relationship to the lower corrosion experienced at Norfolk, although
factors such as temperature are probably of even greater importance.
It can be noted in Table 3 that the SO£ concentrations are relatively low compared
with those for large coal- and oil-fired boilers, which usually range from 2500 to 3000
ppm. On the other hand, such boilers do not contain the HC1 found in incinerators.
Even at low SO2 levels, as the next section of this report shows, accumulation of sul-
fates in the incinerator deposits reach the same levels (ZO to 30 weight percent) in a
few hundred hours as are commonly found in power-station boilers after thousands of
hours of service. Although the HC1 levels in the incinerator gases are generally of the
same magnitude as the SO£ levels, the amounts of chlorides found in the deposits are
smaller than the amounts of sulfates, presumably because of the conversion of chlorides
to sulfates and loss of the chlorine.
803 was not believed to be a significant factor. One such measurement at Miami
County showed 1 ppm 803 and 303 ppm SO£ in the flue gases at the position where the
probe was exposed.
It is concluded that SO^ and HC1 were present in sufficient concentrations in all
three incinerators to be of concern from the standpoint of possibilities of corrosive
conditions .
The amount of HF found at Norfolk (6 ppm) is appreciably greater than the
amounts detected at the other two sites. No indications have been found that this gas
is of great significance in the tube-metal-wastage process.
-------
34
Concentrations of nitrogen oxides were in general lower (4 to 7 ppm) at Norfolk
than at the other sites, where they ranged up to 100 ppm. On the other hand, greater
amounts of organic acids (up to 340 ppm) were measured at Norfolk than at Miami
County (136 ppm). The high values for organic acids most likely stem from the large
amounts of wood burned at Norfolk. While the nitrogen oxides and organic acids con-
tribute to the acidity of the furnace gases, they probably play larger roles in scrubber
corrosion, where they are absorbed in wash water, than in fireside corrosion. Lab-
oratory experiments made during this study have shown that formic acid vapors have
negligible corrosion effects on the steels used.
The combustion gases in the three incinerators contained typical levels of the
normal combustion gases CC>2, QZ> and CO. Less excess air is used at Norfolk, as
temperature reduction is provided more effectively by the water walls and the boiler
convection surfaces.
Gas velocities were measured at two areas in the Miami County incinerator. The
flow was about 2100 fpm at Location 1 and about 450 fpm near Location 2.
Relation of Solid Waste to Gas Composition
In order to relate the nature of the solid waste burned to the composition of the
furnace gases, observations were made during six consecutive operating days at the
Miami County incinerator in April, 1970. A checklist of typical classes of materials in
the solid waste -was drawn up, and the presence of a relatively large amount (5 percent
or more) of any item was noted. At the same time, the gases in the incinerator were
analyzed for SO^, HC1, HF, nitrogen oxides, organic acids, and particulates. A sum-
mary of the data, with duplicate analyses in some cases, is presented in Figure 13.
The upper half of the figure shows the concentrations of the gaseous components of the
furnace gases, while the lower half shows the amounts of various materials in the hop-
per feed to the incinerator during the sampling period. Any materials present in
amounts of 5 percent or more are included in the figure. Lesser amounts of material
could not be estimated, so the totals for each day are slightly less than 100 percent.
Although it is difficult to draw any firm conclusions from the nature of the solid
waste and the furnace gases, some observations can be made:
(1) Organic acids were the predominant component on four of the days; on
these days, corrugated boxes, paper products, or wood were present
in large amounts.
(2) The amount of HC1 found was significantly below the 100 ppm level
which was observed on several other occasions. The presence of
5 to 10 percent of plastics in the waste did not cause any large in-
crease in HC1 concentration. The type of plastic being burned was
not identifiable, however.
(3) The very small amount of HF detected on these days agrees with that
found on other occasions, and probably is typical.
(4) The large amount of nitrogen oxides observed on the last day may be
related to the fact that this was the only day on which garbage was
present in a notable amount.
-------
E
CL
Q.
c"
o
o
L.
"c
IB
o
c
o
o
IUU
90
80
70
60
50
40
30
20
10
n
u t 136
n
-
Sulfur oxides
£ Organic acids
A o Nitrogen oxides
"A 4 x HC1
o HF
A
a
a
-
X A
X
X
X
a
D X
X X
- x a x ,
x
0 0 0 * ° *
ivu
90-
80-
70-
1 60-
o
ri
r
/
^
/
/
/
A
2
PI
\
s
/
{
V
/>
/
X
p
9
V///A Corrugated boxes
1 i Wood
KXXXX Plastic
K\\\N Aluminum foil
K/VAI Paper
Illlllllll Metal cans
Garbage
i i Textiles, tree limbs,
grass,leaves, bottles,
metal objects(notcans),
rubber
m
\
H ^ -iw/]
- : / i^ta
; [j|; ^J
ril April
30
FIGURE 13. COMPARISON OF FURNACE-GAS COMPOSITION AND SOLID WASTE AT
MIAMI COUNTY INCINERATOR
Materials in quantities of less than 5 percent are not included on the bar
graphs, so the totals for each day are slightly less than 100 percent.
-------
36
The amounts of particulates found in the furnace gases are shown in Table 4. They
range from 0. 36 to 2.40 lb/1000 Ib of gas.
TABLE 4. PARTICULATE CONTENT OF FURNACE GAS AT MIAMI
COUNTY INCINERATOR
Date
April 21
April 22
April 24
April 28
April 29
April 30
Paniculate
Grains/SCF
0.25
0.23
0. 35
1.27
0.23
0.27
0.49
0.19
Loading
Lb/1000 Lb Gas
0.48
0.44
0.66
2.40
0.44
0.51
0.93
0.36
C02,
percent
3 to 4
--
3 to 7
3 to 15
2 to 4
4
4 to 6
On April 24 a fairly large quantity of particulate matter was present, i. e. ,
2. 40 lb/1000 Ib gas; on that day, the chief item being burned was aluminum foil from an
industrial source. This resulted in large amounts of aluminum oxide dust. In fact, on
some occasions small particles of unoxidized aluminum have been found in the probe
deposits at the Miami County incinerator.
It appears that in order to obtain a reliable correlation between the solid waste
being burned and the nature of the furnace gases, it would be necessary to add known
amounts of various materials and determine the changes in the furnace-gas components.
Analysis of Deposits
In order to establish the causes and mechanisms of incinerator boiler-tube corro-
sion, analysis of the tube deposits is important. This analysis provides information
regarding the distribution of corrosive components in different temperature zones and
also provides identification of specific compounds that are present. Emission spec-
trography and wet-chemical techniques have been used to determine the concentration
levels of the various elements in the deposits. Specific compounds present have been
identified by X-ray diffraction.
Miami County Incinerator Deposits
Deposits from the corrosion probes exposed at the Miami County, Ohio, incin-
erator have given insight into the deposition process and the chemical reactions
involved.
The distributions of chemical elements found in significant quantities on the cor-
rosion probe as a function of time are shown in Figure 14. These exposures cover
periods of about 100 to 828 hours and constitute analyses of deposits from seven probe
exposures, all made over the bridgewall of the incinerator. Four temperature zones
along the length of the probe have been selected for consideration. The changes in sul-
fur and chlorine concentration have been emphasized by connecting the points in the
-------
-£-
«
o.
ro
,o
m
OJ
o
R
o
o
tO
m
8
8
ro
o
o
ir>
i
o
IT)
(M
"3
0
O
NJ
3
-*->
rt)
a
0)
H
W
H
W
PH
X
W
h
O
^
O
t I
H
O P4
Z O
P H
h <;
-------
o
o
in
r-
o
o
m
"8
c
o
N
d
"S
4)
ft
a
Q
W
P
O
O
w
rt
o
-------
0>
e
o
o
o
ri
I
O
IT)
o
V
o
N
-------
o
LO
I
o
o
o
(D
Pi
o
N
«
f-i
-u
ni
f-i
o
0)
H
Q
H
O
O
W
O
-------
41
figure, as these elements are of great significance in the corrosion process. The sul-
fur concentration goes through a maximum at about 400 hours in three temperature
zones, and the chloride does the same in the lowest temperature zone. However, these
peaks represent data obtained from a single probe and may reflect some unusual nature
of the waste being burned during the period in question. Additional data would be re-
quired to establish the sulfur maximum as a fact. Nevertheless, it is important from
the aspect of corrosion that the sulfur level builds up rapidly in the deposits and after
100 hours is almost as high as it ever becomes. At the higher temperatures the sulfur
level is somewhat reduced, probably because of the volatility of some of its compounds,
but it is still present in sufficient quantity to cause corrosion.
The same is true of the chlorine content of the deposits. At low temperatures the
chlorine persists in the deposits because of lowered volatility of chlorides and slower
rates of conversion to sulfates . However, even the small quantities present in the high-
temperature zones can cause serious corrosion.
High percentages of sodium and potassium were found in the low-temperature zone
of the shortest probe run, with decreasing amounts in the longer exposure periods.
Some differences may be attributed to the variations in the -waste being burned at any
given time, but these results also indicate that the sodium and potassium compounds
deposited initially have been converted to other forms during longer exposure times.
It is likely that the sodium and potassium deposited initially as oxides or possibly chlo-
rides were converted to sulfate by reaction with SC>2 and oxygen during the exposure
period. Sulfates are the dominant form of alkalies in tube deposits in coal-fired boiler
furnaces. This conversion to compounds of higher molecular weight results in lower
percentages of sodium and potassium in the deposits. X-ray-diffraction data have
shown that NaCl, KC1, Na2SO^., and (Na,K)SO^. all exist in the deposits, but the amount
of chloride is always less than that of sulfate even though there is as much HC1 as SC>2
in the furnace gases; this indicates conversion of chloride to sulfate.
The conversion of sodium and potassium oxides to chlorides by HC1 in the furnace
gases and ultimately to sulfates by reaction with SC>2 would result in the deposition of
potentially corrosive chloride-sulfate mixtures on the metal surfaces.
Both the lead and the chloride concentrations were highest in the low-temperature
zone of these probes. This result may reflect higher volatility of the compounds con-
taining these elements at high temperatures, as well as more rapid conversion of ox-
ides and chlorides to sulfates under such conditions. The fact that PbO, 4PbO- PbSC>4,
and PbSC>4 nave been identified in the deposits by X-ray diffraction shows that there is
stepwise conversion of the lead oxide to lead sulfate via the intermediate oxysulfates.
The unusually large amount of lead found in the 400-hour deposit probably resulted
from the burning of some waste with a high lead content during the exposure period.
As the exposure time was lengthened, aluminum became the predominant element
in the deposits, and the amounts of sodium, potassium, calcium, and silicon were sig-
nificant. This indicates that clay-like materials such as silicates containing sodium,
potassium, calcium, and aluminum were formed on long-term exposure. Materials of
this type, forming a hard deposit, can build up above the metal surface a protective
layer that is chemically inactive and serves as a barrier to diffusion of sulfur oxides,
oxygen, and HC1. The decrease in the corrosion rates with time are indicative of such
protective action.
-------
42
The long-term exposure of Probe 9 (8Z8 hours) resulted in a buildup of deposits
up to 3 inches in depth along the probe. Visual examination of the deposits indicated
that there were differences between the bulk of the deposit and the scale that formed
near the specimen surfaces. Consequently, this deposit was separated into layers:
(1) the scale next to the metal and (2) the bulk deposit used for analysis. The differ-
ences observed are shown in Figure 15. There was a greater concentration of potas-
sium, lead, zinc, iron, chlorine, and sulfur in the scale than in the deposit at all
temperatures. On the other hand, aluminum and silicon were present in significantly
greater amounts in the bulk of the deposit at all temperatures. It will be noted that the
amount of lead and chlorine in both deposit and scale decreased as the temperature
increased. These data are in agreement with the results of the electron-microprobe
examination of the probes, which revealed that potassium, lead, sulfur, and chlorine
were present at the metal-deposit interface on corroded specimens.
The analyses of deposits on Probes 5 and 7, which were in a region of high gas
temperature near the flame zone of the incinerator, are shown in Figure 16. The most
notable features in these deposits are the high iron and silicon contents and the low
sulfur. Iron and silicon were present in substantially greater amounts than in corre-
sponding temperature zones of the probes which were not directly exposed to the flames,
while the sulfur was quite low. The analysis indicates a more highly oxidized condition
in these deposits than in those collected further from the flame. Although the metal-
specimen temperatures for three of the zones were the same as those in other probes,
the deposit was subjected to higher gas temperatures which could have volatilized some
of the compounds formed in the deposits.
A relatively short experiment was run at the Miami County incinerator in which a
water-cooled probe was withdrawn at intervals of 2, 8, 24, and 88 hours for removal
of deposits. The results of the deposit analyses are shown in Figure 17. However, no
analyses were made for sulfur in these deposits. Extremely high concentrations of
chlorine and of iron were found in these deposits, indicating that chloride corrosion
begins early in the exposure. As the chlorides are converted to sulfates and pyro-
sulfates later in the exposure, corrosion by these compounds adds to the mutual wast-
age. In the early stages of the deposit history, the lead concentration also is high,
probably as PbO, and the relative amount of lead decreases as this compound is con-
verted to PbSC>4. The potassium and zinc concentrations remained fairly constant
during the time in which these deposits accumulated, while the aluminum and silicon
increased slightly.
Norfolk Incinerator Deposits
Three probe exposures were carried out at the Norfolk incinerator. The distri-
bution of elements in the deposits as a function of temperature is shown in Figure 18.
It is significant that lead and sulfur predominated in the deposits. The average chlorine
content was higher than that found in the long-term exposures at the Miami County in-
cinerator. These facts show that there was more than enough potentially corrosive
material present in the Norfolk incinerator deposits to cause metal wastage, if other
conditions were favorable. The factors that minimize corrosion at Norfolk appear to
be the relatively low metal temperatures and the small temperature gradient through
the deposit.
------- An error occurred while trying to OCR this image.
-------
e
i\ \\\ \\ \\x\\ \\\ \\x\\\\\ \\\ \\\ \\
^\.\\\\\\\\\\\\\\\
I\"\\\\\\\\\\X\\\'
IXXXXXXXX
E
IX \X\\\\\\\\\\\ X
ES
l^s.xxxx
«*
ixxxx \x\\\\ \\x\\ xxx \\ \\ \\x\x\\\
i
IX \\X\\X\\\\\\ XXX
1
n) |V\ \\X\\\X
7
iv XX.XXX *X \\X\\X\\\'
ES^
|\\\\\\\\ \
F^X:
r"
rc\\ \X\\\\\\\\ XX XXXXXXXXXX'
1
\
CM -*FJ,\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ \\X\\\
fvj »a
L_ _
(^ XXX\.\\\\\\\ XX Vj
i^sks
Ix xxx \ xxx \
1
^^2^s\\\\"*
n[__
IXXXXXXXXXXXX'
1
KXXXXXXX'>>
r
Q Pv\\\\\\X\\\\\\\\\\\\\\\\\\\\\\\\\\\\
ff\'*~i
^v XXX XX XXX X
1
IXXXX \\X\\\X
1
IX XXXXX XX
* , A
IXXXXXXX \\
1
(/}
(O
,-f
1*"
M
a.
o
z
(O
-^
0)
Li-
<=
N
0
z
o
.^
CO
a>
"
N
Q.
o
O
O
Z
tn
O
.
^~
^
if
c
N
Xi
Q.
O
O
^-
9
z
U.
1
O
(O
CJ
1 1
o
If)
CM
O
O
o
|(
o
1
o
tn
Is-
li.
o
U"»
K^
1
O
o
tn
CO
H
i i
CO
O
IXXK t^
(1 _^
* z*
a> co ^
-§ fe! ^
" ^ CO
n rvi rTi
"- MH MH
§§
pLJ *~^
n to ^
" ' Q O
.*-* i-^- ro
«n Z g
O %
-§ EH ptj
ft & °
CQ O
2 ^
E"1 O
Q to
^ i
0
td
^3
O
HH
tn
cd
0)
s
ai
i i
M-(
fl
H
Tl
CO
o
0)
CO
*
CO
a»
,£3
O
f-t
a.
cu
0)
^
EH
CM
-------
-iff
o
co*:
£
c o
o
o
8
o
GO
8
o
IT)
O
ro
CM
o o
(
o"
O
ro
m
CVJ
8
m
m
cr
w
C4
P
C/3
O
A
X
w
w
O
ffi
O
MH 7
W H
P U
W §
O
h O
o u
O
O
u
r^
ii
H
D
O
-------
C\J
15
"c
i
* 10
cf
o
p
W
4
c
8
5
»Pb
i
>Pb
s J
_^**^
^****^
s^-^^^
i
i
i
i
i
1
<
C
^^
iZn.Si
'Ca
iK '
'Al
Na
jRj
,Zn,Si ;
ca
'K !
M '
Al
r _
1 "~-~- J^
a "*^^-
^^
q
.
'Ca.Zn
Pb
;K
Na
>Fe
a
250-500 500-750 750-950
Temperature Zones, F
FIGURE 18. PROBE-DEPOSIT COMPOSITION AS A FUNCTION
OF TEMPERATURE FOR EXPOSURES IN THE
NORFOLK INCINERATOR
-------
47
Some specific points of difference between the Norfolk deposits and the Miami
County deposits were noted:
(1) The lead concentration was significantly higher on Probe 10 at Norfolk,
particularly at temperatures above 500 F. However, this high lead
concentration did not result in greater corrosion.
(Z) Sulfur was the predominant element in all temperature zones, although
it was not present in unusually large amounts. No such predominance
of one element has been noted in the Miami County deposits.
(3) In general, deposits on the probes at Norfolk have been less voluminous
and more powdery than those at Miami County.
(4) A significant observation concerning Probe 10 at Norfolk was that less
sulfide was present on the specimens, as compared with probes at
Miami County or Probe 8 at Norfolk. In addition, fewer acidic areas
as revealed by moistened pH indicator paper were present on the mildly
corroded specimens from Norfolk as compared with those from Miami
County. This means that less reaction with sulfur-containing salts had
occurred and that fewer compounds such as pyrosulfates were present.
Sulfide and pH Tests
All specimens except those used for special electron-microprobe or X-ray stud-
ies were subjected to testing by sodium azide solutions for the presence of sulfide, and
to moist pH indicator paper for acidity or alkalinity. These tests were made on speci-
mens with the bulk deposit removed mechanically.
A summary of the results obtained on nine probes is provided in Table 5. The
presence of sulfide is indicated by a + sign. A double ++ denotes a very strong reaction,
i. e. , evolution of many N2 gas bubbles. The pH values were estimated from the color
obtained on Universal pH paper, and show gross effects only.
