Edited by J.B. Kitto and S.C.

Stultz
 The Babcock & Wilcox Company

Table of Contents
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii to ix
System of Units: English and Système International . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
Editors’ Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Introduction to Steam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intro-1 to 17
Selected Color Plates, Edition: 41 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plates 1 to 8
Section I – Steam Fundamentals
Chapter 1 Steam Generation – An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 to 1-17
2 Thermodynamics of Steam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1 to 2-27
3 Fluid Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 to 3-17
4 Heat Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 to 4-33
5 Boiling Heat Transfer, Two-Phase Flow and Circulation . . . . . . . . . . . . . . 5-1 to 5-21
6 Numerical Modeling for Fluid Flow, Heat Transfer, and Combustion . . . . 6-1 to 6-25
7 Metallurgy, Materials and Mechanical Properties . . . . . . . . . . . . . . . . . . . . 7-1 to 7-25
8 Structural Analysis and Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1 to 8-17
Section II – Steam Generation from Chemical Energy
Chapter 9 Sources of Chemical Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1 to 9-19
10 Principles of Combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1 to 10-31
11 Oil and Gas Utilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-1 to 11-17
12 Solid Fuel Processing and Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-1 to 12-19
13 Coal Pulverization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-1 to 13-15
14 Burners and Combustion Systems for Pulverized Coal . . . . . . . . . . . . . . . 14-1 to 14-21
15 Cyclone Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-1 to 15-13
16 Stokers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-1 to 16-11
17 Fluidized-Bed Combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-1 to 17-15
18 Coal Gasification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-1 to 18-17
19 Boilers, Superheaters and Reheaters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-1 to 19-21
20 Economizers and Air Heaters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-1 to 20-17
21 Fuel Ash Effects on Boiler Design and Operation . . . . . . . . . . . . . . . . . . . . 21-1 to 21-27
22 Performance Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1 to 22-21
23 Boiler Enclosures, Casing and Insulation . . . . . . . . . . . . . . . . . . . . . . . . . . 23-1 to 23-9
24 Boiler Cleaning and Ash Handling Systems . . . . . . . . . . . . . . . . . . . . . . . . 24-1 to 24-21
25 Boiler Auxiliaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-1 to 25-23
Section Ill – Applications of Steam
Chapter 26 Fossil Fuel Boilers for Electric Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26-1 to 26-17
27 Boilers for Industry and Small Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-1 to 27-21
28 Chemical and Heat Recovery in the Paper Industry . . . . . . . . . . . . . . . . . 28-1 to 28-29
29 Waste-to-Energy Installations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29-1 to 29-23
30 Wood and Biomass Installations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30-1 to 30-11
31 Marine Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31-1 to 31-13