Considering first the results for the presence of sulfide:
(1) Sulfide was found on the majority of specimens.
(2) At Miami County, the lowest concentrations of sulfide were found pre-
dominantly on the high-temperature specimens of Probes 2 and 6 and
scattered over the length of Probes 5 and 7. This is consistent with
the fact that Probes 5 and 7 were operated at higher temperature.
(3) At Norfolk, very little sulfide was found on the low temperature speci-
men on Probe 10; several pieces at higher temperatures also gave a
negative reaction.
It should be mentioned that in some instances it was necessary to scratch through
the oxide scale and expose base metal before the sulfide test was positive. This is
taken to mean that the sulfide was formed very near or in the metal surface.
-------
48
TABLE 5. RESULTS OF SULFIDE AND PH TESTS
Specimen
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Probe
Sulfide
4
+
4
trace
44
4
44
+
4
4
4
4
44
44
44
4
4
44
44
4
4
trace
4
4
trace
4
trace
4
trace
trace
neg
neg
trace
trace
2
PH
3
3
4
4
4
2
2
2
2
7
3
3
3
4
5
4-8
4
4-8
4-8
7
3
3
3
Probe 3
Sulfide pH
4 2
4 2
44 7
44 2
44 3
44 2
44 3
4 2
44 3
4 3
4 3
4 4
44 2
44 3
44 3
44 7
4 4
4 4
4 7
4 4
4 7
4 4
4 4
4 4
44 7
Probe
Sulfide
4
4
44
4
++
44
4
44
4
4
44
44
44
44
44
44
4
4
4
44
4
4
+
4
44
44
4
4
44
44
4
444
44
4
4
PH
3
2
2
3
3
3
3
4
4
3
3
4
4
3
7
7
7
7
7
7
1
7
7
7
7
7
7
7
7
7
7
7
7
7
Prooe
Sulfide
neg
neg
trace
4
4
44
neg
trace
4
44
neg
4
trace
neg
neg
4
4
4
4
44
5
PH
4-8
4-8
9
4
4-8
10
10
10
4-8
4-8
4-8
4-8
4-8
4
4-8
4-8
4-8
4-8
4-8
Probe
Sulfide
4
44
44
4
4
4
4
4
44
4
4
4
44
4
4
4
4
4
4
44
4
44
44
44
4
4
4
neg
neg
neg
4
neg
4
4
6
PH
3-8
3-8
3
2
2
3
3
2
3
4
3
3
3
4
7
7
7
7
4
4
9
4
9
4-9
4-9
4
4
7
7
7
7
7
7
7-9
Probe
Sulfide
4
neg
4
4
4
4
neg
44
44
44
4
4
4
4
4
neg
4
neg
4
7
PH
3
9
4
7
9
3
3
3
7
10
7
9
8
7
7
8
7
7
7
Probe
Sulfide
4
44
44
4
4
44
4
44
4
44
44
44
44
44
44
4
4
44
44
4
4
4
4
4
44
44
neg
4
44
44
44
neg
neg
4
8
PH
9
3
3-8
4-8
4-8
4-8
4-8
4-8
4-8
3-8
3-8
3-8
3-8
4-8
4-8
4-8
3-5
4-8
4-8
4-8
5-8
3-8
9
3-8
4-8
4-8
4-9
4-9
9
9
5-9
5-9
5-9
5-9
Probe 9
Sulfide
44
44-
4
44
44
neg
44
4
44
44
44
44
44
44
44
44
44
4-4
44
44
44
44
44
44
44
44
44
44
pH
3
3-8
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
Probe
Sulfide
neg
neg
neg
44
4
4
44
44
4
44
44
44
4
44
44
44
4
4
44
44
4
4
44
neg
44
4
4
neg
4
4
neg
4
4
4
10
PH
3-8
3-8
3-8
3-8
3-8
4-8
9
4-9
4-9
9
4-9
4-9
4-8
4-9
5-9
5-9
4-9
9
9
4-9
5
4-9
4-8
4-8
4-9
7
7
8
9
9
9
3-8
6-9
5-9
-------
49
It can be seen in the table that an acidic spot was often detected near one giving an
alkaline reaction (see Probes 5, 8, and 10). There was no direct correlation between
acidity and sulfide but in many cases alkalinity was associated with the absence of sul-
fide. This would be consistent with the hypothesis that the sulfur-containing salts such
as K2S>2<^->7 an(^ KHSO^ produce the sulfide. Furthermore, the pH values were lower for
the longer time exposures i.e., Runs 9 and Z.
The alkaline areas are probably associated with deposits of K^O or Na;?O which
transform to their corresponding hydroxides in the presence of water.
Electron-Microprobe Analyses
The electron microprobe is an instrument which permits examination of a surface
to determine the location and relative concentration of individual elements. This tech-
nique was used to examine the area at the scale-base metal interface on sections cut
from corroded portions of boiler tubes removed from the Oceanside incinerator in 1968,
1969, and 1970. Sections from Miami County Corrosion Probes Z and 3 were also ex-
amined, as were sections of Probe 10 from Norfolk, Virginia. None of these speci-
mens were descaled prior to sectioning so that the areas could be viewed with some
scale intact, but with the bulk deposit removed.
An illustration of the types of data obtained is provided in Figure 19 for a 1968
Oceanside tube. The position and intensity of the white areas in the photographs indi-
cate the location and concentrations of the individual elements. This figure shows the
relative concentrations of zinc, chlorine, tin, lead, copper, iron, potassium, and
calcium in the scale and at the metal/scale interface. The base metal is located in the
lower portion of the photographs, and the interface is at the rough-appearing area
shown in the image photographs and indicated by the lines at the right side of the figure.
The chlorine, zinc, and copper are found directly adjacent to the metal surface. In
fact, some selective penetration of the zinc and chlorine into corroded areas of the
metal is indicated at the pitted-appearing center area. The sulfur in this instance as
in many others is concentrated in the deposit in a layer somewhat removed from the
metal-deposit interface.
The microprobe results for the 18 specimens examined are summarized in
Table 6.
The column designated 'inner layer' represents the interface between metal and
adherent scale, while that designated the 'outer layer' is in the scale, slightly removed
from the metal surface (approximately Z mils). They both are beneath what is nor-
mally considered as deposit. In summary, all three Oceanside boiler tubes showed
high concentrations of chlorine at the interface, and the 1968 and 1969 samples also had
high concentrations of zinc in that region. The large amount of iron represents metal
that diffused outward from the original tube surface in the form of corrosion products.
Other unusual features are the large amounts of potassium in the inner layer of the 1969
sample, and high copper for the 1968 sample. Sulfur was found to be high only in the
outer layer of the 1968 tube, as was lead. Potassium, calcium, and aluminum were
concentrated in the outer layers of the deposits.
-------
TABLE 6. RESULTS OF MICROPROBE ANALYSES
Relative Concentration of Elements
Element
Layer
Oceanside boiler
tube, 1968
Oceanside boilei
tube, 1969
Ocean side boiler
tube, 1970
Miami County Probe 2,
A106B, 633 F
Miami County Probe 2,
Til steel, 680 F
Miami County Probe 2,
304 stainless, 1130 F
Miami County Pfobe 2,
321 stainless, 1150 F
Miami County Probe 3,
A106B steel 300 F
Miami County Probe 3,
A106B steel, 450 F
Miami County Probe 3,
A106B steel, 820 F
Miami County Probe 3
Til steel, 500 F
Miami County Probe 3,
321 stainless, 1005 F
Norfolk, Va Probe 10
Til steel, 340 F
Norfolk, Va Probe 10
A 106 steel 440 F
Norfolk, Va Probe 10,
321 stainless, 540 F
Norfolk, Va Probe 10
A106 steel 660 F
Norfolk, Va Probe 10
A 106 steel ,7 50 F
Norfolk, Va Probe 10,
321 stainless, 805 F
Chlorine
Inner
H
H
H
L
H
L
L
M
L
H
H
M
M
M
L
L
Outer
M
H
L
M
L
L
L
L
M
L
M
H
M
M
L
M
Sulfur
Inner
M
L
M
M
L
L
M
H
H
H
L
VL
L
VL
M
H
Outer
H
M
VL
H
L
VL
L
1
L
VL
VL
H
L
M
VL
H
H
Lead
Inner
L
L
L
L
L
L
L
-
M
H
-
L
L
VL
VL
Outer
H
M
L
H
L
L
L
-
H
L
-
M
M
M
H
Zinc
Inner
H
H
L
VL
L
L
M
VH
H
H
-
M
VL
M
H
L
Outer
M
VH
L
L
L
M
M
L
VL
M
-
M
M
1*
M
H
H
Iron Copper Sodium
Inner
H
H
H
VH
VH
VH
VH
VH
VH
VH
H
VH
VH
M
VH
VH
VH
VH
Outer Inner Outer Inner Outer
H H L - -
M It L VL L
VL L L -
VL - - VL L
VL - - L L
VL
H
H H M L L
L - - - -
VL - - L L
VL - - L L
H - - L L
VH
L - - - -
VH - - -
VH VL VL
L - - - -
Potassium Calcium
Inner
M
H
-
L
L
M
L
M
H
VH
L
H
M
H
L
Outer Inner Outer
H VL H
VH VL M
-
H VL L
H L H
L L M
L VL M
M
L
VL
L
H L M
VH - L
M - -
H
H
Aluminum Silicon Tin Nickel Chromium
Inner Outer Inner Outer Inner Outer Inner Outer Innet Outer
L L - - - -
LM- - VLVL-- --
LH- -------
LLLMLL----
MLVLH----MVL
L M L H - -H VLHVL
LMMH--HMVHL
M VH --------
LH- -LL-- --
L VH ----- VH M
- - - - - - VH VH L VL
VLVLM M M M - - - -
VL M VL M L L H VL VH VL
VL VL VL VL L L - - - -
L M VL L L L VH VL VH VL
Note VH veiy high, H higti, M moderate, L - low, VL - very low Dash indicates element not found m image area
-------
51 and 52
FIGURE 19.
RELATIVE CONCENTRATIONS OF SELECTED ELEMENTS IN
METAL-DEPOSIT INTERFACE OF CORRODED BOILER TUBE
REMOVED FROM OCEANSIDE INCINERATOR IN 1968
Electron-Microprobe Pulse-Mode Photographs.
-------
-------
53
For the corrosion-probe specimens exposed in the Miami County incinerator,
chlorine was high at the inner layer for four of the nine specimens examined and was
present on the remaining ones. Similarly, sulfur was high on three pieces and was
present on the other six. Potassium, aluminum, zinc, and lead were also found in
large amounts on many of the pieces. Copper was found only on Probe 3, the 300-F
specimen.
In general, the chlorine was less concentrated on the specimens from the Norfolk
Probe 10. It -was present on all six pieces examined, but was high only in the outer
layer of the 340-F specimen. Sulfur was present in large amounts on the specimens
from 750 and 805-F positions, but was absent on the 440-F specimen. Zinc was pres-
ent on all six specimens and was high on three of them. Smaller amounts of lead were
found. Potassium was high on half of the specimens.
Potassium, chlorine, and sulfur were found deep within a grain boundary on the
540-F specimen, a Type 3Z1 steel.
The microprobe results show that compounds containing zinc, potassium, chlo-
rine, sulfur, and sometimes lead are usually present at the corroded tube surfaces.
Thus, these results coupled with X-ray studies and laboratory experiments with various
salts help impart an understanding of the cause of the corrosion. This point is dis-
cussed more fully later.
X-Ray Diffraction Studies
Since chemical assays have revealed that a large number of elements are retained
in the deposit extracted from air-cooled probes of refuse incineration, and since
electron-microprobe analyses have shown several of these elements to be concentrated
in the adhering scale, phase studies have been conducted to determine how these ele-
ments are combined. The phase studies have been conducted by obtaining X-ray-
diffraction data on the bulk deposit and on the individual layers revealed by microscopic
examination. Figure 20 presents typical analyses for various probe temperatures and
a photomicrograph illustrating the general location and appearance of the scale and de-
posit as seen under reflected light. The colors indicated in the tabulation were deter-
mined under oblique illuminations from an incandescent lamp. It is evident from these
results that distinct phase changes occur at various distances from the tube-metal sur-
faces as well as at different probe temperatures.
It can be seen first, that unlike the normal oxidation of iron which results in a
layer sequence Fe-FeO-Fe3O4-Fe2O3 above 1040 F, or Fe-Fe3C>4-Fe2O3 below 1040 F,
the scale in contact with the substrate metal is found to contain FeCl^-ZH^O throughout
the entire probe-temperature range 300 to 1ZOO F. The hydrate form of FeCl2 undoubt-
edly resulted from exposure to humid air after the probe was removed; in the inciner-
ator the form is probably FeC^. At the moderately high temperature end of the probe,
it was found that the FeCl2 layer melts and collects into small pools. As can also be
seen in Figure 20, a layer of FeS was also found on specimens from the high-
temperature end of the probe.
The electron microprobe showed that some zinc compound was also present in
this area near the metal. It appears that this element is not in a form readily detected
-------
XSil
MiruMai
O
HH
EH
U
Q
W
Q
i i
w
W
H
-------
55
by X-ray techniques, but may be incorporated in the FeCl2 as (Fe, Zn)Cl2 or in the
FeS as (Fe, Zn)S with the host structures. Lattice-parameter data were not obtained,
but shifts in line position were observed from specimen to specimen.
These interface layers are of the order of 6 to 10 microns thick on the low tem-
perature specimens. The thickness increases somewhat with temperature.
Between the interface scale and the bulk deposit is a multilayered scale of iron
oxides: Fe2O-:3 and Fe3C>4. It is interesting to note that a thin layer of alpha Fe^O^
forms at the FeCl2-mixed oxide interface. Part of the mixed-oxide scale adheres to
the substrate and part to the external deposit. The mixed-oxide (Fe2C>3 -Fe^C^) layer
is a. hard, brittle, magnetic, gray-black material which thickens with temperature and
is made up of several layers with loose red Fe2C>3 between layers. This suggests that
reduction or reaction with other elements occurs in or below this layer as is discussed
later. At the higher temperature where FeS forms, the mixed oxide scale is less ad-
herent, and in this area, more of the mixed oxide is removed with the outer deposit.
The topography of the corroded metal here changes from relatively smooth to dimpled.
Now as one examines the phases present in the deposit up and beyond the separa-
tion line shown in Figure ZO, it can be seen that a great number of compounds have been
identified by X-ray diffraction. These compounds have been arranged in classes and
are listed below along with the compounds found near the tube.
PbSO4 SiO2 NaCl CaCO3
ZnSO4 aFe2O3 KC1 Al
Na2SO4 Fe3°4 FeC12 KOH
(Na, K)2SO4 PbO Mg2SiO4
CaSO4 PbO- PbSO4 FeS
K2Pb(SO4)2 4PbO- PbSO4
KA1(S04)2
It should be mentioned that the examinations with the microscope indicated that
some melting had occurred near the scale. These studies showed the presence of re-
crystallized continuous phases (NaK)2SO4 and PbO PbSO4 on the external surface of the
mixed iron oxide scale. The (NaK)2SO4 phase is also dispersed throughout the iron
oxide scale that adheres to the bulk deposit. It appears that separation in the iron oxide
scale occurs at the depth to which this salt phase has permeated in sufficient quantity to
destroy the integrity of the iron oxide. It is proposed that in the initial stage of incin-
erator operation, these salt phases as well as mixed chloride salts possibly ZnCl2 and
PbCl2 permeate the iron oxide scale and destroy its protective characteristics.
As is discussed later, these salt phases when molten also contribute to maintain-
ing a low oxygen partial pressure in the adherent scale. While the melting points even
for mixtures are significantly higher than those indicated by thermocouples placed in
the metal probes, the furnace-gas temperature will be much higher than the tempera-
ture of the probe because of the thermal gradient in the deposit. This condition is sub-
stantiated by the fact that molten pools of FeCl2 were detected where no deposit had
formed on a high-temperature (1100 F) probe sample, but did not melt on the opposite
side of the same section where the deposit had formed. It is interesting to note that the
-------
56
molten salt at the deposit-mixed oxide interface changes color and phase with time and
temperature. In short times the deposit consists of NaCl, KC1 and an unidentified
phase containing large quantities of Zn or Pb while for longer exposures the melted de-
posit on low-temperature sections (~300 F) was found to be PbO- PbSO^., at intermediate
temperature Na2SC>4 plus KOH and at high temperatures (NaK)2SC>4 plus
-------
57
CORROSION FROM INCINERATOR DEPOSITS
UNDER HUMID CONDITIONS
Stress-Corrosion Cracking
Since all incinerators must be shut down on occasion, and since high humidity
conditions can often cause the deposit on the tube surfaces to become moist, it is
important to assess the severity of attack under these conditions. The most damaging
forms of attack to be anticipated are stress-corrosion cracking (SCC) and pitting.
Probably the most hazardous of these is stress-corrosion cracking.
When an alloy is simultaneously subjected to surface-tensile stresses and a
variety of specific corrosive environments, it may experience premature failure be-
cause of stress-corrosion cracking. This process is unique in that failure does not
occur when the individual factors mentioned above are acting alone.
One aspect of the phenomenon is that the metal at the failed areas does not ex-
hibit ductile tearing, but instead exhibits a brittle fracture. Crack propagation, once
initiated, is quite rapid. The propagation is in a plane perpendicular to the direction
of the applied stress. The cracks may be either transgranular or intergranular,
depending on the alloy and the composition of the environment.
The tensile stresses required for cracking may be either applied or residual.
The latter may result from welding or from press fits or other assembly procedures.
Also, the presence of crevices may favor the initiation of stress-corrosion cracks.
All common structural materials such as carbon steel, low-alloy steels, stain-
less steels, aluminum alloys, copper alloys, and nickel alloys are susceptible to SCC
to varying degrees in some specific environments. The mechanism of the cracking is
not completely understood. From an engineering standpoint, however, it is important
to know which combinations of materials and environments will initiate cracking.
Mild steels will crack in certain nitrate, caustic, and carbonate-bicarbonate
aqueous solutions and in some gaseous CO-CO2~H2O mixtures, anhydrous ammonia,
and H^S-containing environments. Obviously, these substances cannot all be avoided
in incinerators. It is considered unlikely, however, that the specific combinations
of these substances which promote SCC of mild steel will be encountered in incinera-
tors except on rare occasions. Aqueous nitrates and gaseous CO-CO2~H2O mixtures
would be the most likely causes of SCC in mild-steel incinerators.
While SCC of stainless steels can occur in caustic environments, the most im-
portant substance promoting the SCC of stainless steels is chloride ion. Trace quanti-
ties of this substance are enough to cause austenitic stainless steels to develop stress-
corrosion cracks rapidly. While higher temperatures generally favor SCC, Type 304
will crack even at room temperature. (34) Stainless steel alloys will crack when
stressed to as low as 2000 psi in the presence of a few parts per million of chloride
ions.
Since chloride ions are commonly present in incinerator deposits, two groups
of materials were evaluated for SCC while in contact with incinerator deposit under
humid conditions. The first group of alloys included Types 304, 310, and 446 stainless
-------
58
steels and carbon steels A106, Grade B, and A213, Grade T-ll. The Group 2 speci-
mens included Type 316L stainless steel and three other alloys more resistant to SCC
in chloride environments, namely, Inconel 600, Inconel 601, and Incoloy 825. Bent
specimens (C-rings) were stressed to a level halfway between the yield stress and the
ultimate tensile strength (at the SCC test temperature). Some U-bend configurations
were also used. Some specimens of Type 304 stainless steel were sensitized for
2-1/2 hours at 1200 F to increase their susceptibility to SCC. Other specimens were
tested as received.
The tensile-stressed surface of the specimens was exposed to incinerator deposit
in a moist-air atmosphere at about 170 F. Group 1 specimens were exposed to deposits
from Miami Probe 3, Samples 1 through 8, or Probe 4, Samples 1 through 7. Group 2
specimens were exposed to deposits from Miami Probe 12, Samples 1, 2, and 3.
The apparatus for these moist-air SCC experiments is similar to a double boiler:
the specimens are placed on dry pulverized incinerator deposit in the inner vessel
while distilled water in the outer vessel is heated to humidify the atmosphere.
In the Group 1 alloys, the Type 304 specimens (both sensitized and annealed) showed
cracks after an exposure of one week. Fracture occurred on the sensitized Type 304
piece during the eighth week of exposure. Cracks were first observed in Type 310
steel after 10 weeks. Deep-stress-corrosion cracks developed in all three materials
during 20 week's exposure. There were fewer cracks on the Type 310 specimens but
they -were very deep. The attack on this material was predominantly transgranular as
is illustrated in Figure 21.
The nature of the cracking on Type 304 stainless steel at the end of three weeks
is illustrated in Figure 22. That metallographic examination of cracked specimens of
Type 304 revealed a number of branching cracks in both specimens. The cracks in the
unsensitized specimen follow a predominantly transgranular mode of propagation; that
is, most cracks cut through grains. The sensitized specimen, however, revealed a
considerable degree of intergranular propagation of cracks.
On the other hand, no SCC has been observed with Type 446 steel after 22 weeks
of exposure. The specimen, however, was quite rusty and pitted. After 12 weeks,
the A106 and T-ll steels were quite rusty and scaled but no stress-corrosion cracking
was observed. The surface roughening on these steels was noticeable at 15X but was
not great. There was no indication of deep pitting as is sometimes seen in corrosive
environments.
The Group 2 alloys were exposed for 26 weeks. Only the Type 316L specimens
exhibited cracking and this was at 22 weeks. However, some rusty appearing spots
appeared on this specimen in about four weeks. As can be seen in Figure 23, the
cracking started at one of the rusty areas. A photomicrograph taken of a section cut
through the crack indicates that the attack is transgranular in nature.
At 26 weeks, the Inconel 601 specimen showed many fine, shallow pits. Some of
these appeared at four weeks as rusty spots.
The Inconel 600, at 26 weeks, was mottled in appearance as if it were etched in
spots. This alloy also showed a few rusty spots after four weeks.
-------
59 and 60
100X
C4007
FIGURE 21. STRESS CORROSION CRACK IN C-RING OF TYPE 310
STAINLESS STEEL EXPOSED TO INCINERATOR
DEPOSIT UNDER HUMID CONDITIONS
The cracking is largely transgranular.
-------
-------
61 and 62
7X
C-3937
7X
C-3938
a. Sensitized 2-1/2 Hours at 1200 F
b. Unsensitized
100X
C-3939
c. Photomicrograph of Sensitized
Specimen
Vii-
100X
C-3940
d. Photomicrograph of Unsensitized
Specimen
FIGURE 22. STRESS-CORROSION CRACKS RESULTING FROM EXPOSURE OF
STRESSED TYPE 304 STAINLESS STEEL TO MOIST INCINERATOR
DEPOSIT AT 170 F
-------
-------
63 and 64
r
\
/ :-U
»%
i
v * , »
I*' >
, /
W
H
* * V "»v. v. ^
' ° '^^''-'' ->^7v a
> * - *. - /^' - .**> "^ > *. / *i' / H
/*-,S; ''^''V *. */// w
*« a l_jr sr-^^X :if/l T -*-- ^ * w
** ^Jjf"*4 'i *V ****» » -«- w
'.-: x :4'-%%
^^M^j^fe^.-^^
W
Q
ss
< c^
r^ W
W Q
X ^ tf
o ^ o
W
o
o
^
U
u g u
I-H O
O ft
o ^
en W
(M
W
CiJ
P
O
Q
X
o
(M
T)
OJ
o
rt
IH
U
-------
-------
65
The Incoloy 825 showed good resistance throughout the 26-week period. It is,
therefore, concluded that this alloy is a good choice when resistance to downtime
conditions is considered. Since the steels A106, Grade B; and A213, Grade T-ll are
corroded in a fairly uniform manner under these conditions, they can also be con-
sidered for construction materials. The relatively good performance of the tubes in
the boiler at the Navy Public Works Salvage Fuel Boiler provides additional evidence
for consideration of carbon steels.
The austenitic-type stainless steels such as 304, 310, 316, etc. , do not appear
to be good choices because of the likelihood of the incidence of stress-corrosion
cracking.
The ferritic-type stainless steels as exemplified by 446, have the limitation of
possibly being subject to seve're pitting under the downtime conditions.
Apparently, these laboratory conditions were more severe than those experienced
in the field because the corrosion probe specimens exposed at Miami County did not
show bad pitting even when downtime occurred weekly.
Ferritic steels like Type 446 are thought to be immune to SCC in chloride media,
but they may be embrittled by hydrogen, particularly when sulfur is present. (36)
Stressed C-ring specimens of Types 304, 310, and 446 steel have also been
exposed to pulverized incinerator deposit at 1000 F for 50 and 250 hours in a simulated
incinerator atmosphere containing HC 1 and SC>2- As expected from previous work, (37)
none of the four stainless steel specimens developed any evidence of SCC under these
conditions. However, all four specimens were severely corroded on the surfaces that
contacted the deposit. A later section of this report gives the corrosion rates for
pieces exposed with the C-rings.
Other stressed C-ring specimens were exposed at 1000 F for 50 and 250 hours to
a salt mixture of 29. 6 g K2SC>4 + 0. 3 g NaCl in the same gaseous environment. This
environment was selected because it borders test results with high and low corrosion
rates and, accordingly, it would be more likely than most of the other environments
investigated to induce SCC. Again, none of the specimens developed stress-corrosion
cracks, although all were somewhat tarnished on surfaces that contacted the salt.
Additional stressed C-ring specimens were exposed to the simulated incinerator
atmosphere (as above) but in the absence of contact with either salt mixture or incin-
erator deposit. Neither cracking nor much corrosion attack resulted from these
exposures.
Metallographic examination of specimens exposed to hot salts and incinerator
deposits revealed some instances of surface-layer intergranular attack, but no
evidence of SCC.
The results of these investigations confirm that SCC is not likely to be a
problem with stainless steels at the operating temperatures of incinerators. However,
as was shown previously, the damp incinerator deposits can cause cracking.
-------
66
LABORATORY CORROSION STUDIES
The objective of the laboratory experiments was to determine the importance of
individual factors on corrosion in incinerators and also to help establish the mechanism
by which metal wastage occurs. Since the experiments could be carried out under
carefully controlled conditions, it was possible to determine the role of such factors
as gaseous components, salts, or deposits, metal composition, and temperatures. The
analyses made by the microprobe, X-ray, and analytical chemistry have served to
define corrosive salt mixtures to be studied.
The approach taken in the research described was to subject alloys to simulated
flue gases with and without direct contact with known salt mixtures. Weight losses
were used to assess the amount of corrosion which occurred.
Apparatus and Procedures
Metal specimens were placed in porcelain boats or crucibles located within quartz
or Vycor tubes passing through resistance-heated tube furnaces. Gases were preheated
in a section which was packed with quartz wool to enhance heat transfer.
Synthetic furnace gases (FG), consisting of typical mixtures of CC>2>
water as encountered in the incinerator, were used with SO2 and HC1 separately and
in combination. The standard composition used was: 80 percent air, 10 percent CO2,
and 10 percent H^O with 250 ppm SO2- The gas flow was about 0.16 cfm, which
corresponds to a linear velocity of 0. 2 fps. HC1 was added on occasion in amounts of
200 and 2000 ppm.
The equipment upstream of the reactor consisted of flowmeters for metering
carbon dioxide and air to the reactor, manometers to regulate the pressure of these
two gases at 19. 7 psia, and a manometer for determining pressure drop through the
system and exit portion of the setup. The controlled and small amount of SOz was
brought into the stream of carbon dioxide through capillary tubes. The flow of sulfur
dioxide was determined and checked periodically by the hydrogen peroxide method. By
sparging air through heated water, water vapor was introduced into the system and was
controlled by the flow of air and the temperature of the water reservoir.
Corrosion specimens about 0. 75 x 0. 175 x 0. 125 inch were made from A106
Grade B, A213 Grade Til, and Type 321 stainless steel.
Exposure times were usually 50 hours and the extent of corrosion was evaluated
by weight-change measurements and by metallographic examinations. The sodium
azide test described earlier was also used to detect the presence of sulfide on the cor-
roded steel specimens.
-------
67
Results
Gas-Phase Reactions
As might be expected, the attack of metals by gas-phase reactions was not great
at normal operating temperatures - i. e. , 800 F. Details of the extent of the attack
from furnace gases containing 200 ppm HC1 on A106 and Til steel at 800, 1000, and
1200 F are summarized in Table 7. It can be seen that the corrosion rates increased
rapidly as the temperatures increased in a manner similar to that noted on many in-
cinerator corrosion probes although the estimated wastage of these laboratory
specimens was much less than for the probe specimen. The presence of HC 1 had little
effect on these corrosion rates.
TABLE 7. GAS-PHASE CORROSION AT
ELEVATED TEMPERATURES
Corrosion Rate,
Alloy Temperature mils per month
A106
A106
A106
Til
Til
Til
800
1000
1200
800
1000
1200
0.
8
36
0.
6
29
9
8
Sulfate-Chloride Mixtures at 1000 F
As was mentioned earlier in this report, chlorine- and sulfur -containing salts
were found in deposits from the incinerator corrosion probes. Microprobe analyses
of corroded surfaces from the same probes have shown that these salts are very near
or adjacent to the bare metal surfaces. Thus, such salts can be suspected of being of
importance from the corrosion standpoint.
Experience in oil- and coal-fired boiler furnaces has amply demonstrated the
role of chlorine in fuels as an accelerating agent in corrosion processes. (38-42) Cor-
rosion by mixed sulfate- chloride salts has been observed. However, most of the
reported work had been done at temperatures above 1000 F and with alloy steels rather
than with carbon steels.
Most of the work at 1000 F was carried out with a base corrosion mixture of
and Na£SO *n ^e m°lar ratio of 3 to 1.
Detailed conditions concerning the corrosion mixture composition and gas atmo-
sphere along -with wastage rates at 1000 F are summarized in Figure 24. The lower
line on this figure also shows the results of the azide test for the presence of sulfide
on the corroded metals. A triple + ++ mark indicates a very strong test result. The
-------
corrosion is expressed in units of mils/month as calculated and extrapolated from the
weight loss values for 50 hours of exposure.
As can be seen in Figure 24, Run 3, the K2SO4-Na2SC>4 mixture with Fe2C>3
added did not accelerate corrosion above that obtained in flue gas alone (Run O).
However, when 1 percent NaCl was added, the corrosion was greatly increased as in-
dicated in Runs 1, 5, and 20. Runs 5 and 20 were repeat runs and demonstrate fairly
good reproducibility.
The addition of Fe2C>3 was made to determine whether reactions to form alkali
iron trisulfates took place. Such compounds are known to be corrosive at 1100 F.
Since the addition did not accelerate corrosion, it was concluded that the corrosive-
ness of the mixture or the possible formation of alkali-metal ferric trisulfates from
sodium and/or potassium sulfates, iron oxide, and sulfur trioxide was not of major
concern in the temperature range used.
As shown by Run 7, K^SO^ with 1 percent NaCl was less corrosive than the
mixture of K2SO4 and Na2SO4 with NaCl.
Increasing the chloride content had negligible effect although the action of KC1 on
A106 and Til steels at 5 percent level seemed to be somewhat less than for NaCl.
(Runs 10, 11).
It is significant that sodium chloride alone in the presence of flue gas was quite
corrosive to the carbon steels, i. e. , Run 13. Of even greater significance is the fact
that sulfide as revealed by the azide test, was found on these specimens. Thus, some
reaction involving the SC>2 in the flue gas was of importance since that was the only
source of sulfur. Sulfide had been detected, of course, in all preceding runs in which
corrosion had occurred. There was also a correlation between the severity of cor-
rosion and the amount of sulfide detected.
In order to get a better understanding of the possible reactions involving sulfur
and SC>2> several experiments were conducted with SC>2 eliminated. In addition, ex-
periment No. 21 was carried out where the SCu content of the flue gas was increased
from 250 to 2500 ppm.
As can be seen from Runs 15, 16, and 18, the elimination of SO2 from the flue
gas decreased markedly the corrosion for chloride alone and for sulfate-chloride
mixtures. Runs 16 and 18 were repeat experiments and it can be seen that results
are in quite good agreement. It is important to note that no sulfide was detected on the
metal surfaces of the specimens for the three runs just mentioned, i. e. , with
SO2 absent.
Similarly the corrosion was low and no sulfide was formed when the exposures
were carried out under helium with no SO2 present (Run 17). The addition of SO£
(2500 ppm) to the helium increased the corrosion several times as indicated for
Run 19. No sulfide was detected on these specimens.
In the experiment with the synthetic flue gas containing 2500 ppm SO2 (Run 21),
a slight increase in corrosion rate was measured on the A106 steel, but the corrosion
rates on the Til steel specimens were comparable to those when the level of SO2 in
the gas was 250 ppm (Run 10).
-------
8c
1-1
901 -V
1-1
901-V C
-i i
901 -V
ll-l
901-V
ll-l
901-V I
i-i
ON
ll-l I
1-1
l-l
901-V 1
ll-l C
901-V C
I7C
ll-l
MOI-W
IZ£ C
l-l
901-V 1
1-1 C
ll-l
901-V 1
I2£ I
i-i
muoiu jad spin '
. ,
8 i
CO CO _ £ O
o
(M 13
o
0
0
w
o
s
o «»
t\J
w
o
-------
70
In other experiments it was found that extremely rapid attack of Type 321 stain-
less steel resulted from lead chloride at temperatures near 1000 F. Metallographic
studies revealed intergranular penetration in addition to the general attack. The alloy
was also attacked by molten zinc and zinc chloride vapor.
The A106-Grade B and low-alloy A213-Grade Til carbon steels were attacked
about five times more rapidly at 1000 F when lead chloride was present than when it
was absent.
Incinerator Deposits
A few experiments were carried out in the laboratory using deposits taken directly
from the Miami County corrosion probes. For comparison, exposure tests were con-
ducted in synthetic furnace gas only and in Na^SO^ containing 1 percent NaC 1. Results
are summarized in Table 8. The corrosion which occurred is expressed, first, as a
weight loss in milligrams; second, as penetration in mils; and third, as a rate in mils
per month.
TABLE 8. LABORATORY CORROSION STUDIES WITH INCINERATOR DEPOSITS
Chemical
Environment
Gas onl/a)
Na2SO4 + 1 fy NaCl
in furnace gas
Deposit from
Probe 3 in
furnace gas
Gas only(a)
Na2SO4 + 1 °to NaCl
in furnace gas
Deposit from
Probe 3 in
furnace gas
Very low temperature
deposit from Probe 9
in furnace gas
Units of
Corrosion
Run 60 -
Wt loss, mg
Wt loss, mg
Penetration, mils
Rate, mils/ month
Wt loss, mg
Penetration, mils
Rate, mils/month
Run 61 -
Wt loss, mg
Wt loss, mg
Penetration, mils
Rate, mils/month
Wt loss, mg
Penetration, mils
Rate, mils/month
Run 63 -
Wt loss, mg
Penetration, mils
Rate, mils/ month
Steel Type
310
50 Hours at
0.4
6.3
0.07
1.0
348
3.9
57
250 Hours at
1.0
9.1
0.1
0.3
950
10.8
31
50 Hours at
S304
1000 F
2.8
9.0
0.1
1.5
417
4.8
70
1000 F
8.4
24.3
0.3
0.9
953
10.8
31
1000 F
304
3.4
39.5
0.45
6.6
411
4.7
70
12.7
104
1.2
3.6
994
11.3
33
446 A106 Til
8.2
10.6
0.12
1.8
375
4.3
63
11.9
34.8
0.4
1.2
775 1296 1586
8.8 14.7 18.0
26 43 52
219 80
1.0 0.4
15 5.9
321
14.5
0.07
1.0
(a) 80 % air, 10 %
CO2, 10 % H2O, 250 ppm SO2, 2000 ppm HC1.
Runs 60 and 61 were conducted for 50 and 250 hours, respectively. The deposit
mixture for these runs was a portion taken from Specimens 1 through 8 of Probe 3. It
can be seen in Table 7 that the incinerator deposit was much more corrosive than was
the Na2SO4-NaCl mixture or the furnace gas alone. It is also evident that the corrosion
rate was greater for the 50-hour exposure than for the 250-hour exposure. This cor-
responds to the observation discussed for the field work in that the corrosion appeared
-------
71
to decrease with increasing times of exposure. The results also show greater attack
in the A106 and Til steels than on the stainless steels, as would be expected. Results,
in general, agree fairly well with those obtained in the field.
The deposit sample used in Run 63 was taken from the water-cooled section of
Probe 9. The corrosion resulting from this deposit was much less than that from the
deposit taken from Probe 3.