vi Steam 41
 The Babcock & Wilcox Company

Chapter 21
Fuel Ash Effects on Boiler
Design and Operation

The effective utilization of fossil fuels for power gen- other heat absorbing surfaces in the convection pass.
eration depends to a great extent on the capability of In extreme cases, uncontrolled ash deposits can de-
the steam generating equipment to accommodate the velop to the point where flow passages in tube banks
inert residuals of combustion, commonly known as ash. are blocked, impeding gas flow and ultimately requir-
The quantity and characteristics of the ash inherent ing the unit to be shut down for manual removal.
to a particular fuel are major concerns to both the de- Large deposits in the upper furnace or radiant super-
signer and the operator of the equipment. heater can become dislodged and fall, damaging pres-
With few exceptions, most commercial fuels contain sure parts in the lower furnace. Under certain condi-
sufficient ash to warrant specific design and operat- tions, ash deposits can also cause fireside corrosion on
ing considerations. The following focuses on these tube surfaces.
design and operating considerations, primarily as they Minimizing the potential for these ash-related prob-
relate to pulverized coal firing. Fuel ash characteris- lems is a primary goal of both the designers and op-
tics relating to petroleum fuels are also discussed. erators of coal-fired boilers. The extent to which coal
Ash dilutes the heating value of fuel, placing addi- ash characteristics affect boiler design is illustrated in
tional burdens on fuel storage, handling and prepa- Fig. 1 which compares the relative size of a gas-fired
ration equipment. Extensive facilities are also needed and coal-fired boiler. Both are sized for the same steam
to collect, remove and dispose of the ash. These mate- generating capacity and similar steam conditions.
rial handling requirements represent significant costs While the combustion characteristics of coal play a role
in terms of equipment and real estate which are di- in sizing the furnace, the deposition and erosion po-
rectly proportional to the amount of ash in the fuel. In tential of the ash are the primary design consider-
the case of coal, ash quantities can be substantial. Con- ations driving the overall size and arrangement.
sider, for example, a 650 MW utility steam generator fir- The variability of ash behavior is one of the biggest
ing a coal with a heating value of 10,000 Btu/lb (23,250 problems for boiler designers and operators. Although
kJ/kg) containing 10% ash by weight. The unit would boilers are often designed to burn a wide range of
burn approximately 300 tons per hour (272 tm/h) of coal, coals satisfactorily, no unit can perform equally well
generating more than 700 tons per day (635 tm/d) of ash. with all types of coal.
In pulverized coal-fired boilers, most of this ash is
carried out of the furnace by the gaseous products of
combustion (flue gas). Abrasive ash particles sus- Ash content of coal
pended in the gas stream can cause erosion problems The ash content of coal varies over a wide range.
on convection pass heating surfaces. However, the This variation occurs not only in coals from different
most significant ash-related problem is deposition. geographical areas or from different seams in the same
During the combustion process, the mineral matter region, but also from different parts of the same mine.
that forms ash is released from the coal at tempera- These variations result primarily from the wide range
tures in the range of 3000F (1649C), well above the of conditions that introduced foreign material during
melting temperature of most mineral matter com- or following the formation of the coal. (See Chapter
pounds. Ash can be released in a molten fluid or sticky 9.) Ash content can also be influenced by extraneous
plastic state. A portion of the ash, which is not cooled mineral matter introduced during the mining opera-
quickly to a dry solid state, impacts on and adheres to tion. Before being sold, some commercial coals are
the furnace walls and other heating surfaces. Because cleaned or washed to remove a portion of what would
such large total quantities of ash are involved, even be labeled ash in the laboratory. However, the ash con-
a small fraction of the total can seriously interfere with tent of significance to the user is the content at the
boiler operation. Accumulation of ash deposits on fur- point of use. The values noted below are on that basis.
nace walls impedes heat transfer, delaying cooling of Most of the coal used for power generation in the
the flue gas and increasing the flue gas temperature United States (U.S.) has an ash content between 6 and
leaving the furnace. Elevated temperatures at the fur- 20%. Low values of 3 to 4% in bituminous coals are
nace exit raise steam temperature and can extend rare and these coals find other commercial uses, par-
deposition problems to pendant superheaters and ticularly in the metallurgical field. On the other hand,

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-1

 The Babcock & Wilcox Company

Fig. 1 Size comparison of gas-fired and coal-fired utility boilers.

some coals may have ash contents as high as 40%. Many bottom arrangement. The coal-fired boiler in Fig. 1 is
high ash fuels can be successfully burned in utility (elec- typical of this design. In a dry-bottom unit most of the
tric power generation) boilers. Their use has increased ash, typically 70 to 80%, is entrained in the flue gas
in areas where they offer an economic advantage. and carried out of the furnace. This portion of the ash
Evaluation of ash content on a weight percentage is commonly known as flyash. Some of the flyash is
basis alone does not take into account the heat input collected in hoppers arranged under the economizer
associated with the coal, which is also related to mois- and air heaters, where coarse particles drop out of sus-
ture content. It is common, for design and fuel evalu- pension when gas flow direction changes. The finer
ation purposes, to consider ash content on the basis ash particles remain in suspension and are carried out
of weight per unit of heat input, generally expressed of the unit for collection by particulate control equip-
as pounds of ash per million Btu. This factor is calcu- ment. (See Chapter 33.) The remaining 20 to 30% of
lated as follows: the ash that settles in the furnace, or is dislodged from
the furnace walls, is collected in a hopper formed by
Ash (% by weight) the frontwall and rearwall tube panels at the bottom
× 104 = lb ash/106 Btu
HHV (Btu/lb) of the furnace. This bottom ash is discharged through
a 3 to 4 ft (0.9 to 1.2 m) wide opening that spans the
or (1)
entire width of the hopper.
Ash (% by weight) Slag-tap furnaces were originally developed to re-
× 103 = kg ash/MJ
HHV (kJ/kg) solve ash deposition and removal problems when fir-
ing coals with low ash fusion temperatures in dry-
where HHV is the higher heating value of the fuel. bottom furnaces. These units are intentionally de-
The relevance of this factor is illustrated in Table 1,
which provides proximate analyses for three selected
coals. Each coal has a moderate ash content of 9 to 10% Table 1
by weight. However, on a heat input basis, ash quanti- Proximate Analyses of Three Selected Coals ⎯
ties vary significantly. The lignite in this example would Ash Content as Weight Per Unit of Heat Input
introduce almost three times as much ash as the high
volatile bituminous coal at an equivalent heat input. High Volatile
Rank Bituminous Subbituminous Lignite

Furnace design for ash removal Moisture, % 3.1 23.8 45.9

Historically, two distinctly different types of furnace Volatile matter, % 42.2 36.9 22.7
Fixed carbon, % 45.4 29.5 21.8
design were used to handle the ash from coal firing
Ash, % 9.4 9.8 9.6
in large utility boilers. These are commonly referred Heating value,
to as the dry-ash or dry-bottom furnace and the slag- Btu/lb 12,770 8683 4469
tap or wet-bottom furnace. lb Ash/106 Btu 7.4 11.3 21.5
All modern pulverized coal-fired boilers use the dry-

21-2 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

signed to maintain ash in a fluid state in the lower A conditioning period is required before significant
furnace. Molten ash is collected on the furnace walls deposition occurs. Assuming there is no direct flame
and other surfaces in the lower furnace and drained impingement, as ash particles approach a clean tube,
continuously to openings called slag-taps in the fur- most tend to be resolidified due to the relatively lower
nace floor. Water tanks positioned beneath the slag- temperature at the tube surface. The particles frac-
taps solidify the liquid ash for disposal. ture on impact and partially disperse back into the flue
Slag-tap furnaces have been used with both pul- gas stream. Over a period of time, however, a base de-
verized coal and CycloneTM furnace firing systems. posit begins to form on the tube. The base deposit may
(See Chapter 15.) Application is limited to coals hav- be initiated by the settling of fine ash particles or the
ing ash viscosity characteristics which would ensure gradual accumulation of particles with very low melt-
that ash fluidity could be maintained over a reason- ing point constituents. As the base deposit thickens,
able boiler load range. Much of the coal ash research the temperature at its outside face increases signifi-
conducted by The Babcock & Wilcox Company (B&W) cantly above the tube surface temperature. Eventu-
concerning the viscosity-temperature relationship of ally, the melting point of more of the ash constituents
coal ash was initially directed at defining coal ash is exceeded and the deposit surface becomes molten.
suitability limits for wet-bottom and Cyclone furnace The process then becomes self-accelerating with the
applications. A minimum coal ash content was also plastic slag trapping essentially all of the impinging
specified to ensure sufficient ash quantities to main- ash particles. Ultimately, the deposit thickness reaches
tain the required slag coating. One benefit of wet- an equilibrium state as the slag begins to flow, or the
bottom firing was a significant reduction in flyash deposit becomes so heavy that it falls away from the
quantity. In pulverized coal wet-bottom applications, tubes. Depending on the strength and physical char-
as much as 50% of the total ash was collected in the acteristics of the deposit, sootblowers using steam,
furnace. Units equipped with Cyclone furnaces could compressed air or water as cleaning media (see Chap-
retain up to 80% of the ash in the furnace. ter 24) may be able to control or remove most of the
The application of slag-tap units for pulverized coal deposit. However, the base deposit can remain at-
firing began to decline in the late 1940s, primarily due tached to the tube, allowing subsequent deposits to
to design improvements in dry-bottom units that mini- accumulate much more rapidly.
mized ash deposition problems. Slag-tap units Fouling is defined as the formation of high tempera-
equipped with Cyclone furnaces continued to be ap- ture bonded deposits on convection heat absorbing
plied until the early 1970s when the federal Clean Air surfaces, such as superheaters and reheaters, that are
Act mandated control of nitrogen oxides (NOx) emissions. not exposed to radiant heat. In general, fouling is
The high furnace temperatures required for wet-bottom caused by the vaporization of volatile inorganic ele-
operation were highly conducive to NOx formation. ments in the coal during combustion. As heat is ab-
sorbed and temperatures are lowered in the convec-
tive section of the boiler, compounds formed by these
Ash deposition elements condense on ash particles and heating sur-
Regardless of the firing method, when coal is burned, face, forming a glue which initiates deposition.
a relatively small portion of the ash will cause deposi- Areas where slagging and fouling can occur are
tion problems. Ash passing through the boiler is subject shown in Fig. 2. Figs. 3 and 4 show heavily slagged
to various chemical reactions and physical forces which and fouled surfaces. The characteristics of coal ash and
lead to deposition on heat absorbing surface. The pro- their influence on slagging and fouling are discussed
cess of deposition and the structure of deposits are vari- in the following sections.
able due to a number of factors. Particle composition,
particle size and shape, particle and surface tempera-
tures, gas velocity, flow pattern and other factors influ- Characteristics of coal ash
ence the extent and nature of ash deposition.
Due primarily to the differences in deposition Sources of coal ash
mechanisms involved, two general types of high tem- Mineral matter is always present in coal and forms
perature ash deposition have been defined as slagging ash when the coal is burned. This mineral matter is
and fouling. usually classified as either inherent or extraneous.
Slagging is the formation of molten, partially fused (See Chapter 9.) Inherent mineral matter is organi-
or resolidified deposits on furnace walls and other sur- cally combined with the coal. This portion came from
faces exposed to radiant heat. Slagging can also ex- the chemical elements existing in the vegetation from
tend into convective surface if gas temperatures are which the coal was formed and from elements chemi-
not sufficiently reduced. cally bonded to the coal during its formation. Extra-
Most ash particles melt or soften at combustion tem- neous mineral matter is material that is foreign to the
peratures. The time-temperature history or cooling organic structure of the coal. This includes airborne
rate of the particle determines its physical state (solid, and waterborne material that settled into the coal
plastic or liquid) at a given location in the furnace. deposit during or after formation. It usually consists
Generally, in order to adhere to a clean surface and of mineral forms associated with clay, slate, shale,
form a deposit, the particle must have a viscosity low sandstone or limestone and includes pieces ranging
enough to wet the surface. from microscopic size to thick layers. Other extraneous
Slag deposits seldom form on clean tube surfaces. material may be introduced through the mining process.