These experiments were carried out at the same time as the stress-corrosion-
cracking studies discussed elsewhere in this report.
An oily liquid collected at the cool end of the furnace tube during these experi-
ments. This liquid was quite acidic and analysis showed it to be H^SO^.. It is believed
that compounds such as pyrosulfates or bisulfates in the deposits are being thermally
decomposed during the heating to release SO3, which then combines with moisture to
form HoSO^. In experiments with bisulfates and pyrosulfates described later, a
similar liquid condensed at the exit by the furnace tube, particularly when bisulfates
were heated.
Corrosion at 800 F
Seven experiments were carried out at 800 F under conditions similar to those
discussed in the series at 1000 F. Results are summarized in Figure 25. In general,
the accelerating effect of NaCl can be seen although wastage rates are much lower
than at 1000 F. The effect of varying the SO2 was not explored at this temperature.
E
g. 20
J2
E
o>~ 10
^Jl
U
> r\
CD
0
< =
t
c\j n i
ro
co n
1$ CO 0
~ CD £ CO O '
O *~ O T <
iE« ^ n^ On i
Run 08469
K2S04 % - 79 58 78 75
Na2S04 % - 21 16 21 20
Na Cl % - - 1 1 5
Misc. % - - 25FeP0,
-
CO H
o _
' CO
< O _
^~ -^
n-
-------
72
No attack on Type 321 stainless steel was observed at 800 F, although this
material had corroded in some experiments performed at 1000 F. The effects of
alkali-metal chloride addition and of the presence of SO? in the flue gas at 800 F were
similar, in general, to those observed at 1000 F with carbon steels. The corrosive
action of KC1 and NaCl at the 5 percent level was less at 800 F than at 1000 F.
Corrosion at 600 F
Since the field studies showed that fireside wastage could occur at metal tem-
peratures of 600 F and lower, extensive laboratory studies were carried out with a
variety of salt compositions. The standard synthetic flue gas was used, with additions
of HC las noted.
The most corrosive salts examined at 600 F were the bisulfates and pyrosulfates.
These were picked because of their low melting points and because they had been
demonstrated to be of importance in low-temperature boiler-tube corrosion in fossil-
fuel-fired boilers .(43) They were also likely suspects here because the field studies
showed that sulfur-bearing salts were involved in the corrosion.
The results with potassium bisulfate mixtures are summarized in the bar graphs
in Figure 26. It can be seen in Run 48 that appreciable corrosion occurred in the pure
salt and that the stainless steel was more severely attacked than were the carbon
steels. The reason for this attack on the stainless steels is not apparent. A base-line
comparison of a run without bisulfate can be found in Figure 27, Run 22. The addition
of 5 percent NaCl to the bisulfate increased the attack by a factor of 3 or 4 (Run 50).
The addition of 5 and 10 percent ZnCl2 along with the NaCl did not increase corrosion
(Runs 57 and 58). Similarly, the introduction of 2000 ppm HC 1 to the flue gas did not
accelerate corrosion (Run 59). The addition of ZnSC>4 and PbO to the NaCl -bisulfate
mixture decreased corrosion (Run 52) and so did the addition of PbCl2 and ZnCl2
(Run 55). It should be pointed out, however, that at 1000 F, PbCl2 and ZnCl2
accelerate corrosion.
It should be noted that all these mixtures were molten during tests and that the
corroded metals gave a positive test for sulfide.
As can be seen in Figure 27, K2S2C>7 caused greatly accelerated attack on car-
bon and stainless steels at 600 F (Run 28); however, dilution with inert components
greatly reduced the action. The addition of NaCl caused very little increase in the
attack (Run 44). The high rate of the attack on stainless steel was maintained when
ZnCl2 was added, but the attack on the carbon steels was decreased (Runs 36 and 41).
Other additions of oxides, chlorides, and sulfates decreased the corrosiveness of the
KzSzC1? (Runs 46, 40, 30, and 37). Figure 27 shows that the more corrosive mixtures
were those which had been molten during the test.
The mixture of KzSC^-NazSC^ and 5 percent NaCl which corroded the steels at
800 and 1000 F did not cause much attack at 600 F in 50 hours, as can be seen in
Figure 27 (Run 22).
-------
.111=
MMI V |
111
901V C
IXV I
II 1 |
yuiv I
1 cJt f
u i I
QMI W |
l^t [
II 1 1
401 V 1
\r/.(r \
II 1 1
901 V 1
IZt 1
901 V 1
1 1 1 1
3 O O O O C
) «a- ro CM
muow jed s|iui 'aBojsoM
m
m
CO
in
CO
m
en
m
r^
in
o
m
CO
^~
>
c
tic
OJ
O O .a - in i
m in ro a. ' CM '
m
CO O Q
o 1 c J3 in i
m m i N CL CM '
0 ^
ro
o
0 o 0 | J0. |
00 ' C\J '
._ o
^ 9 o
S in in 1 J(i 0
^ ^
o i m i
o
m i i m .
CD m ! ' c\j i
8111 § i
^ o-° 0>8 0^° | §.
__w Q_ QL
^ z N if ii co x
"S
+ r
. "O
+ Q)
+ T
+ if
, -0
+ ^
+ u_
~^J
T zj
+ Ll_
t ,=
+ u_
, -a
+
+ 5
u_
a>
P
"5
M -C
< Q.
o
o
xO
O
o
u
!2
O
*
O
in
O
EH
U
W
W
o
-------
I2£ 1
111 d
901V 1=
IZ£
111 C
901V C
me
901 V C
l|ODUD ON |7£
111 C
901V C
IXV 1
111 HZ!
901V d
I^V 1
111 C
901V C
\/.V 1
II 1 |
qoiw i
IX? 1
II 1 1
QOIW 1
IZ£ 1
111
901V 1
1 1 1 1
3000 o c
r> ' ' ' N
m
m | | | o
m
^ l l m l
CM ' '
^^^
i i i i
C^ i l l l
ifiS- '
wO CO
°8
m g
s ,
CM
CM
m 1
CM
CM
§ 1
^
10 1
1
CM
O
m |
CM
CO X
+ 1
CD
o
+ CD
+ -x.
+ 0
+ w
+ g
f" *"
+ if
J. (/)
T =>
+ U_
4- "S
+ u_
E
OJ
to ""
< CL
o
o
O
n
8
O
u
O
(M
CO
(M
O
EH
U
W
w
(M
w
o
-------
75
DISCUSSION OF CORROSION MECHANISMS
The results of studies of the deposits and scale formed on fireside tubes and
probes in incinerators combined with the laboratory studies have provided an explana-
tion of the cause of corrosion. Other discussions on the same subject can be found in
References 44 through 54. As would be expected, several reactions are coupled and
interrelated so that the overall picture is necessarily complex. The following para-
graphs outline the processes which are believed to be taking place.
The effects of various constituents are discussed individually; they can act in
combination in practice.
Effect of Chlorides and Metal Salts
It is significant that chemical, electron- microprobe, and X-ray analyses show
that chlorides are present throughout the deposit and at the scale/metal interface. The
fact that FeCl2 has been identified by X-ray analysis in the scale at the interface layer
over the entire temperature range studied is considered to be important. It is thought
that this iron chloride corrosion product is initially formed from the reaction of iron
with the hydrogen chloride or elemental chlorine released at the scale /metal interface.
Hazardous corrosive conditions from deposited chlorides have already been pointed
out by Cutler and his associates(^3)? particularly as they apply to burning fossil fuel
containing chlorine.
It is suggested that the corrosive agent hydrogen chloride is released by the
reaction:
2KC1 + SO2 + 1/2 O2 + H2O -K2SO4 + 2HC1 . (1)
The hydrogen chloride then reacts with the iron surfaces to form ferrous chloride as
indicated by:
Fe + 2HC1 ~FeCl2 + H2 . (2)
As indicated by the laboratory studies at Battelle and supported by the work of
Brown, DeLong, and Auld(^^' (see Figure 28), the corrosive effects from HC1 would
not be expected to be severe below temperatures of about 600 F.
Since Fed-, has been detected on corrosion-probe samples exposed at temperatures
well below 600 F, it is believed that some other mechanism is operating in this tem-
perature range.
Although elemental chlorine has not been identified in incinerator furnace gases
and would not be expected at high gas temperatures, it is believed that it may play a
role in the corrosion reaction. It is postulated that metal oxides, possibly Fe2O3 or
PbO, on the tube surfaces catalyze the reaction:
2HC1 + 1/2 02 -H2O + Cl2 . (3)
-------
w
£
l-H
A
O
J
ffi
u
5
j_
rj in
W^
H W
w Q
O
Pi
cd
O
O
oo
(M
W
o
II
fr
ffi
LIIUOUJ jad SIILU '
-------
77
The chlorine, which is formed only near the catalytic surface, then combines directly
with the iron by Reaction (4):
Fe + C12 »FeCl2 . (4)
It has been shown that once a layer of FeCl2 has formed on the iron surface,
reactions forming iron oxides (FegC^ and Fe3O^) and sometimes sulfides take place.
Microscopic examination of the scales has shown that a molten salt phase has been pre-
sent at points within the scale and as an interface layer between the scale and the deposit.
This, of course, would act as a barrier to the motion of gases in and out of the layer
adjacent to the metal surfaces. Under oxidizing conditions, it would be expected that
FeCl2 (and also FeS) would react with oxygen to form Fe2O3.(56, 57) Under reducing
conditions, on the other hand, it would be expected that this oxide would be reduced to
Fe3O4. It is suggested that once a thin scale has formed on the metal tubes, such a
cycling takes place. The mixed-oxide and molten- salt layers limit the availability of
oxygen and retain chlorine and sulfur within the adherent scale.
It has been shown by Fassler^-3^ that the reaction involved could be:
4FeCl2 + 3O2 -2Fe2O3 + 4C12 . (5)
The chlorine retained within the scale deposit then reacts further with the iron accord-
ing to Reaction (4).
Three observations suggest that chlorine may be involved. First, as was just
mentioned, ferrous chloride is found next to the corroded tube wall even at the low-
temperature end of the corrosion probe. Second, the reaction of chlorine with carbon
steel is extremely rapid at temperatures of 400 F and higher. The curve in Figure 28
taken from the work referred to earlier^") shows the high corrosion rate of carbon
steel in dry chlorine at low temperatures. Third, the Deacon Process (Reaction 3) is
favored at temperatures near 500 to 700 F in the presence of catalysts. Thus it is
suggested that chlorine reactions with steel are important at the low temperatures, and
that, possibly, HC1 and C12 reactions take place at the high temperatures.
The release of C12 and HC1 directly adjacent to the tube wall is an important part
of the mechanism just proposed. It is of course necessary to explain how the chlorides
reach the tube surfaces. It is suggested that the relatively short residence time in an
incinerator would favor the deposition of any chloride originally present in the refuse.
It has been well demonstrated that solid- phase chlorides can be volatilized during
burning processes. (58, 59)
The observation that measurable amounts of chlorides are found in incinerator
deposits agrees quite well with the work reported by Bishop'^"' on the combustion of
coal containing 0. 9 percent chlorine. He showed that initial deposits could contain up
to almost 40 percent sodium chloride at temperatures near 900 F.
Other chlorides in the deposits come via HC1 which is released when chlorinated
plastics such as PVC are burned. This HC1 then combines with the K2O and Na2
-------
78
It is proposed further that some of the chloride found on the corrosion probes
has been transferred as PbGl2 or ZnCl2. According to Sarvetnick("0) compounds such
as dibasic lead phosphate, dibasic lead stearate, lead sulfate, lead chlorosilicate,
dibasic lead phthalate, zinc 2- ethylhexoate, zinc laurate, and zinc stearate are added
to PVC as thermal stabilizers. They act in this capacity because of their ability to
react with the HC1 liberated from the polymer as it is thermally decomposed. This
can account at least in part for the presence of lead and zinc salts on the corrosion
probes. Additional lead and zinc compounds of course could come from the volatiliza-
tion of metal scrap in the refuse. They are probably first released as oxides in the
flame and then are converted to sulfates and chlorides on the tubes.
Part of the deleterious function of the lead and zinc salts, particularly chlorides,
could be the formation of molten-salt layers, since their melting points are low and
since when mixed with other chlorides (NaCl, KCl) they form eutectic compositions
which melt at even lower temperatures. These low-melting materials permeate and
destroy the initial oxide scale so that the substrate metal becomes exposed to chlorine
as it is released from the chlorides in subsequent reaction with furnace gases.
Effect of Sulfur Compounds
It is believed that sulfur-bearing compounds play an important role with regard to
corrosion at both low and high temperatures. For example, the most corrosive salts
studied in the laboratory at 600 F were the pyrosulfates and the bisulfates. It is sug-
gested that this high corrosivity is related to the fact that these materials have rela-
tively low melting points, i. e. , about 575 F for K2S2(~>'7 anc^ 415 F for KHSO4. Since
the corrosion reactions, which are electrochemical in nature, can take place more
readily in liquids than in solids or gases, the presence of a liquid phase is of great
importance. It is believed, as was implied earlier, that the melting points of these
salts is further modified, possibly lowered, by the zinc and lead salts also shown to be
present on the corroded samples.
It is suggested that one of the significant corrosion reactions causing wastage of
the metals in incinerators at temperatures near 500 F and higher is the same one pro-
posed many years ago by Corey and his associates'-^, 61) ancj recently by Coates'"^) to
account for boiler-tube corrosion in fossil-fired furnaces, namely:
Fe2O3 + 3K2S2O7^2K3Fe(SO4)3 . (7)
It is proposed further that the following reactions also take place to some extent:
2KHS04 + 3Fe ~Fe2O3 + FeS + K2SO4 + H2O (8)
K2S2O? + 3Fe^Fe2O3 + FeS + K2SO4 . (9)
In addition to the laboratory evidence showing the corrosivity of these materials,
it should be pointed out that the azide and pH tests made on the individual corrosion-
probe specimens show the presence of sulfide and the presence of acidic hydrolyzing
salts on the majority of the specimens. The presence of sulfide helps validate
Reactions (8) and (9) above.
-------
79
Sulfur was also detected with the microprobe at areas near the metal surface.
The fact that sulfide is formed in the corrosion scale suggests that the oxygen
pressures beneath the scale must be very low, probably of the order of 10" atmosphere.
This is indicated because it has been shown(57) that a sulfur pressure of 10" ^ atmo-
sphere is sufficient to produce FeS when SC>2 and C>2 pressures are below 10" ^"* atmo-
spheres. At higher SC^ or C>2 pressures, iron oxides form. When oxides are present
on the iron surface, higher sulfur pressures and perhaps moisture(°3) are needed to
form FeS.
The dimpled surface structure of the specimens mentioned earlier is related to
the presence of both FeS and FeCl2 phases at the metal/scale interface. It is suggested
that the reason for the dimpling is the variation in the corrosion through these phases
which coat adjacent areas on the tubes. Because of the cation vacancies in the FeS
structure, there is a higher diffusion rate of iron ions through the FeS film than through
the FeCl2 film, which is more ionic in character, has fewer cation vacancies, and thus
a lower cation diffusion rate. The driving force for this diffusion in both cases is the
iron concentration gradient established by the formation of the more stable iron oxides
at the exterior surface of these films. The oxide formation depends upon the SC>2 or C>2
penetrating the mixed-oxide scale and the other deposits. The presence of FeCl2 or
FeS at the metal/scale interface suggests that the environment at the metal surface is
sufficiently low in oxygen pressure to prevent formation of iron oxides which have
10 to 20 percent larger free energy of formation than either the chloride or sulfide over
the temperature range of interest. However, the chloride and sulfide are slowly
oxidized to Fe2C>3 at the surface away from the metal substrate so that the chlorine or
sulfur in the respective phase is increased and the reaction proceeds by attack of the
substrate metal.
The low pH of the waste deposits also suggests the presence of the acid sulfates
or pyrosulfates. Of course, zinc chloride and iron chloride would also give an acidic
reaction upon hydrolysis.
The presence of bisulfates, pyrosulfates, or alkali trisulfates has not been proved
by X-ray diffraction. This, of course, does not mean that they are not present, since
previous studies at Battelle and elsewhere have shown that these materials are difficult
to detect in small amounts. Dilution of these phases with normal sulfates reduces the
attack on the steels by very significant amounts.
The sequence of the chemical reactions discussed in the preceding paragraphs
has been summarized in Figure 29, which illustrates the proposed corrosion
mechanism.
-------
Deposit
S02 + 02 + H20
FIGURE 29. SEQUENCE OF CHEMICAL REACTIONS EXPLAINING
CORROSION ON INCINERATOR BOILER TUBE
-------
81
CONCLUSIONS
General
These field and laboratory studies have demonstrated that the wastage in water-
wall refuse boilers can be more severe than that normally encountered in fossil-fuel-
fired boilers. The complex nature of the refuse used as the fuel and the relatively
poorer control of burning in an incinerator combine to increase the possibility for corro-
sion. The contributors to the attack are corrosive gases and low-melting chloride and
sulfur-containing salts which exert a fluxing action on the protective films on the metal
surface. These low-melting salts primarily contain compounds such as zinc and lead
chlorides along with potassium bisulfate and potassium pyrosulfate. The data developed
reveal that the gases SO£, SO-3,, HC1, and C\2 are also playing a major role in the
wastage processes.
Analyses of tube deposits and furnace gases confirm the belief that sufficient
quantities of the deleterious salts and gases are present in all municipal incinerators
to warrant careful consideration from a corrosion standpoint.
Tube Wastage
The work carried out with the corrosion probes inserted in the incinerators has
demonstrated that the wastage rates of boiler-tube metals are directly related to the
operating metal temperatures. Some attack will be experienced in the 300 to 600 F
range and greatly increased rates can be anticipated as temperatures are raised to
1000 F. The temperature gradient between the metal surface and the furnace atmo-
sphere is also important. Thus, relatively low metal temperatures are to be preferred
in flame areas.
The corrosion rates shown on the bar graphs in earlier sections of this report
should not be used to project wastage values to be expected after long exposure periods
such as a year or more. This is because the initial rates are always high and the rates
decrease appreciably as the exposure time increases. The protective action of the
tightly adhering scale probably accounts for this decrease in attack. Thus, rates for
carbon steel tubes at about 500 F might be expected to range below about 10 mils per
year for extended exposure periods rather than the much higher rates calculated using
the bar-graph data. However, scale removal to improve heat transfer could adversely
affect the corrosion behavior.
While the stainless steels, particularly Types 310 and 446, furnish good resis-
tance to fireside corrosion, their use is limited because of the deleterious effects
which can take place during downtime under humid conditions. The Type 310 can fail
because of stress-corrosion cracking and the Type 446 can be severely pitted. How-
ever, only minor pitting occurred in incinerator operation. The other austenitic stain-
less steels evaluated - Types 304, 316, and 321 - also were more resistant than the
carbon steels but are subject to SCC. The Type 416 stainless steel did not perform
well enough to be seriously considered. The Inconel 600 and 601 materials were quite
resistant over the lower temperature region 300 to 500 F but were severely attacked
at higher temperatures and thus are not good choices.
-------
82
Incoloy 8Z5 gave encouraging results over the entire temperature range and is
much more resistant to SCC than the austenitic stainless steels so it could be considered
for use in high-temperature areas where the carbon steels would not be recommended.
While the aluminum, chromium, and inorganic coatings furnished some initial
protection to the tube surfaces they do not appear to be sufficiently durable to be con-
sidered for long-time boiler operation.
Incinerator Operation
Several criteria of boiler operation can be suggested from the research results
which should increase incinerator boiler life and reduce maintenance costs.
(1) It is concluded that water-wall incinerators should not be used to generate
high-temperature superheated steam, but should be operated at relatively low metal
temperatures, near 500 F, to minimize tube wastage. If high-temperature superheated
steam is desired for more efficient power production, it appears that there is merit in
using the refuse to heat the water in a separate furnace and do the additional heating in
another furnace with fossil fuel as the heat source. Thus corrosive deposits and gases
will be kept away from tubes operating at high metal temperatures.
It is of interest in this connection to compare the nominal operating conditions of
some of the water-wall installations on this continent with those in Europe and England.
Given in Table 9 are steam temperatures and pressures and the nominal tube-metal
temperature, arbitrarily chosen to be 50 degrees above the steam temperature, for
several stations operating and several being constructed. It can be noted that the
European practice, except for those at Rotterdam and Coventry, is to operate at rela-
tively high temperatures and pressures. Severe metal wastage has been reported for
some of these stations. On the other hand, the units at Norfolk have not shown corro-
sion. It is believed that the excellent performance of the Norfolk installation to date is
related to the fact that the operating pressure there has been kept very low, i. e. ,
175 to 200 psi. Allowing for a 50 F rise through the metal wall, this corresponds to a
metal surface temperature of about 425 to 440 F. Since the cool water flows to the
bottom of the wall tubes and then rises, the metal temperatures in the flame area are
probably even lower than 425 F.
It is interesting to note that the trend in design of new water-wall incinerators on
this continent has been in the direction of operation at moderate boiler tube
temperatures.
(2) It is concluded that downtime could be a critical corrosion period for a re-
fuse boiler in that hygroscopic salts on the tubes could become wet and cause attack.
Such attack, however, has not occurred at Norfolk where the two boilers are operated
alternately every other week. Nevertheless, it is believed that maintaining warm, dry
tubes at all times is worthwhile. As a corollary it would appear that operation of the
units 7 days a week, 24 hours a day would be optimum. Furthermore, metal tempera-
tures in the boiler and in the effluent gas stream should be kept high enough to avoid
either H^SO^. condensation or HCl/H^O dewpoint condensation. Staying above about
300 F should prevent the former and above 225 F for the latter.
-------
83
TABLE 9. NOMINAL OPERATING CONDITIONS OF WATER-WALL INCINERATORS
Location
Milan, Italy
Mannheim, Germany
Frankfurt, Germany
Munster, Germany
Moulineaux, France
Essen Karnap, Germany
Stuttgart, Germany
Munich, Germany
Rotterdam, Netherlands
Edmonton, England
Coventry, England
Amsterdam, Netherlands
Montreal, Canada
Chicago (N. W. ), Illinois
Oceanside, New York
Norfolk, Virginia
Braintree, Massachusetts
Harnsburg, Pennsylvania
Hamilton, Ontario
Steam Pressure,
psig
500
1800
960
1100
930
--
1100
2650
400
625
275
600
225
265
460
175
265
275
250
Steam Temp,
F
840
980
930
980
770
930
980
1000
680
850
415
770
395
410
465
375
410
460
400
Metal Temp,
F (Approx. )
890
1030
980
1030
820
980
1030
1050
730
900
465
820
445
460
515
425
460
510
450
(3) Although not examined in the current program, the European practice of at-
tempting to avoid all operation where reducing conditions can occur has much merit.
In this connection the control of the burning and the feeding of the refuse can be of great
importance. It should be pointed out that many years of research were required to
develop the best techniques for burning the fossil fuels coal and oil with a minimum of
corrosion. Fuel preparation and burning control are an important part of that tech-
nology. Similarly, if refuse is to be considered a fuel it should be treated as such and
not fired as it comes from the pickup trucks.
-------
84
INCINERATOR WET SCRUBBER CORROSION
(Research Grant EP-00325-03S1)
During Year 3 of the grant program, a supplemental study was conducted on
corrosion problems associated with wet scrubbers used with municipal incinerators.
This grew out of the observation that scrubber corrosion was as serous or even more so
than fireside corrosion. Two areas of study were pursued on this phase of the program.
First, a survey of the current corrosion status of many operating scrubbers throughout
the eastern United States was conducted by means of plant visits and correspondence or
by telephone. Second, an on-site corrosion study was conducted in the scrubber at the
North Montgomery County Incinerator, near Dayton, Ohio.
Corrosion Status Survey
The nature and extent of incinerator scrubber corrosion was discussed with the
operating personnel during visits to the following municipal incinerators:
Crookshank, Cincinnati, Ohio
East Hartford, Connecticut
East 73rd Street, New York City
Framingham, Massachusetts
Greenpoint, Brooklyn, New York
Montgomery County, North, Dayton, Ohio
Montgomery County, South, Dayton, Ohio
Miami County, Troy, Ohio
Waterbury, Connecticut.
Correspondence or verbal communication on incinerator operation and scrubber
corrosion has also been received from the following:
Almeida, Richard, Ft. Lauderdale, Florida
Axtell, Wm. R. , Ethyl Corporation, Louisiana
Hall, P. B. , City of Alexandria, Virginia
Hilsheimer, H. , Mannheim, Germany
Hollander, Herbert I. , Roy F. Weston, Engineers, Pennsylvania
Nowak, Frank, Stuttgart, Germany
Odle, James, Homes Road Incinerator, Houston, Texas
Sebastian, F. P. , Envirotech Corporation, California
Tuckett, Norman, Ft. Lauderdale, Florida
Velzy, Charles, O. , Charles R. Velzy Associates, New York
Whitewell, Joseph A. , Chemical Construction Corporation.
Useful information on incinerator corrosion was also obtained from:
The International Nickel Company
Engineering Foundation Research Conference, August 23-27, 1971,
Deerfield, Massachusetts
-------
85
Solid Waste Management Office, Incinerator
Corrosion Conference, April 12, 1971
ASME Incinerator Division meetings.
The assistance furnished by all of the above individuals and organizations is
gratefully acknowledged.
There is general agreement that the corrosion in wet scrubbers for incinerators
is an extremely severe problem. The rapid attack of most metals and alloys is not too
unexpected when the composition and complexity of the effluent gases and dusts coming
from the burning chamber are considered. These gases include corrosive constituents
such as HC1, SO^, SO,, HF, and organic acids. A detailed discussion of the gaseous
by products from an incinerator are found in earlier sections of this report along with
measurements of the composition of the dusts carried by the gases. Extensive mea-
surements of incinerator effluent gases have been reported by Carrotti and associates
during the past few years. (64, 65, 66) One of their earlier measurements at the
East 73rd Street Incinerator in New York City indicated an HC1 evolution of up to
8 Ib/ton of refuse. This incinerator did not use a scrubber but did contain water-spray
chambers. Thus, some of the HC1 was probably removed ahead of the stack where the
measurements were made. Since well-designed scrubbers will remove much more of
the soluble gases, it would follow that a relatively small incinerator of say 200 T/day
could be absorbing up to 1600 Ib/day of HC1 in the scrubber water. Large amounts of
sulfur-containing acids would also be absorbed.
The presence and concentration of HC1 is of particular concern because this ma-
terial accelerates pitting of most construction materials. It also limits the use of
austenitic stainless steels because of the possibility of stress-corrosion cracking. The
importance of HC1 in causing severe corrosion of equipment has been well documented.
Hanna and Curley, for example, showed that even trace amounts were damaging. (67)
Kear reported increased corrosion from HC 1 in combustion gases. (6°) Piper and
VanVliet some time ago pointed out that accelerated corrosion results from HC1 as the
water dewpoint was reached in a power station burning coal. (69)
There are many possible sources of HC1 in an incinerator. Some, but not all, of
this gas comes from burning plastics. The subject of combustion products including
HC1 from burning plastics has been widely discussed, as in References 70 to 79.
It has been found at many locations that the pH of the scrubber water soon drops
to a low value, i. e. , about 2. 0. This is particularly true if the water is recirculated
as is the practice in many places. (°0, 8 1, 82) An example of such an operation is at
Montgomery County, Ohio. Data furnished by that office have been tabulated in Fig-
ure 30, which shows the buildup of various constituents in the scrubber water at the
South Plant over about a 2-week period. It can be seen that the pH remained quite low
near 2. 5 and that the chloride rose from about 9, 000 ppm to 40, 000 ppm. The sulfate,
on the other hand, remained near 2, 000 to 3, 000 ppm and then abruptly rose to
25,000 ppm. Other dissolved solids, conductivity, etc., also rose as the operation
continued. Thus, with all these acidic components present, it is easy to see why the
corrosion is so severe in these systems. It is estimated that the buildup of chloride in
the circulating solution in the example cited above amounted to about 500 Ib/day.
-------
£ i is
2 I i
- i I «3
2 U 3
"> § -
^ 4. <->
E
£ ° -g §5 K
^ jfl^
TD $ « ^
-gS-g-o 2 0> 2 ot w 3
1- OT» <0
i C?cri2
- OT(rf^
X 3 (AJ»
^ ^ Ss
S " Z°
t- < OT
i ^ -c|-p-
. et $ -m-
E to _x & s vi
$. i- "!- «5 i vi
r1 I ^ >
^ CO CM H Jj
< s
i£ o
bfl
PJ f
M "
to ^ co g o
ocvi g S
I, fH
«> ^i^S * ?
co ro o Q
CO W
r (1J
h J3
o ^
^ 10 S rt
^BS o ^
g g
~S3 g 3
fe §
o S
u «
_ 31 10 I, *
0 CM g ^
n nj
en W ^
UJ 0> pi r$
" % £
^ 01
^ dj
0 U H
fc
| | 8" it 8" ff 1 1" if 1 g 1| §
~" ex o ^
Ludd'uO!|DJ;u90uoo 05 5 "o. ^
-------
87
05
H
O £
H ~
« i
w jj
ii cd
O £
§ 2
o
u
tf
o
o
H
£
O
O
O
ii
H
U
W
CO
W
Pi
13
O
ii
fc
T3
0)
t-t
(1) H
PJ P
to O
£H 4l
.2 bo
co CJ
s?
5 «
-------
88
Other factors leading to high corrosion in incinerator scrubber are
(1) Elevated temperatures (140-190 F)
(2) Abundant oxygen supply
(3) High velocities.
The survey showed that a valuable accumulation of corrosion data is available.
In summary, there is general agreement that
(1) Carbon steels, cast iron, brass and bronze, are rapidly corroded.
(2) Performance of stainless steels has been variable, ranging from
little corrosion to severe wastage and pitting.
(3) The more resistant alloys are Hastelloy C, titanium, and Inconel
625. However, they are quite expensive.
^4) Rubber and plastic coatings have shown some encouraging results,
particularly in piping application.
(5) Fiberglas-reinforced plastic pipe and structural components show
good resistance to the corrosive solutions as does acid-brick
construction.
(6) Both deposits and corrosion can be a problem on induced-draft fans.
(7) Corrosion can be a problem on stacks.
(8) All components and areas in the scrubber are susceptible to corrosion,
with the mist eliminator area being the most susceptible to attack.
(9) Concrete supports and flumes can be severely corroded by the hot
acid solutions.
Scrubber Corrosion Studies
Procedures
As summarized in the preceding sections a considerable background of corrosion
data for incinerator scrubbers is available. The search for this information, however,
revealed that no systematic study of stress-corrosion cracking had been conducted in
scrubber environments. Some plant failures, of course, had been traced to stress-
corrosion cracking. In order to gain a better understanding of corrosion and stress-
corrosion cracking in scrubber systems, a fairly extensive exposure program was
carried out in the North Montgomery County Incinerator. The specimens were im-
mersed in the hot water (170-190 F) passing through the concrete flume at the base of
the scrubbers. The racks holding the specimens were alternated between Units 1 and 2
as the incinerator operation schedule dictated. Figure 31 shows the location of the
-------
89
specimens on a section drawing of the plant. The pH of this scrubber water varied
from 4. 0 - 5. 1 as shown in Table 10, along with compositional data on the effluent
water. A duplicate set of specimens was also exposed in the same water at Columbus,
Ohio, under static, aerated, conditions at pH 2. 0-2. 5 to determine the effect of lower
pH. The following tabulation lists the eighteen metals and alloys evaluated in these
scrubbers. The compositions of the alloys can be found in Table 1. Most were chosen
because of their good corrosion resistance and resistance to stress-corrosion cracking,
although a few stainless steel composition were included for comparison purposes. The
specimen configuration used is illustrated in Figure 32, The specimens to the left are
stressed C-rings cut from pipe, those in the center are 4 x 4-inch flat plates con-
taining 2-inch diameter circular welds, and those to the right are U-bends cut from
sheet stock. The welded pieces provide an indication of weld durability as well as the
effect of stress.
Materials Evaluated in Scrubber
Stainless Steel Hastelloy
304 C
310 C-276
316 L, F
446 G
Inconel Carpenter
600 20 S
TypeS-816
Titanium
75A
6A1-4V
Armco
22-13-5
USS
18-18-2
The C-rings and U-bends were mounted in pairs with centers touching as shown.
This arrangement provided some indication of the importance of crevice areas. The
specimens were held between Teflon strips as illustrated in Figure 33. The C-rings
and U-bends were stressed with Type 304 stainless steel bolts. The welded plates were
supported on Teflon rods with Teflon spacers. These spacer areas provided additional
crevice areas.
A third set of stressed specimens of selected alloys was exposed at Columbus,
Ohio, in contact with deposit removed from the ID fans and housing and maintained
under hot, humid conditions above scrubber water.
Results From North Montgomery County
The specimens exposed in the scrubber water at North Montgomery County were
examined after 8, 17, 38, 63, 96, and 129 days. These specimens rapidly became
coated with a very adherent deposit which could be removed only by vigorous scrubbing
-------
vD
OO
OO
in
CO
H
oj
H
co
p
2
O co
1-1 rr1
CO ft
O W
O £
O pj
>H W
P CQ
p W
H D
co ccj
O U
W
CO
W
CO ^
f^ O
On
CO
co O
W ti
PH CTi
>H O
H U
CO
W
P4
8
-------
1885
FIGURE 33. MOUNTED SPECIMENS USED IN SCRUBBER STUDY
-------
92
with a bronze-wire brush. This coating very likely afforded some corrosion protection
to the specimens. Spectrographic analyses of the coating showed a major amount of
calcium and appreciable amounts of lead, barium, silicon, and aluminum. The low
solubilities of calcium, barium, and lead sulfates is consistent with the presence of
these deposits. X-ray diffraction studies confirm the presence of PbSCs and
CaSO4-2H2O.
A large portion of the water in this scrubber was re circulated, but a sizeable
make up was required. This is evident from the data in Table 10 which includes analyses
over a 4-month period. It can be seen that the chloride concentrations never exceeded
2400 ppm, which is ten times lower than was reported for the South Plant where com-
plete recirculation is employed. Furtheimore, the results for 12/8/71 in Table 10
show extremely low concentrations of dissolved salts. This is because the scrubber
during that period was being operated on a once-through water flow.
Details regarding each specimen exposed are found in Table 11 which lists the
first instance of corrosion and the severity of attack at the conclusion of the exposure.
Corrosion rates based on weight-loss measurements have not been presented because
the attack when present was always localized in nature, so overall corrosion rates
based on weight loss would be misleading.
The alloys are arranged with the most resistant ones at the top of each section
(C-ring, U-bend, circular weld) and the least resistant at the bottom. Pit depths are
shown for the plate specimens.
The most durable materials were Ti-6Al-4V, Hastelloy C, Inconel 625,
Hastelloy G, and Hastelloy C-276, as shown by good ratings, in two groups of specimens.
Some other alloys were rated good in one section but not in others. The least resistant
were Type 304, 304 sensitized, Armco 22-13-5, USS 18-18-2, 445 and Inconel 600. The
other materials were intermediate between these extremes. The unalloyed titanium
(Ti 75A) was quite resistant and the weld cracks reported in the table may have been
present in the as-welded specimens.
The only definite cracking as a result of exposure was on the Type 304, 304 sensi-
tized, and the Inconel 600 specimens. Figures 34 and 35 illustrate these effects after
63 days of exposure. In many respects the attack has the appearance of stress-
accelerated corrosion rather than stress-corrosion cracking, particularly for the
Inconel. Later sections of this report illustrate typical stress-corrosion cracks.
The appearance of the less durable C-rings and U-bends (Type 304, 304 sensi-
tized, 446, USS 18-18-2, and Inconel 600) after 129 days in the scrubber are shown
in Figure 36.
Cracking was not observed in the circular welded panels, but many of them were
pitted, often predominantly in the area near the weld. For example, Figure 37 shows
a fairly deeply pitted area on Type 316L near the weld. The back side of the same panel
after 63 days contained strings of pits as well as crevice attack at the support holes.
The strings of pits may be similar to those which finally produced leakage in the actual
scrubber wall.
Similar selective attack at the support hole crevices and the pits in the weld area
on the back of the Type 310 stainless steel panel were evident.
-------
93
TABLE 10. SCRUBBER WATER COMPOSITION DURING EXPOSURE PERIOD
Date
Measurement
Total hardness as CaCO-^ppm
Calcium as Ca, ppm
Magnesium as Mg, ppm
Sulfate as SO^, pprn
Chloride as Cl, ppm
Specific conductance ^mho/cm
Total solids, ppm
Suspended solids, ppm
Total acidity as CaCO3, ppm
PH
8/31/71
2330
568
262
1261
2355
7499
5916
114
402
4-4
10/8/71
2760
784
230
1308
2355
7639
5900
14
388
4-3
11/2/71
6250
1704
573
1933
1610
11, 160
11,386
10
352
5-1
12/8/7l(a)
272
71
27
231
183
833
732
114
68
4-0
(a) On this date the scrubber water was not being recirculated.