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-3

 The Babcock & Wilcox Company

niques, have identified an enormous variety of min-

eral species in coal, encompassing the entire spectrum
of major mineral forms found in the earth’s crust. Most
of these minerals fall into one of several groups: clay
minerals (aluminosilicates), sulfides/sulfates, carbon-
ates, chlorides, silica/silicates and oxides. Some of the
more common minerals in these groups are shown in
Table 2.
Chemical composition
Because both quantitative and qualitative evalua-
tion of mineral matter forms are extremely difficult,
relatively simple chemical analyses are commonly used
to determine the percentages of the major elements
in the ash. Elemental ash analysis is performed on a
coal ash sample produced in accordance with the
American Society for Testing and Materials (ASTM)
D 3174 ashing procedure. Pulverized coal is burned
in a furnace with an oxidizing atmosphere at 1292 to
1382F (700 to 750C). The elements present in the ash
are quantitatively measured using a combination of
emission spectroscopy and flame photometry and are
reported as weight percents of their oxides. Coal ash
is consistently found to be composed mainly of silicon,
aluminum, iron and calcium with smaller amounts of
magnesium, titanium, sodium and potassium. The
elemental analysis also identifies phosphorus as P2O5
and sulfur as sulfur trioxide (SO3). Phosphorus is usu-
ally present in very small quantities and is sometimes
omitted. Sulfur is reported as SO3 because it is nor-
mally present as the sulfate form of one of the metals.
Fig. 2 Deposition zones in a coal-fired boiler. Percentages of the individual elements vary over a
wide range for different coals; however, characteris-
tic differences are evident between the older, high
Mineralogical composition rank coals common in the Eastern U.S. and the
There are no standardized methods that are used younger, low rank Western coals. Bituminous coals
routinely for determining the specific mineral constitu- typically have higher levels of silica, aluminum and
ents of coal. Mineralogical analysis requires the use iron, while the lower rank subbituminous coals and lig-
of a low temperature ashing technique to separate the nites generally have higher levels of the alkaline earth
mineral matter from the organic portion of the coal. metals, calcium and magnesium, and the alkali metal
Standard high temperature ashing procedures would sodium. These trends are evident in the ash analyses
significantly alter the mineral forms. However, a num- shown in Table 3.
ber of researchers, using a variety of low temperature Although the ash constituents are reported as ox-
ashing methods and sophisticated analytical tech- ides, they actually occur in the ash predominately as

Fig. 3 Heavily slagged surface. Fig. 4 Heavily fouled surface.