-------
94
TABLE 11. CORROSION RESULTS OF STRESSED AND WELDED SPECIMENS EXPOSED TO SCRUBBER
SOLUTIONS AT NORTH MONTGOMERY COUNTY,
Specimen
Alloy
T1-6A1-4V
Hast C
625
Hast G
(welded)
Hast C276
(welded)
316 L
Carp 20
Hast F
Armco
22-13-5
T75A
S816
USS 18-
18-2
825
601
310
316 L
304
304 S
600
446
Number
Jl
J2
Dl
D
Nl
N2
SI
S2
Rl
R2
HI
H2
Cl
C2
El
E2
Tl
T2
Kl
K2
Fl
F2
Ul
U2
Al
A2
Ml
M2
1C
1H
HI
H2
3G
3H
2G
2H
Bl
B2
4G
4H
Corrosion Results
U-Bends
Good resistance 129 days
Good resistance 129 days
Good resistance 129 days
Good resistance 129 days
Good resistance 129 days
Good resistance 129 days
Good resistance 66 days
Good resistance 66 days
Good resistance 66 days
Good resistance 66 days
Good resistance, one pit 129 days
Good resistance 129 days
Very fine pits 63 days, same at 129 days, good resistance
Very fine pits 63 days, same at 129 days, good resistance
Very fine pitting 63 days, slight edge attack 129 days
Good resistance 129 days
Good resistance 33 days
Good resistance 33 days
Weld-grains etched, possible cracks in weld, 129 days
Weld-grains etched, possible cracks in weld, 129 days
Slight roughening 63 days, same at 129 days
Slight roughening 63 days, same at 129 days
Bad edge and end-grain attack at 33 days
Bad edge and end-grain attack at 33 days
C -Rings
Good resistance 129 days
Good resistance 129 days
Good resistance, slight corrosion at machining lines 129 days
Good resistance, slight corrosion at machining lines 129 days
One pit near crevice 63 days, slight pitting near crevice 129 days
Slight pitting near crevice 63 days, same at 129 days
Some edge attack 63 days, worm track and edge attack 129 days
Edge attack 63 days, same at 129 days
Pitting 17 days, rows of pits across face 63 days, possible cracks 129 days
Pits 8 days, rows pits across face 63 days, possible cracks 129 days
Pitting 38 days, rows pits across face 63 days, possible cracks 129 days
Pitting 38 days, rows pits across face 63 days, possible cracks 129 days
Pitting 17 days, deep grooves 63 days, same at 129 days
Edge attack 8 days, deep grooves 63 days, same at 129 days
Pitting 17 days, severely corroded, one penetration 129 days
Pitting 8 days, severely corroded, deep edge canyons 129 days
Circular Welded Plates
Depth of Attack, mils
Ti-6Al-4V
625
Hast C
Hast F
S816
Carp 20
825
Armco
22-13-5
316 L
310
600
304
J
N
D
E
F
C
A
T
H
I
B
G
Pit
Good resistance, grains outlined 129 days 0
Good resistance 129 days 0
Good resistance 129 days 0
Good resistance 129 days 0
Good resistance 129 days 0
Some crevice attack 129 days 0
Some crevice attack 129 days 0
Pitting and crevice attack 33 days 2
Pitting and crevice attack 63 days, worse at 129 days 8
Pitting 38 days, deep pits and crevice attack at 129 days 14
Shallow pits 8 days, bad attack surface, weld, edge, 30
crevice at 129 days
Pitting 8 days, deep pits, more at heat-affected zone 129 days 20
Crevvce
0
0
0
0
0
4
5
15
20
32
28
31
(a) pH 3. 0-5. 1
Temperature 170-180 F.
Most specimens examined at 8, 17,
38, 63,-96, and 129 days.
-------
95 and 96
,?
J^i^S1*-
i
^^ - - ** '
I "", * '
^ T -*_'
o w
ro H
fl "fj
v <3
o ^
v W
rt
0)
if' ^ J
j*. --i
O
W
w
X
^ W rcj
S g W
w MI
10 J §
W P5
W O
H w
w tf
w
O
O
.*.<«**£'
'
$
*>^.
dff^»
«r-jt »_
-tv P
' O
' "
-* 4-"
f ^
O
H
o
ff5
W
ft
<
00
W
PS
o
ID
w
O
ft
><
W
rt
u
so
rt
o
in
X
-------
-------
97 and 98
W
Pi
en
O
X
W
Pi
W Pi
H W
h H
en ^
W
W
U
W
Q W
W Pi
en H
M m
4 CQ
Pi ^
Pi
U
en
o
o
vO
-1 2
W <
O P5
U W
~ O
O §
gg
^ en
5 Q
W <2
Si
W
Pi
D
O
O
-------
-------
3923
FIGURE 36. APPEARANCE OF MORE SEVERELY CORRODED SPECIMENS
FROM SCRUBBER AT NORTH MONTGOMERY COUNTY
The C-ring specimens were exposed for 129 days.
The USS-18-18-2 specimen was exposed for 33 days.
-------
100
Type 304 stainless steel was similarly affected, except that more linear strings
of pits were present.
Inconel 600 was usually severely attacked at corners and edges and crevices. It
will be recalled, however, that stressed C-rings of this alloy showed deep grooves.
Results From Columbus, Ohio
The scrubber solutions used in the exposures made at Columbus, Ohio, were
obtained from the scrubber effluent stream at the North Montgomery County Incinerator.
The solutions were changed three times during the 137-day period. A set of speci-
mens identical to those exposed at North Montgomery County was used. The specimens
were completely submerged in the scrubber water contained in a large Pyrex jar main-
tained at 170-180 F. Air was slowly sparged through the solution during the exposure.
The pH of the solutions was maintained at 2. 0-2. 5 by means of occasional additions of
sulfuric or hydrochloric acids. Specimens were examined at 20, 42, 95, and 137 days.
The results in Table 12 have been arranged and tabulated in a manner similar to that
used for the specimens described in the preceding section. The following materials
furnished the best resistance when rated on the basis of good performance in all con-
figurations in which they were run: Ti6Al-4V, Hastelloy C, Inconel 625, Alloy S-816,
and Hastelloy F, Hastelloy G, and Hastelloy C-276.
It will be noted that stress-corrosion cracking was found on the Type 304 panels
and C-rings. Figure 38 shows a crack running across the panel on each side of the
circular weld. The cracked C-rings are shown in Figure 39. Cracking was also noted
at an area in the weld and base metal of the Type 316L panel.
Stress-accelerated corrosion (trenches) running across the face of the specimen
was found for the Inconel 600 and 601 specimens, as illustrated in Figure 39.
Pitting occurred on many of the specimens as noted in Table 12. It will be noted
in Table 12 that most of the panels which showed bad pitting were also severely attacked
at the crevices under the Teflon spacers near the support holes in the corners of the
panels.
Fan Deposit Studies at Columbus, Ohio
As was pointed out elsewhere in this report, there have been field service failures
of induced-draft fans, particularly from SCC when austenitic stainless steels were used.
Carbon steel fans have also corroded rapidly by general attack and by pitting. This sub-
ject has been explored here experimentally by subjecting C-rings and U-bends of various
alloys to deposits taken from the fan housing at the incinerator at North Montgomery
County.
-------
101 and 102
15X
C4016
FIGURE 37. TYPICAL PITTING IN TYPE 316L PANEL EXPOSED AT
NORTH MONTGOMERY COUNTY SCRUBBER
FOR 129 DAYS
The pitting was worse in areas near the heat-affected
zone adjacent to the weld.
-------
-------
103
TABLE 12. CORROSION RESULTS OF STRESSED AND WELDED SPECIMENS EXPOSED
TO SCRUBBER SOLUTIONS AT COLUMBUS, OHIO
-------
104
Deposit Analyses
Deposits from both the blade and the housing of the induced-draft fan at the
Montgomery County North Incinerator were analyzed during these studies. Results are
summarized in Figure 40. These deposits were substantially different from those found
on the corrosion probes at the Miami County and Norfolk incinerators. The difference
results primarily from the fact that a wet scrubber has removed most of the particulate
material from the gas stream before it reaches the induced-draft fan. As a result, the
amount of iron, aluminum, silicon, and calcium found in the fan deposit is low. These
are the components of the clay-like materials in the fly ash which were removed in the
scrubber. The lead, zinc, and potassium content was high, and the compounds K.£Pb
(804)2 and K2Zn(SO^)2 were readily identifiable in the deposits by means of X-ray dif-
fraction. The presence of these compounds in the fan deposits shows that some particles
escape the scrubber water, as these sulfates are soluble but will precipitate as PbSCs
when exposed to water.
The chlorine content of these deposits was unusually high. In fact, the amount of
chlorine found in the fan housing deposit was greater than that determined in any of the
other buL< deposits analyzed on this program. The chloride concentration of almost
16 percent in the housing deposit would account for the severe attack by this deposit of
various metals used in our laboratory experiments with the deposit as described later.
This buildup of chloride in the ID fan deposits shows that the wet scrubber is not re-
moving all the HC1 vapors.
Corrosion
The stressed specimens were placed edge down and buried about one-half inch
in the deposit, which was held several inches above a sample of scrubber solution. The
entire assembly was contained in a closed vessel which opened at the top to take a re-
flux condenser. The scrubber solution was heated to 170-180 F to produce humid con-
ditions within the container. Thus, the stressed samples touched damp deposit and
vapors from the scrub water.
The results for the twelve alloys including carbon steel which were exposed up to
6 weeks are given in Table 13. Type 304 stainless steel and plain carbon steel were
severely pitted after 2 weeks and were removed, while Armco 22-13-5 was severely
cracked and removed at this time. The alloys have been arranged with the most resis-
tant alloys at the top of the list and the least resistant to the bottom of the table. The
degree of rusting, pitting, and cracking is indicated by numbers from 0-9. Pit depths
are also shown. The best resistance was furnished by T1-6A1-4V, Inconel 625, and
Hastelloy C. A small amount of pitting was observed on the Type S-816, Incoloy 800,
and Incoloy 825 specimens. Stress-corrosion cracking and pitting was detected by sur-
face examination of the Carpenter 20, Type 316L, USS-18-18-2, and Armco 22-13-5
specimens.
It can be seen from the results just discussed that the corrosion under the humid
fan deposits was more severe than in the scrubber water. This is undoubtedly related to
the presence of localized high concentration of chloride salts at the metal surfaces and
the absences of scale forming protective deposits.
-------
105 and 106
15X
.₯*
C4017
FIGURE 38. SCC THROUGH TYPE 304 STAINLESS STEEL
PLATE AND WELD
This panel was exposed in the scrubber solution
at Columbus for 137 days. A crack was
first observed after 42 days.
-------
-------
3924
FIGURE 39.
APPEARANCE OF CORRODED C-RING SPECIMENS
AFTER EXPOSURE TO SCRUBBER SOLUTIONS
AT COLUMBUS, OHIO
Stress corrosion cracking has occurred in the
Type 304 and 304 sens pieces. Grooves are
visible on the Iiiconel pieces.
------- An error occurred while trying to OCR this image.
-------
109
l\/\/\
\\y\/\/\
c
o
o
o
u_
CH
JKZ
iy\x
O
QJ
c
00
c
N
o
ro
O
-------
110
The most severe surface cracking was observed on Armco 22-13-5. More rusting
and pitting took place on the USS 18-18-2 alloy, although cracks were seen in the face
view at 7X.
The examination of cross sections was required in some cases to reveal the SCC.
This was either because of the presence of severe rusting, which masked the cracks as
for Type 304, or because the cracks were small, as for Type 316L and Carpenter 20.
It will be noted that the depth of the cracks in the alloy just mentioned when
viewed in cross section was about 40 mils. Similar penetration of stress-corrosion
cracks was observed with Type 316L stainless steel. However, in this case less surface
rusting occurred. The depth of the attack on the Carpenter 20 specimen was only about
5 mils.
Conclusions From Scrubber Studies
The studies of alloys under three scrubber exposure conditions have furnished
useful information on their general corrosion and stress-corrosion cracking behavior.
(1) The superior resistivities of Ti-6Al-4V, Inconel 625, and Hastelloy C as re-
ported by other investigators (such as the International Nickel Company) have been con-
firmed. Unalloyed Titanium 75A may also be good, but since some fine cracks in the
weld were visible it has been rated lower. It is possible that these cracks were present
in the as-welded material.
Hastelloy F and S-816 have also shown good resistance, as have Hastelloy G
(welded) and Hastelloy C-276 (welded). The last two alloys were exposed for only 66 days,
however.
(2) The studies indicate that the alloys mentioned in the above paragraphs are
also resistant to stress-corrosion cracking under the conditions used.
(3) Conditions present when metals are in contact with ID fan deposits are more
likely to produce stress-corrosion cracking than are the conditions present in scrubber
waters. Thus, the stainless steels - Types 304, 316L, USS 18-18-2, Armco 22-13-5,
and Carpenter 20 were cracked to varying degrees when in contact with these deposits.
In scrubber solutions, on the other hand, Types 304 and 316L were the only ma-
terials that cracked. This was true for both sensitized and unsensitized Type 304 ma-
terial. It is significant, however, that stress apparently had a marked effect on the
corrosion of Inconel 600 and 601 in scrubber solutions because deep grooves are
produced in stressed area.
Pitting and crevice attack seemed to be a more severe problem with many alloys in
the hot scrubber solutions. For example, pitting was noted on Type 304, 310, 316L,
446, Carpenter 20, USS-18-18-2, Armco 22-13-5, and Incoloy 825. The leakage seen in
some scrubbers constricted of 316L stainless steel reflects the pitting tendency of this
alloy in scrubber environments.
-------
Ill
No preferential attack was noted at crevices where the C-rings or U-bends
touched. However, severe attack as noted in the tables was found with many alloys at
areas under the Teflon spacers near the corner holes in the 4 x 4-inch panels.
Because of the cost of the more corrosion-resistant materials for incinerator
scrubber service, it appears that consideration should be given to the use of nonmetallic
construction materials as suggested by Ellison and Kempner. (83, 84, 85) The good per-
formance of such materials at the East 73rd Street Incinerator also is an important
consideration.
-------
112
ACKNOWLEDGMENTS
The cooperation and assistance of the following personnel in arranging the studies
at Miami County, Ohio; Montgomery County, Ohio; Norfolk, Virginia; and Oceanside,
New York are gratefully acknowledged:
Miami County Incinerator
Mr. Ronald Karnehm, Superintendent
Mr. Nick M. Brookhart, County Sanitary Engineer
Navy Public Works Center Salvage Fuel Boiler
at Norfolk, Virginia
CDR R. E. Deady, CEC USN Operations Officer
Mr. Wm. Osteen, Superintendent
Mr. Roy Evans, Foreman
North Montgomery County Incinerator
Mr. Charles Bennett, Superintendent of Incineration
Mr. Francis Tamplin, Superintendent
Oceanside Incinerator
Mr. Charles R. Velzy, and Charles O. Velzy, Charles R. Velzy Associates, Inc.
Mr. Howard Smith, Superintendent
Mr. Wm. Landman, Commissioner of Sanitation, Town of Hempstead
Mr. Al Vilardi, Chief Engineer
In addition, the assistance of Mr. Henry J. Gates, Vice President, International
Incinerator, in obtaining the deposit specimens from the Atlanta Incinerator was much
appreciated.
The encouragement and assistance furnished by personnel at the Solid Waste
Research Division, EPA - Mr. Robert C. Thurnau, Mr. Daniel J. Keller,
Mr. Alvin G. Keene, and Mr. Louis W. Lefke - were very valuable.
-------
113
PUBLICATIONS RESULTING FROM THE RESEARCH
Discussions on Paper, "Considerations in the Construction of Large Refuse
Incinerators", by F. Nowak, National Incinerator Conference, 1970.
Miller, Paul D. , and Krause, Horatio H. , "Fireside Metal Wastage in Municipal
Incinerators", ASME Paper 70-WA/Inc-2, Annual Meeting, New York City,
December, 1970.
Miller, Paul D. , and Krause, Horatio H. , "Factors Influencing the Corrosion of
Boiler Steels in Municipal Incinerators", Corrosion, 27 31-45 (January, 1971).
Boyd, Walter K. , and Miller, Paul D. , "Materials Selection for Design of Pollution
Control Equipment", presented at Design Engineering Conference, April 19-21,
1971, ASME Preprint 71-DE-12.
Boyd, W. K. , and Miller, P. D. , "Materials Selection for Pollution Control
Equipment", Engineering Digest, Vol. 17, October, 1971, pp 21-25.
Miller, Paul D. , and Krause, Horatio, H. , "Metal Corrosion in Incineration",
presented at AIChE Meeting, Cincinnati, Ohio, May 16-19, 1971.
Miller, Paul D. , and Boyd, Walter K. , "Corrosion in Wet Scrubbers in Municipal
Incinerators", presented at North Central Regional Conference NACE, Charleston,
West Virginia, October 27, 1971.
Miller, Paul D. , "Incinerator Corrosion", presented at Engineering Research
Foundation Conference, Deerfield, Massachusetts, August 23-27, 1971.
Miller, Paul D. , and Krause, Horatio H. , "Corrosion of Carbon and Stainless
Steels in Flue Gases from Municipal Incinerators", submitted to ASME for the
June, 1972, National Incinerator Conference.
Miller, Paul D. , Krause, Horatio H. , Vaughan, Dale A. , and Boyd, Walter K. ,
"The Mechanism of High-Temperature Corrosion in Municipal Incinerators"
submitted to NACE.
Miller, Paul D. , Krause, Horatio H. , Zupan, Janez, and Boyd, Walter K. ,
"Corrosive Effects of Various Salt Mixtures Under Combustion Gas Atmospheres",
submitted to NACE.
-------
REFERENCES
(1) Nowak, F. , "Considerations in the Construction of Large Refuse Incinerators",
Proceedings of the National Incinerator Conference, 1970, pp 92-96.
(2) Hilsheimer, H. , "Experience After 20, 000 Operating Hours - The Mannheim
Incinerator", Proceedings of the National Incinerator Conference, 1970, pp 93-106.
(3) Regan, J. W. , "Generating Steam from Prepared Refuse", Proceedings of the
National Incinerator Conference, 1970, pp 216-223.
(4) Fife, J. A. , "Design of the Northwest Incinerator for the City of Chicago",
Proceedings of the National Incinerator Conference, 1970, pp 249-260.
(5) Hansen, Erwin G. , and Rousseau, Henri, "An Engineering Approach to the Waste
Disposal Crisis", Combustion, 41, 8-13 (1970).
(6) Editor, "St. Louis Sees Trash Power", Electrical World, 173, 24
(April 27, 1970).
(7) Ficktner, Wolfgang, Maurer, Karl-Georg, and Miller, Harold, "The Stuttgart
Refuse Incineration Plant: Layout and Operation Experience", ASME Paper 66
WA/P1D-10, Annual Meeting, 1966, ASME.
(8) Eberhardt, H. , "European Practice in Refuse and Sewage Sludge Disposal by In-
cineration", Combustion, 28, 8-15 (September, 1966).
(9) Moore, H. Carlton, "Refuse Fired Steam Generator at Navy Base, Norfolk, Va.",
Proc. MECAR Symposium - Incineration of Solid Wastes, March 21, 1967,
pp 10-23.
(10) Larkin, John F. , "Waste Heat Utilization at Hempstead - Merrick Refuse Disposal
Plant", Proceedings of the Incinerator Conference, 1964, pp 95-98.
(11) Watson, R. H. , and Burnett, J. M. , "Municipal Refuse as a Fuel", Proc. Inst.
Mech. Eng. , 183, Part 1, 1-8 (1968-69).
(12) Pepe, P. D. , and Turner, G. M. , "Design of the First Large United Kingdom
Power Producing Refuse Disposal Plant", Proc. Inst. Mech. Eng. , 183, Part 1,
9-22 (1968-69).
(13) Editor, "Possible Recovery of Heat", Symposium on Refuse Incineration, Na-
tional Industrial Fuel Efficiency Service, London W-l, 1967.
(14) Nowak, F., "Corrosion Problems in Incinerators", Combustion, 40, 32-40
(November, 1968).
(15) Rousseau, H. , "The Large Plants for Incineration of Domestic Refuse in the
Paris Metropolitan Area", Proceedings of the National Incinerator Conference,
1968, pp 225-231.
-------
115
(16) Eberhardt, H. , and Mayer, W. , "Experiences With Refuse Incinerators in
Europe Prevention of Air and Water Pollution, Operation of Refuse Incinera-
tion Plants Combined with Steam Boilers, Design and Planning", Proceedings of
the National Incinerator Conference, 1968, pp 73-86.
(17) Matsumoto, K. , Asukata, R., and Kawashima, T., "The Practice of Refuse In-
cineration in Japan Burning of Refuse With High Moisture Content and Low
Calorific Value", Proceedings of the National Incinerator Conference, 1968,
pp 192-196.
(18) Day and Zimmerman Engineers, "Special Studies for Incinerators for the Govern-
ment of the District of Columbia Dept. of Sanitary Engineering", U. S. Dept. of
Health, Education, and Welfare, Public Health Service Solid Waste Program
Cincinnati, 1968, pp 55-73.
(19) Kautz, K. , and Kirsch, H. , "New Results of Corrosion Research in the Field of
Refuse Incineration", Mitteilungen VGB, 5j_, pp 223-227 (June, 1971).
(20) Tichatschke, J. , "Investigations of Emissions of Refuse Incinerating Installa-
tions", Mitteilungen VGB, 51, pp 219-223 (June, 1971).
(21) Defeche, J. , "Corrosions Produced by the Incineration of Domestic Refuse",
Schweiz. Zeit. Hydrologie, 3_1_, pp 488-508 (1969).
(22) Billings, Clayton H. , "Incineration of Refuse-State of the Art", Public Works,
102, pp 77-79 (April, 1971).
(23) Wisely, F. E. , Sutterfield, G. Wayne, and Klumb, David L. , "St. Louis Power
Plant to Burn City Refuse", Civil Engineering (January, 1971).
(24) Arpin, JeanV., "Montreals1 New Incinerator", APWA Reporter, pp 15-19
(January, 1971).
(25) Editor, "Milan Trims Garbage Costs With Power Generating Unit", Electrical
World, 170, p 32 (September 30, 1968).
(26) Windecker, G. , and Engelmann, F. C. , "Automatic Refuse Disposal Plant-
Hamburg", Chemical and Process Engineering, 52, pp 95-96 (June, 1971).
(27) Bender, Rene, J. , "Solid Refuse Dilemna-To Waste or to Recycle? ", Power
115, pp S27-32 (June, 1971).
(28) Editor, "Northwest Incinerator Plant is Largest Complex of Kind in Western
Hemisphere", Solid Wastes Management, 14, pp 150-152 (May, 1971).
(29) Bender, Rene, J. , "Steam-Generating Incinerators Show Gain", Power, 114,
pp 35-37 (September, 1970).
(30) Rogus, Cosimir A. , "Incineration With Guaranteed Top-Level Performance-
Harrisburg", Public Works 10\_, pp 9 2-97 (September, 1970).
(31) Editor, "Amsterdams" Automated Incinerator", Public Works, 101, p 67
(August, 1970).
-------
116
(32) Dryers, R. W. , andKerekes, Z., "Recent Experience With Ash Deposits in
Refuse Fired Boilers", 68-WA/CD4, The American Society of Mechanical
Engineers (1968).
(33) Szymanski, Walter A. , "Method for Rapid Detection of Sulfur Embrittlement of
Nickel", Materials Protection, 1, 100 (September, 1962).
(34) Harston, J. D. , and Scully, J. C., "Stress Corrosion of Type 304 Steel in
H2SO4/NaCl Environments at Room Temperature", Corrosion, 25(12),
493-501 (1969).
(35) Kirk, W. W. , Beck, F. H. , and Fontana, M. G. , "Stress-Corrosion Cracking
of Austenitic Stainless Steels in High Temperature Chloride Waters", Physical
Metallurgy of Stress Corrosion Fracture, (1959), pp 227-236.
(36) Logan, H. I. , The Stress Corrosion of Metals, John Wiley and Sons, New York
(1966), pp 100-106.
(37) Pickering, H. W. , Beck, F. H. , and Fontana, M. G. , "Rapid Intergranular
Oxidation of 18-8 Stainless Steel by Oxygen and Dry Sodium Chloride at Elevated
Temperature", Trans. ASM, 53_, 793-803 (1961).
(38) Bishop, R. J., and Samms, J.A.C., "The Corrosion of Austenitic Superheater
Steels by Alkali Sulfate/Chloride Mixtures in Air and in a Flue Gas Atmosphere",
Arbeitsgemeinschaft Korrosion, 26th Event of the European Federation of
Corrosion, Frankfurt, April, 1965.
(39) Shirley, H. T., "Sulfate-Chloride Attack on High Alloy Steels and Nickel-Base
Alloys", Proceedings of the Marchwood Conference, Mechanism of Corrosion by
Fuel Impurities, edited by Johnson and Littler, Butterworths, London, 1963,
pp 617-628.
(40) Alexander, Patricia A. , and Marsden, Ruth A. , "Corrosion of Superheater
Materials By Residual Oil Ash", ibid. , pp 542-555.
(41) Fontaine, P. I. , and Richards, E. G. , "Hot Corrosion of Nickel-Iron-Chromium-
Cobalt Alloys by Sulfate-Chloride and Sulfate Vanadate Simulated Fuel Ashes",
Hot Corrosion Problems Associated with Gas Turbines, ASTM STP 421, 1967,
pp 246-269.
(42) Bishop, R. J. , "The Formation of Alkali-Rich Deposits by a High Chlorine Coal",
J. Inst. Fuel, 41_, 51-65 (February, 1968).
(43) Corey, R. C. , Grabowski, H. A. , and Cross, B. J. , "External Corrosion of
Furnace Wall Tubes III. Further Data on Sulfate Deposits and the Significance of
Iron Sulfide Deposits", Trans. Am. Soc. Mech. Eng. 71, 951-963
(November, 1949).
(44) Rasch, R. , "Formation of Iron II Chloride and Iron III Chloride at High Temper-
ature Corrosion in Furnaces", Battelle-5, Frankfurt.
(45) Fassler, K. , Leib, H. , and Spahn, H. , "Corrosion in Refuse Incineration Plants",
Mitt. V.G.B., 48, 126-139 (April, 1968).
-------
117
(46) Wickert, K. , "The Accelerators of Corrosion in Furnaces", Warme, 74, (4),
103-109 (1967).
(47) deJong, K. H. , "Refuse Incineration Problems", Brennstoff-Warme-Kraft, 22,
518-529 (November, 1970).
(48) Teller, W. , and Bohne, H. , "Hydrogen Chloride in Flue Gases From Incinera-
tors", Mull-Abfall-Abwasser, No. 10, 28-29 (1969).
(49) Rasch, Rudolf, "Fireside Corrosion in Incinerators", Brennstoff-Warme-Kraft,
j_9, 498 (October, 1967).
(50) Hirsch, M. , "Fireside Corrosion by Burning Chlorine Containing Refuse",
Brennstoff-Warme-Kraft, 20, 499 (October, 1967).
(51) Huch, R. , "Hydrogen Chloride Corrosion in Incinerators",, Brennstoff-Warme-
Kraft, 18_, 76-79 (February, 1966).
(52) Angenend, J. , "Behavior of Materials for Boiler Tubes in Gases Containing
Hydrogen Chloride", Brennstoff-Warme-Kraft, J_8, 79-81 (February, 1966).
(53) Cutler, A. J. B. , Halstead, W. D. , Laxton, J. W. , and Stevens, C. G. , "The
Role of Chloride in the Corrosion Caused by Flue Gases and Their Deposits",
Paper 70-WA/CD-l presented at ASME Annual Meeting, New York, December,
1970.
(54) Reid, W. T. , External Corrosion and Deposits - Boilers and Gas Turbines,
Elsevier (1970).
(55) Brown, M. H. , DeLong, W. B. , and Auld, J. R. , "Corrosion by Chlorine and
by Hydrogen Chloride at High Temperatures", J. Ind. Eng. Chem. , 3_9_,
839-844 (July, 1947).
(56) Teterevkov. , A. I., Vil'nyanskii, Ya. E. , and Lapshinn, M. K. , "Action of
Oxygen on Ferrous Chloride at High Temperature", Zhurnal Prikladnoi Khimii,
43_ (3), pp 487-491 (1970); Translation UDC 661.41: 546.72: 546.131.
(57) Ross, T. K. , "The Distribution of Sulfur in Corrosion Products Formed by
Sulphur Dioxide on Mild Steel", Corrosion Science, 5_, 327-330(1965).
(58) Halstead, W. D. , and Raask, E. , "The Behavior of Sulfur and Chlorine Com-
pounds in Pulverized-Cool-Fired Boilers", J. Inst. Fuel, 42, 344-349
(September, 1969).
(59) Bishop, R. J. , and Cliff, K. R. , "Condensation of Sodium Chloride Vapor From
a Moving Gas Stream", J. Inst. Fuel, 42, 283-285 (July, 1969).
(60) Sarvetnick, Harold A. , Polyvinyl Chloride, Van Nostrand and Reinhold ( 1969),
pp 89, 98, 99.
-------
118
(61) Corey, R. C., Cross, B. J. , and Reid, W. T. , "External Corrosion of Furnace
Wall Tubes II. Significance of Sulfate Deposits and Sulfur Trioxide in Corrosion
Mechanisms", Trans. ASME, 67, 289-302 (May, 1945).
(62) Coats, A. W. , Dean D.J. A. , and Penfold, D. , Phase Studies on the Systems
Na2SO4-SO3, K2SO4-So3, and Na2SO4-K2SO4-SO3", J- Inst. Fuel, 41, 129-132
(March, 1968).
(63) Kubas chewski, O., and Hopkins, B. E., Oxidation of Metals and Alloys,
Academic Press, Inc. , New York (1953), pp 211.
(64) Carotti, Arrigo A. , and Smith, Russell A. , "Air Borne Emissions from
Municipal Incinerators", Contract PH 86-72-62 and PH 86-68-121, U. S. Depart-
ment of Health, Education, and Welfare, Bureau of Solid Waste Management
(July, 1969).
(65) Kaiser, Elmer R. , and Carotti, Arrigo A. , "Municipal Incineration of Refuse
With 2 Percent and 4 Percent Additions of Four Plastics: Polyethylene,
Polystyrene, Polyurethane, and Polyvinyl Chloride", Report to Society of
Plastics Industry (June 30, 1971).
(66) Carotti, Arrigo A. , and Kaiser, Elmer R. , "Concentrations of Twenty Gaseous
Chemical Species in the Flue Gas of a Municipal Incinerator", Paper, Air
Pollution Control Association, Atlantic City (June 27-July 2, 1971).
(67) Hanna, George M. , and Curley, Lawrence C. , "Corrosion of Combustion Equip-
ment by Chlorinated Hydrocarbon Vapors", Air Engineering, Vol. 7, pp 38-42
(April, 1965).
(68) Kear, R. W. , "The Effect of Hydrochloride Acid on the Corrosive Nature of
Combustion Bases Containing Sulfur Trioxide", J. Applied Chem. , Vol. 5,
pp 237-242 (May, 1955).
(69) Piper, John D. , and Van Vliet, Hagen, "Effect of Temperature Variation on
Composition, Fouling Tendency, and Corrosiveness of Combustion Gas From a
Pulverized Cool Fired Steam Generator", ASME Paper 57-A281, ASME
Annual Meeting (December, 1957).
(70) O'Mara, M. M. , Crider, L. B. , and Daniel, R. L. , "Combustion Products
From Vinyl Chloride Monomer", American Industrial Hygiene, 32, pp 153-156
(March, 1971).
(71) Boettner, E. A., Ball, Gwendolyn, and Weiss, Benjamin, "Analysis of Volatile
Combustion Products of Vinyl Plastics", J. Applied Polymer Science, 13,
pp 377-391 (1969). ~
(72) Coleman, E. H. , and Thomas, C. H. , "The Products of Combustion of
Chlorinated Plastics", J. Applied Chem. , 4, pp 379-383 (July, 1954).
-------
119 and 1ZO
(^3) Ranby, Bengt, "Technical Environmental Data and Plastics", Fosfatbolaget-
Sweden (1969).
(74) Leib, Heinz, "Corrosion by Burning Plastics", SKYDD, Swedish Fire Protection
Association, Symposium, pp F1-F5 (April 23, 1969).
(75) Gralen, Nils "Plastics From an Environmental Standpoint", IVA Meddelunde,
Royal Swedish Academy of Sciences, Stockholm (September, 1969).
(76) Fuller, William R. , Bieler, Barrie H. , and Morgan, David C. , "Method of
Reducing Halogen Emissions From the Incineration of Halogen Containing
Plastics", U. S. Patent 3, 556-024, Dow Chemical Company.
(77) Fong, G. P., Mack, A. C., and McDonald, J. N. , "Role of Plastics in Solid
Waste-A Status Review", Proc. Inst. Solid Waste, Framingham, Massachusetts,
pp 53-79 (May, 1970).
(78) Editor, "Can Plastics be Incinerated Safely?", Environmental Science and
Technology, j>, pp 667-669 (August, 1971).
(79) Heimburg, Richard W. , "Environmental Effects of the Incineration of Plastics",
AIChE, Sixty-Eighth National Meeting, Houston, Texas (February 28
March 4, 1971).
(80) Cross, F. L. , Jr. , and Ross, R. W. , "Effluent Water from Incinerator Flue-
Gas Scrubbers", Nat. Inc. Conf. Proc., pp 69-72 (1968).
(81) Schoenberger, R. J. , and Purdom, P. W. , "Characterization and Treatment of
Incinerator Process Waters", Proc, Nat. Inc. Conference ASME, pp 204-215 (1970).
(82) Achinger, W. C. , and Daniels, L. E. , "An Evaluation of Seven Incinerators",
ibid, pp 32-64.
(83) Ellison, W. , "Control of Air and Water Pollution from Municipal Incinerators
With the Wet Approach Ventur Scrubber", ibid, pp 157-166 (1970).
(84) Ellison, William, "Wet Scrubbers Popular for Air Cleaning", Power, 115,
pp 62-63 (February, 1971).
(85) Kempner, S. K. , and Seller, E. N. , "Performance of Commercially Available
Equipment in Scrubbing Hydrogen Chloride Gas", Journal of Air Pollution Control
Association, 2_0; pp 139-143 (March, 1970).
(86) Task Group T-70-8, "Procedures for Quantitative Removal of Oxide Scales
Formed in High Temperature Water and Steam", Materials Protection, 6,
pp 69-72 (July, 1967).
-------
-------
APPENDIX
-------
-------
A-l
APPENDIX
Probe-Specimen Handling
The 34 corrosion specimens were separated from each other by tapping with a
fiber mallet. The amount of metal wastage was determined by chemically or electro-
chemically stripping the residual deposits and scale from the majority of the specimens
using standard procedures. (86) The laboratory specimens of Til and A106 were
stripped in Clarkes Solution: 100 g cone HC1, 2 g Sb2O-j, 5 g SnC^ at room tempera-
ture. Most specimens were stripped cathodically in 10 percent f^SO^ containing
1 -ethylquinolinium iodide inhibitor.
Some stainless steel specimens were descaled in a two-step process: first with
11 percent NaOH and 5 percent KMnO4 at 212 F and then with 20 percent HNO3 with
2 percent HF at 130 F. Then the weight loss and dimensional changes were measured.
The original weights of these specimens were in the range of 65 to 80 grams.
Experimental Results
Miami County
Tables A-l, A-2, and A-3 present wastage data by weight loss and penetration for
Probes 1, 2, 3, 4, 5, 6, 7, 9, 12, 13, 14, and 15.
Norfolk
Table A-4 presents wastage data for Probes 8, 10, and 11 exposed at Norfolk.
Laboratory
Tables A-5 and A-6 present corrosion results for studies made in the laboratory
at 1000 F and 800 F, respectively.
Tables A-7, A-8, and A-9 present corrosion results for 600 F for steels A-106,
Til, and Type 321, respectively.