21-4 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

ous mineral matter in some of the wide variety of min-

Table 2 eral forms shown on Table 2. When the coal is pul-
Common Minerals Found in Coal verized, some of the particles will be mostly coal with
only inherent mineral matter, some will be pure min-
Clay minerals: eral matter, and others will be combinations of both.
Montmorillonite Al2Si4O10(OH)2 • H2O
Illite KAl2(AlSi3O10)(OH2)
Because the coal particles are burned discretely in sus-
Kaolinite Al4Si4O10(OH)8 pension, the composition of an individual ash particle
will depend on the specific mineral form or forms that
Sulfide minerals: were included in the coal particle. As a result, indi-
Pyrite FeS2 vidual particle composition can vary significantly from
Marcasite FeS2 the bulk ash composition.
Sulfate minerals: During combustion, ash particles are exposed to
Gypsum CaSO4 • 2H2O temperatures as high as 3000F (1649C) and a vari-
Anhydrite CaSO4 ety of heating and cooling rates. The atmosphere in
Jarosite (Na,K)Fe3(SO4)2(OH)6 the burner zone can range from highly oxidizing to
Carbonate minerals: highly reducing. Depending on the composition of the
Calcite CaCO3 specific particle, mineral forms in the ash can react
Dolomite (Ca,Mg)CO3 with each other, with the organic and inorganic con-
Siderite FeCO3 stituents of the coal, and with gaseous elements, such
Ankerite (Ca,Fe,Mg)CO3 as sulfur dioxide (SO2), in the flue gas. The compounds
Chloride minerals: that are ultimately formed by these interactions are
Halite NaCl the materials that cause deposition problems. The com-
Sylvite KCl pounds can have a wide variety of melting tempera-
tures and viscosity-temperature characteristics. Some
Silicate minerals: compounds combine to form eutectic mixtures that
Quartz SiO2
Albite NaAlSi3O8
have melting temperatures lower than either of the
Orthoclase KAlSi3O8 original compounds. Particles that melt at lower tem-
peratures and stay sticky long enough to reach a fur-
Oxide minerals: nace wall become slag deposits. Volatile compounds
Hematite Fe2O3 that vaporize in the furnace tend to condense on and
Magnetite Fe3O4 foul cooler convective heating surfaces.
Rutile TiO2
Elemental ash analyses do not directly identify the
compounds that cause deposition, or directly identify
the mechanisms of deposit formation. Despite these
a mixture of silicates, oxides and sulfates, with smaller limitations, no other data pertaining to coal ash com-
quantities of other compounds. The silicates originate position are as widely available as the chemical analy-
mainly from quartz and the clay minerals which con- ses of ASTM ash. A large part of the coal ash research
tribute silicon, aluminum, sodium and much of the that has been conducted over the last sixty years has
potassium. A principal source of iron oxide is pyrite been directed at correlating analysis data and other
(FeS2) which is oxidized to form Fe2O3 and sulfur ox- characteristics of ASTM ash to observed ash behavior
ides. Part of the organic and pyritic sulfur that is oxi- both in full scale boilers and in test facilities that closely
dized combines with calcium and magnesium to form simulate full scale conditions. Various evaluation
sulfates. Calcium and magnesium oxides result from methods have been developed based on these corre-
the loss of carbon dioxide from carbonate minerals such lations to characterize ash behavior and predict depo-
as calcite (CaCO3) and dolomite [(Ca, Mg)(CO3)]. In sition potential.
low rank coals, a major portion of the sodium, calcium
and magnesium oxides can originate from organically
bound elements in the coal. Ash fusibility
Laboratory ash is prepared from a coal sample in a The measurement of ash fusibility temperatures is
controlled atmosphere at controlled temperatures to by far the most widely used method for predicting ash
provide a reproducible and uniform ash. The actual behavior at elevated temperatures. The preferred pro-
ashing process during combustion in a pulverized coal- cedure in the U.S. is outlined in ASTM Standard D
fired furnace is a much more complex process. In a 1857, Fusibility of Coal and Coke Ash. An ash sample
boiler furnace, pulverized coal is burned in suspen- is prepared by burning coal under oxidizing conditions
sion as discrete particles. If all of the mineral matter at temperatures of 1470 to 1650F (799 to 899C). The
were evenly distributed through the coal, the compo- ash is pressed in a mold to form a triangular pyramid
sition of each resulting ash particle would be the same (cone) 0.75 in. (19 mm) in height with a 0.25 in. (6.35
as the bulk ash composition determined by the analy- mm) triangular base. The cone is heated in a furnace
sis of ASTM ash. A coal with no extraneous mineral at a controlled rate to provide a temperature increase
matter might approach this hypothetical case, because of 15F (8C) per minute. The atmosphere in the fur-
organically combined inherent material would be ex- nace is regulated to provide either oxidizing or reduc-
pected to be evenly distributed. In reality, however, ing conditions. As the sample is heated, the tempera-
all coals contain non-uniformly distributed extrane- tures at which the cone fuses and deforms to specific