-------
A-2
CO CO CD
ii ii CD ^f O1
CXJ LO ^3- LO "ZT
r-^ LO LO CD <i co CD <« r-. CD CD CD CD *=* co o~> CD cn cxj «=a- cxi CD CD CD co .t LO CD
LO OJ CXI LO LO CO OO CD 'I "=* CO CXI OO LO CO LO* LO LO LO LO CD LO CD CO CD I* CD CD
CD cn r
i cO ro
co 01 CD i-f~3 CD CD i CD co LO LO r < r i i
oo LO CD
LO CD TH CO ^ CD CXJ
«=3- LO CD* *=3T LO -«£r LO'
T* °?
co r"
CD « CD
3 P 3
r-. CD
T *-^ i cj2 T" cD CNI Cxl CVJ Cxl CXJ CN| ^ ^ CNJ ^ c^3 oo ^^ ^^ ^
} co ro <£. I ^ co co co c-o co co co co co co CO co co co co
CO
O
Q_
>-
CO
CXJ
CO
CD LO ii CD CD CD
CO LO r~-~ LO CD CD
CXI CO CD CXI CD LO CO CXI r-^ "d-
CO CD OO CO CXI CO ^J- CO CD CD
J2
LO CD -J^CD "=rcxj^-(COCDCXjr--;cDCO
j-^ CD ^" CD LTi CO CD ^^" CD >' LO CD CD
o
u
^3~ CO OO
LO CO ' ' LO !
cxi co co " ^r CD i. cxj r-~^
' CD CD CO LO CXI OO CO "-0 CD
CO CO CO «" i" O~> CXJ CO
P-* Cxt CO Cxi CD » CD r~~~~
z.
o
o
'i co LO r-^ CD
ti^CO^CD^CD^CD^I"^CDc^c>g':^C^C^c^C^C^
C3^C3^HCD^HCD-HCD^Hp_)o^^l.^H^a.F_ll(J.i_l+^L^H.^J.'
CD cxi co CD cO *=!- co r "^
LU
CO
Cxi LO OO CT>
LO cjD CO O~)
CD CO
CO oO
CD CD
CD
co ro
^=r CD CD
CD CT1 CO
111
_l
00
<
I-
CD CD CXJ
CO LO LO
i CO CD I -- CO <"*">
LO CD CC> LO LO '
CO CO CD r-- CXJ
' OO "1
LO co LO CD r~-~
LO "=J" CO CO
CO CO O~> ^"
aS
S <="
CD i" CD
CXJ
CO
CO CD CD CD "=3~ CXI LO
r-. co oo
r-. co < cxi CD
co r. co «=3- .i CD CD r-^ LO cxi co LO o*> co i CD CD CD r-~ LO CD
r~- co co LO ^J- CD <* co CD
co LO -^r I CD cr> CD cp co co I
CD CD cxj Ti .i cxi cxi *=3~ co *=J- co r~-~ r- LO CD co < LO
CD co CD CT> cn I LO co *=3- CD LO >i r CD LO I I r-~. CD '* r-^ '*
CD CD ' < CXJ i i «=*
CDVcD^CD^
i CD ' CD CD >
i CO Q-) CXI ,_(
CO P- CO CD CD - i CXI CO
-------
A-3
cxj CD en LO CD co CD CD
CO -=3~ CD CD OO CD CO
CXI CX/ ,t ,i ,i cxj CXJ it
CD LO =zT CXI «=J" CXJ LO
CO *=T LO CD CD CO «=t-
co i en r-. ^Q, t CD ^_ co co
' LO co co ' LO i< r~- co '=$~ LO I <_o co co ' oo ' r~~- oo *f CD CD co
.i ,i .i cxi cxj ii cxi cxi ,i cxi cxi ' -i
s
o
CL
CD f^-i OO LO CD CD CO rO CVJ CJD LO rO LO CT^ .I
: i ^ i * to r~: ! OT i
CD CD co i ' co i
01 ' LO r co cxi i to
O I CD ^-< CD
5 P <: K 3
CD
co
' _
co 1 * CXl OO *=J~ LO CD P-- CO CT)
CD CD CD ' ' C* CD CD CO ^f CD CD CD CD CD CO
i 1 , 1 *=T ' ' ** r-H , 1 *:e- CD « ' 1 "=3- *« ' 1 ^
CD =* cxj . i LO
CO CO CO CO
CD
CO
c^
UJ
00
i CD CO CO
cxj ( CD ' CD cxi co r^- cxi co r-. co o~>
CD ^J~
CO CO
.( CO
CNJ CO
LO oo "=* oo cr>
co CD CD LO oo ~=r
CO CXJ CXI LO CO
CD co 03 r-! 06
OO CO (Xj
CO CO CO
CO «=3-
CD CXJ
CO LO CNJ CD
r- LO en oo
CO
s.
RJ
f
cD
oo
------- An error occurred while trying to OCR this image.
-------
A-5
.8
V)
UJ
CO
o
CD i-n cxj cj-> to
co LO to r-~-
C-J ' CO CD CO LO
I
cxi i r i' cxi co «t to ro co CD co '=r ro CD > Cxj
Cxi cxi cxi "3" i cxi LO uo i t LO r- to
^J CXI CXI CD CXI CO "' LO CO *=T CXI CO CO
_o LO oo r~~ ' ro cxi a~) i co i co
ro to
-r ^ s ^ "? 5 ^
to i to . to c^j r7~
CD " CD 4 CD CD 03
5 P ^ P < < 5
CD ^j- CD
o
li.
cc:
o
o
UJ
o
UJ
UJ
_1
CO
-i CXI ^-» CO C5 CD '
CO CO CO
CO CO CO
CO CO CD CO
CD CD CD
=3- CD
CD " CD
00 CD , '
c^ cxi cxi ro cxi t_o
> CD CD CD CD' CD CD
CD CD CD
co cq |
CD CD CD*
^ c^
OJ CXI CXI
CXI LO CO LO CQ
rO i_O LT5
" cxi cxi cxi CNJ '' '* cxi Cxi ro
< CZ3 CD
CXI r« CD CD CO cxi
CD CD CXI
cxicococo'cocxi--cxi
QO CO ' * CO CXI
CD i cxi cxi cxi co .
i CD <~> ' ' "^-t
LO CO CO CZ> CO ~=T CO CXI
CD f~~* CXI CXI <" CD C3 CXI
' -' _
co co I
icxicxicxic^cxirxicxicxicxjcococococo
I ^coro-^Tcocococo-^-cocococo-^-coco
-------
A-6
TABLE A-5. LABORATORY CORROSION RESULTS AT 1000 F
Run
0
3
1
5
20
7
10
11
13
15
16
18
17
19
21
0
3
1
5
20
7
10
11
13
15
16
18
17
19
21
0
3
1
5
20
7
10
11
13
15
16
18
17
19
21
Atmosphere
F.G.(C)
F.G.
F.G.
F.G.
F.G.
F.G
F.G.
FG
FG
F.G. No S02
F.G. No S02
F.G. NoS02
Helium
Helium . 2500 ppm S02
F G. with 2500 ppm S02
FG
FG
F.G
F.G.
FG
FG
F.G.
F.G.
FG.
F G. No S02
F.G. No S02
F, G No S02
Helium
Helium 2500 ppm S02
F, G .2500 ppm S02
FG
FG,
FG
FG
FG.
FG.
FG.
FG.
F G.
F G No S02
F G No S02
F G No S02
Helium
Helium 2500 ppm S02
F.G. * 2500 ppm S02
_K2!°4_
59
58
78
78
99
75
75
-
-
78
78
78
78
75
-
59
58
78
78
99
75
75
-
-
78
78
78
78
75
-
59
58
78
78
99
75
75
-
-
78
78
78
78
75
Corrosion
Na2S04
16
16
21
21
-
20
20
-
-
21
21
21
21
20
-
16
16
21
21
-
20
20
-
-
21
21
21
21
20
-
16
16
21
21
-
20
20
-
-
21
21
21
21
20
Mixture, percent
NaCI
A- 106
-
-
1
1
1
1
5
-
100
100
1
1
1
1
5
A213-
-
-
1
1
1
1
5
-
100
100
1
1
1
1
5
_Typ_eJ21_
-
-
1
1
1
1
5
-
100
100
1
1
1
1
5
KCI Fe203
- Grade B Steel
_
25
25
_
-
_
_ -
5
-
-
-
-
-
-
-
Grade Til Steel
-
25
25
-
-
-
-
5
-
-
-
-
_
-
-
Stainless Steel
-
25
25
-
-
-
-
5
-
-
-
_.
-
-
~~ ~~
Weight Loss(a),
mg
Spec'men
1
109
80
211
1074
474
281
1003
489
403
192
236
307
56
362
1327
58
65
995
882
704
270
989
536
1243
142
143
120
176
406
1112
<1
13
61
287
217
17
256
509
No
18
17
10
64
67
114
2
110
84
507
1148
992
299
1055
574
508
146
235
293
87
360
1675
76
66
1060
986
607
268
863
562
1159
113
148
104
158
404
955
<1
9
123
160
280
10
251
372
attack
15
10
9
37
59
70
Calculated
Corrosion
Rate(b\mi!s
per month
Specimen
1
8
5
14
71
31
19
66
32
27
13
16
20
4
24
88
4
4
60
58
47
18
66
36
82
9
10
8
12
27
74
0
1
4
19
14
1
17
34
1
1
1
4
4
7
2
8
6
34
76
66
20
70
38
34
10
16
19
6
24
111
6
4
70
65
40
18
57
37
77
8
10
7
11
27
63
0
1
8
11
19
1
17
25
1
1
1
3
4
5
Azide
Test
Results
neg
,7
H
H-t
H
f
t-H
t-H
H-t
neg
neg
neg
neg
neg
H4
neg
\'i
H
m
f*
t-H-
f ,
\ i
neg
+7
neg
neg
neg
"
neg
neg
(
+i-
t
t
-t-H
neg
neg
f
neg
neg
neg
t-
(a! Specimen size, 0.8 x 0.75 x 0.125 inch, weight range, 8 to 10 grams.
ib) Based on exposure of 50 hours.
id F.G. flue gas 80°0 air, 10°0C02, 10°0H20, 250 ppm S02.
-------
A-7
TABLE A-6, LABORATORY CORROSION RESULTS AT 800 F
Run
0
8
4
6
9
12
14
0
8
4
6
9
12
14
0
8
4
6
9
12
14
Atmosphere
F.G.(C)
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
F.G.
K2S04
_
79
58
78
75
75
-
79
58
78
75
75
-
_
79
58
78
75
75
-
Corrosion
Na2S04
_
21
16
21
20
20
-
_
21
16
21
20
20
-
_
21
16
21
20
20
-
Mixture, percent
NaCI KCI Fe203
A=106 - Grade B Steel
_
_
1 - 25
1
5
5
100
A 213 -Grade Til Steel
_ _ _
_
1 - 25
1
5
5
100
Type 321 Stainless Steel
_ _ _
- - -
1 - 25
1
5
5
100
Weight Lo
mg
Calculated Corrosion
,ss(a)r Rate(b)>
mils per month
Specimen
1
9
18
47
69
264
98
34
8
9
14
37
164
233
49
No attack
No attack
3
2
0
0
No attack
2
__
18
44
99
264
110
77
9
24
34
273
152
62
4
0
1
2
Specimen
1
1
1
3
5
18
7
2
1
1
1
3
11
15
3
_
-
<1
<1
0
0
0
2
_
1
3
7
18
7
5
_
1
2
2
18
10
1
_
-
<1
0
<1
<1
0
Azide
Test
Results
neg.
+
+
+
-H-
++
+-H-
neg.
+
+
+
4-+
-H-
-t-H-
neg.
neg.
neg.
neg.
neg.
neg.
neg.
(a) Specimen size, 0.8 x 0.75 x 0.125 inch, weight range, 8 to 10 grams.
(b) Based on an exposure of 50 hours.
(c) F.G. - flue gas 80% air, 10% C02, 10% H20, 250 ppm S02
-------
A-8
-o -a -o
: Q; ^_ CO CD tO OO CD CO i CD CD O~> tO i-O CO CO LO OO r~~. =* CD CO CD
co ro v.xj 'i cxj cxi cxi ' i cxi to cxi LO ro -«a- CD to cxi LO cxj co
co ro <3-,i^d-.i ro ,i
CQ
HJ
O
O
I
1 I I I t I I I I I
o
JD.
O-
o
T
O
u.
o
M
to
111
oe.
I I I I
CD LO CD
O
o
O
u
o;
O
I I I I
I I I I I
Oi
O
ca
UJ
_i
CD
^ f I
I I I I I t I
XXX
XXXXXX
XXX
1111111(1
Cxi CXJ CXJ CXI CXI CXJ CXJ Cx! CXJ CXI CXI CXI CXI CX! CXJ Cxj CXJ CXI CXI Cx
ICxJcxJcxjCXICxjCXJcxjCxICxiCxJcxlCxICXICxjCxICXJCXJcxjcxI
LO LO LO
Cxi CXJ Cxi
XXXXXXXXXXXXXXXXXXXXCD x: X y< X. X. X
cxlcxjcxjcxjcxicxjcvjcxlcxicxlcxjcxicxicxicxicxicx)
X X X X X X X
LO LO LO LO LO LO LO
CXI CXJ CXI CXI CXJ CXJ CXI
-------
A-9
o
cDOJo_>ajcucD_=:
00 00
CXI CXI CXJ ,_ CXI .
ro -^r «^f- co i i
*~
Oi
o
Of
o
o
CXJ
5
I I I I I I I I I I
I I I I I I I I I I I
O O
Cxl 00
I I I I I I I I I I I I I I I I I ^ I I
i i i i i i i i i i i i i i i i i i i i 2 i i i i i i i i i i i i i i i i
Ul
Of
g
00
o
Q£
tt
O
u
a
o
CO
CO
I
UJ
CO
-
"^ I I I I I II
I I I I I I I I I I I I I I I I I I S2 I I I I I
I I t I t I I I I ( \ I I I "° I I I I I I 53 I
<^ i i
1 ' ' ' * '
1 t J 1 1 1
1 ' ' ' ' '
OJ OJ
1 1 1 1 1 1 1 1 1 1 1 1 1 1
CD CD CD
X X ?< X X K
Cxi OJ OJ rxj OJ OJ
1 1 1 1 1 1 1 1 1 1 I 1 1 1
XXX
cr> co en CD en
-- -~
-------
A-10
00 =3 =3 =3 CO ID =1
cxb oi oo ob ao
l/>
UJ
Z
"~l
t/)
CO
Ul
Q.
V
0
LL
1- §
i ^
-J 0-
3 0
o °
o;
o
u
>-
o£
0
<(
Of
o
CO
^
-1
0^
1
UJ
1
-^ 31
OQ c
<^_ O C
0 Cxi <_
1 CD n= 3:
-£I CD
(/> "~1
E o
^ 0 0
Q 31
O
°-
c:
LO LO LO LO LO
OCDC^CPCDCDCDOCDCDCXJOOOC3^0^SC30C:>OC::>
OCDCD0CDCDC.CDCD^-CD^CDcSScD^Cn0ScDCDCD
Cxi O~l <-:) ^ °^
CXJ
CO CD CD CD CD CD CD CD CD ' ' LT> CD Q CD *^t" CH CD ^T LO cP . t CD CD CD
CVI CO
0
C^J
11111 l-llll ' »r ' 0*" ' S
^r,^3- S S CD Q- -^-,
t^-0 TOTO FXlv^ CUOJC>-QQ_
C^> , '-'" ^~ > ^- ( )
L^LJO LJ-)U^ . i c^, U^UD. ^LJOC^J
i i i i i i i i i i i i i i i i i2 i i i i i i i
i i i i i i i "" i i i i i i i i i i i i i S i i
1 1 1 1 1 1 1 I^^^^I^^^O^LO, , , , |
j I LO LO U"> LO LO LO LO LO LO i LO LO i_O LO 1_O LO LO 1 1 t 1 1
1 1 1 1 1 1 1 1 I 1 1 1 1 1 1 1 1 1 ^ 1 1 1 1 1
1 1 I 1 1 1 1 1 t 1 I 1 1 1 *" 1 1 1 I 1 1 S 1 1
, ,CDCDCDcno-1cncncnr^CDo, coco , , , ,CDCD ,CD^
1 i LO LO LO f~-< i . i . i i i co LO i i r-^ r-~- ' CD CD i LO LO , LO co
|2|S iiiiSiiiiiiiiiiiiii
rxj ^xj CXJ
c-"> co co co ro co 'co co co co co co ro co co co co co co co co co co co
CD C^> CD CD CD CP CD CD CD CD CD CD CD CI3 CD CD CD CD CD CD CD CD CZJ CD
" '
CXJ CXI CXI CXJ CXJ CXl Cxi CXJ CXJ fXJ CXJ CXI CXI CXJ CXI Cxj Cx, CXJ CXI CXJ CXJ CXI CXI CXJ
^^^^^^^^0^^^^^^^^^^^^^^^^
xxxxxxxxxxxxxxxxxxxxxxxx
l_O LO LO LO LO LO LO LO LO LO LCI LO LO LO LO LO LO LO LO LO LO LO LO LO
CXI CXI CXI CXI CXI CXJ CXJ CXJ CXJ CXI CXJ CXI CXJ CXl CXI CXI CXI CXI CXJ CXI CXJ CXI CXJ CXJ
co *^r cxi ^r -=a- cxj cxj cxi cxj "^r co co co co co LO 1.0 LO LO co co "=a- *=j- LO
*~*
^^-cS^c^co-^^^^^S^ 8
£
O
cn
c
.Cxjcxicxj^- _ ^. ^^ ^
co cxi co LO cxi co cxi . i jr
E
c
OOLOC^>LOoo^)CZ}Lj->. c r to CZ) ro £j
a-ro-a-^D CNJUJ ro^ Q
q
(/i
o '
£ ?
i i i i i i ^, i i 4. i i i
* 00
O cxj c^
t^l OJ (=
= Li_ M
M => 0
Lf~> ro ro
1 1 1 1 1 1 1 1 1 1 1 i 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1
I i i i """ "^ i I I i i^"^2
L^l^L^,, , ,^,^,^^^^0
i i i i i i i igcSSSgg
i i^SSSSSi
'
E
S22IIIIII1IIII
o
CO
00
fC^J^llllllIIIII ro
ao
CU
3
1 1 1 1 1 1 1 1 1 I 1 1 1 1 ~^
LO to
CXJ ^
1 ' o
CD .c:
X CD
LO L*~*
t i i i i i i i i i i i t i 2 ^
X ^
i co to
i ' O
co a.
cxi cxi cxi cxi cxj cxi cxi cxj cxj cxi cxj cxi cxi cxj QJ a>
CO CO
XX.XXXXXXXXXXXX , £- c
LO LO LO LO LO LO LO LO LO LO LO LO LO LO OJ O
cxj cxj cxj cxi cxj cxi cxj cxj cxi cxi cxj cxj cxj cxi ^ "O
. oo" CO
CXJCX/'C^C^C^^^^^LO'LOLO'LO-LO1 . TO S
------- |