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-5

 The Babcock & Wilcox Company

Table 3
Ash Content and Ash Fusion Temperatures of Some U.S. Coals and Lignite
Low Volatile Sub-
Rank: Bituminous High Volatile Bituminous bituminous Lignite
Seam Pocahontas No. 3 No. 9 No.6 Pittsburgh Antelope
Location West Virginia Ohio Illinois West Virginia Utah Wyoming Texas
Ash, dry basis,% 12.3 14.1 17.4 10.9 17.1 6.6 12.8
Sulfur, dry basis, % 0.7 3.3 4.2 3.5 0.8 0.4 1.1
Analysis of ash, % by wt
SiO2 60.0 47.3 47.5 37.6 61.1 28.6 41.8
Al2O3 30.0 23.0 17.9 20.1 21.6 11.7 13.6
TiO2 1.6 1.0 0.8 0.8 1.1 0.9 1.5
Fe2O3 4.0 22.8 20.1 29.3 4.6 6.9 6.6
CaO 0.6 1.3 5.8 4.3 4.6 27.4 17.6
MgO 0.6 0.9 1.0 1.3 1.0 4.5 2.5
Na2O 0.5 0.3 0.4 0.8 1.0 2.7 0.6
K2O 1.5 2.0 1.8 1.6 1.2 0.5 0.1
SO3 1.1 1.2 4.6 4.0 2.9 14.2 14.6
P2O5 0.1 0.2 0.1 0.2 0.4 2.3 0.1
Ash fusibility
Initial deformation temp, F
Reducing 2900 + 2030 2000 2030 2180 2280 1975
Oxidizing 2900 + 2420 2300 2265 2240 2275 2070
Softening temp, F
Reducing 2450 2160 2175 2215 2290 2130
Oxidizing 2605 2430 2385 2300 2285 2190
Hemispherical temp, F
Reducing 2480 2180 2225 2245 2295 2150
Oxidizing 2620 2450 2450 2325 2290 2210
Fluid temp, F
Reducing 2620 2320 2370 2330 2315 2240
Oxidizing 2670 2610 2540 2410 2300 2290

shapes, as shown in Fig. 5, are recorded. Four defor- specified the use of only a reducing atmosphere and
mation temperatures are reported as follows: had loosely defined criteria for identifying the soften-
ing and fluid points. When the atmosphere is not speci-
1. Initial deformation temperature (IT or ID) – the
fied, it is generally assumed to be reducing. Reported
temperature at which the tip of the pyramid be-
softening temperatures are assumed to be the ST (H
gins to fuse or show signs of deformation.
= W) point unless otherwise specified. Methods for de-
2. Softening temperature (ST) – the temperature at
termining fusibility of ash used by other countries are
which the sample has deformed to a spherical shape
similar to the ASTM procedure but results may vary
where the height of the cone is equal to the width at
considerably due to differences in procedures or the
the base (H = W). The softening temperature is com-
definition of terms.
monly referred to as the fusion temperature.
The gradual deformation of the ash cone is gener-
3. Hemispherical temperature (HT) – the tempera-
ally considered to result from differences in melting
ture at which the cone has fused down to a hemi-
characteristics of the various ash constituents. As the
spherical lump and the height equals one half the
temperature of the sample is increased, compounds
width of the base (H = 1/2 W).
with the lowest melting temperatures begin to melt,
4. Fluid temperature (FT) – the temperature at
causing the initial deformation. As the temperature
which the ash cone has melted to a nearly flat layer
continues to increase, more of the compounds melt and
with a maximum height of 0.0625 in. (1.59 mm).
the degree of deformation proceeds to the softening and
The determination of ash fusion temperatures is hemispherical stages. The process continues until the
strictly an empirical procedure, developed in standard-
ized form, which can be duplicated with some degree
of accuracy. Strict observance of test conditions is re-
quired to assure reproducible results. ASTM specified
tolerances on reproducibility of the individual tem-
perature measurements range from 100 to 150F (56
to 83C) when the test is performed by different opera-
tors and apparatus.
An earlier version of the ASTM D 1857 procedure Fig. 5 Specific shapes as ash fuses and deforms with temperature.

21-6 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

temperature is higher than the melting point of most of responding data from other fuels of known full-scale
the ash constituents and the fluid stage is reached. performance. Even comparisons can be misleading,
Fusibility testing was originally developed to evalu- however, when differences in data are within the range
ate the clinkering (agglomerating) tendency of coal ash of reproducibility of the test. Actual ash viscosity mea-
produced by combustion on a grate. In several re- surements (described later) provide a much more accu-
spects, the test method is a somewhat better simula- rate and less subjective definition of the viscosity/tem-
tion of stoker firing than suspension burning of pul- perature relationship and are considered by B&W to
verized coal. During the fusion test, at a heating rate provide a better assessment of slagging potential.
of 15F (8C) per minute, the transition from the IT to
the FT stage may take up to two hours or more for a
high fusion ash. Rather than slow heating and Influence of ash elements
gradual melting of the ash, the process in a pulver-
ized coal furnace is essentially reversed. Ash particles Ash classification
are rapidly heated, and then cooled at a relatively slow Coal ash is classified into two categories based on
rate, as they pass through the furnace. During com- its chemical composition. Lignitic ash is defined as
bustion, coal particles are heated almost instantaneously having more (CaO + MgO) than Fe2O3. Bituminous
to temperatures ranging up to 3000F (1649C). As heat ash is defined as having more Fe2O3 than the sum of
is removed from the flue gas, the ash is cooled over a CaO and MgO. Bituminous ash is generally charac-
period of less than two seconds to temperatures around teristic of higher rank coals from the eastern U.S.
1900 to 2200F (1038 to 1204C) at the furnace exit. Lower rank western coals typically have lignitic ash.
In practical terms, for dry-bottom furnaces, fusion As a result, bituminous ash is sometimes referred to
temperatures provide an indication of the temperature as eastern ash and lignitic ash is sometimes referred
range over which portions of the ash will be in a mol- to as western ash. However, ash classification is not
ten fluid or semi-molten, plastic state. High fusion specific to ASTM rank or geographical origin. In rare
temperatures indicate that ash released in the furnace cases, lignites and subbituminous coals can have bi-
will cool quickly to a nonsticky state resulting in mini- tuminous ash and bituminous coals can have lignitic
mal potential for slagging. Conversely, low fusion tem- ash. For example, the Utah coal shown in Table 3 is
peratures indicate that ash will remain in a molten or classified a bituminous, but has lignitic ash.
plastic state longer, exposing more of the furnace sur-
face or convective surface to potential deposition. Effect of iron
When temperatures in the furnace are below the Iron has a dominating influence on the slagging
measured initial deformation temperature, the major- characteristics of coals with bituminous type ash. As
ity of the ash particles are expected to be in a dry solid shown in Table 2, iron can be present in coal in sev-
state. In this form, particles impacting on heating sur- eral mineral forms. These include pyrite (FeS2), sid-
face will bounce off and be re-entrained in the gas erite (FeCO3), hematite (Fe2O3), magnetite (Fe3O4)
stream, or, at worst, settle on the surface as a dusty and ankerite [(Ca, Fe, Mg)CO3]. Pyrite is the major
deposit which can be readily removed by sootblowers. form of iron in most Eastern coals. In areas of the
At temperatures above the IT, the ash becomes increas- furnace where there is sufficient oxygen, pyrite is
ingly more plastic in nature and impacting particles have converted to Fe 2O3 and SO2. If the local atmosphere
a greater potential to stick to heating surfaces. is reducing, however, pyrrhotite (FeS) is formed
Fusibility temperatures also provide an indication along with the lesser-oxidized iron forms such as
of deposit characteristics as they relate to control and FeO and metallic iron, Fe. The reduced forms have
cleanability. When the temperature at a deposit sur- significantly lower melting temperatures than the
face is at or above the fluid temperature of the ash, oxidized forms. When completely oxidized to Fe2O3
slag will tend to flow or drip from the surface. While iron tends to raise all four values of ash fusion tem-
fluid slag can not be controlled with sootblowers, the peratures: initial deformation, softening, hemi-
deposits tend to be self-limiting in thickness and do spherical and fluid. In the lesser oxidized form (FeO)
not interfere significantly with heat transfer effective- it tends to lower all of these values. The effect of iron
ness. However, if the deposit surface temperature is in each of these forms is indicated in Fig. 6, plotted
in the plastic range, between the initial deformation for a large number of ash samples from U.S. coals.
and hemispherical temperatures, the slag will be too The data show that as the amount of iron in the ash
viscous to flow and will continue to build in thickness. increases, there is a greater difference in ash fus-
Wide IT to HT differentials can result in deposits that ibility between oxidizing and reducing conditions.
build quickly to large proportions and are difficult to con- These effects may be negligible with coal ash con-
trol, because sootblowers can be ineffective in penetrat- taining small amounts of iron. Coals with lignitic ash
ing the plastic shell that forms on the deposit surface. generally have small amounts of iron and the ash fu-
In practice, very high and very low fusion values sion temperatures are affected very little by the state
are relatively easy to interpret as being troublesome of iron oxidation. In fact, lignitic ash containing high
or non-troublesome with respect to slagging. Unfor- levels of calcium and magnesium may have ash fu-
tunately, however, most coals fall in an intermediate sion temperatures that are lower on an oxidizing ba-
range where evaluations can be much more difficult. sis than on a reducing basis. The ash analysis and
Fusion temperatures have their most valid signifi- fusion temperatures shown for the subbituminous coal
cance when used on a comparative basis against cor- in Table 3 illustrate this effect.

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-7

 The Babcock & Wilcox Company

Fig. 6 Influence of iron on coal ash fusion temperatures.

Base to acid ratio However, the presence of basic constituents in an

acidic ash tends to flux or reduce the melting tempera-
The constituents of coal ash can be classified as ei- ture and viscosity of the mixture. Conversely, the
ther basic or acidic. The basic constituents are iron, melting temperature and viscosity of a basic ash are
the alkaline earth metals calcium and magnesium, and reduced by relative proportions of acidic constituents.
the alkali metals sodium and potassium. Acidic con- When the percent base and percent acid are nearly
stituents are silicon, aluminum and titanium. Bases equal, fusion temperatures and ash viscosity tend to
and acids tend to combine to form compounds with be reduced to minimum levels. The general trend is
lower melting temperatures. Experience has shown shown in Fig. 7. Minimum fusion temperatures typi-
that the relative proportions of basic and acidic con- cally occur at approximately 40 to 45% base which
stituents provide an indication of the melting behav- equates to base to acid ratios in the range of 0.7 to 0.8.
ior and viscosity characteristics of coal ash. Ratios in the range of 0.5 to 1.2 are generally consid-
The elemental analysis is used to calculate the per- ered to indicate high slagging potential.
cent base, percent acid and the base to acid ratio as The base to acid ratio considers all of the basic and
follows: acidic constituents to have equal effects on ash melt-
ing characteristics. However, research has shown that
Percent base = the various acids and bases have different fluxing
( Fe2 O3 + CaO + MgO + Na2O + K 2O ) × 100
(2) strengths which must also be considered.
SiO2 + Al2 O3 + TiO2 + Fe2 O3 + CaO + MgO + Na2 O + K 2 O Studies conducted by B&W on the relationship of
ash composition to ash viscosity have provided addi-
Percent acid = tional factors which improve the simple base to acid
relationship. Ash viscosity is an important criterion for
(SiO2 + Al2O3 + TiO2 ) × 100
(3) determining the suitability of a coal ash for use in a
SiO2 + Al2 O3 + TiO2 + Fe2 O3 + CaO + MgO + Na2 O + K 2 O slag-tap furnace. Experience has shown that slag will
flow readily at or below a viscosity of 250 poise. The
Base/acid ratio = temperature at which this viscosity occurs is called the
Fe2 O3 + CaO + MgO + Na2 O + K 2 O
(4) T250 temperature of the ash. The preferred maximum
SiO2 + Al2 O3 + TiO2 T250 for wet-bottom applications is 2450F (1343C).
Trends in T250 temperatures have been shown to cor-
The range of base to acid ratio extends from ap- relate with ash fusion temperatures. Low T250 tempera-
proximately 0.1 for highly acidic ash to 9.0 for ash that tures indicate low fusion temperatures and increased
is high in base content. slagging potential.
Ash that is either highly acidic or highly basic gen- Ash viscosity can be measured directly in a high tem-
erally has high ash fusion and melting temperatures. perature viscometer. Because viscosity measurements

21-8 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

Fig. 7 Fusion temperatures and viscosities versus acidic constituents.

Fig. 8 Plot of temperature for 250 poise viscosity versus base to

require a considerable amount of coal ash that may acid ratio – based on ferric percentage of 20.
not be readily available and are costly and time con-
suming, methods were developed to determine viscos- tend to reduce fusion temperatures and slag viscosity
ity from chemical analysis of the coal ash. Based on a at a given temperature. The degree of iron oxidation
large number of direct viscosity measurements of bi- is normally expressed as the ferric percentage where:
tuminous and lignitic ash samples, T250 temperatures
were related to ash composition as shown in Figs. 8 Ferric percentage =
and 9. Fig. 8 is for bituminous ash and lignitic ash with
an acidic content above 60%. At base to acid ratios less Fe2O3 × 100 (6)
than 0.3, the silicon (SiO2)/aluminum (Al2O3) ratio is Fe2O3 + 1.11FeO + 1.43Fe
taken into account. Silicon and aluminum are both acidic
constituents; however, higher percentages of silicon tend
to raise the T250 and the melting temperature.
Fig. 9 is for lignitic ash with an acidic content less
than 60%. T250 is a function of both the percent base
and the dolomite percentage which is defined as:

Dolomite percentage =
(CaO + MgO ) × 100
(5)
Fe2 O3 + CaO + MgO + Na2O + K 2O

At a given percent base, higher dolomite percent-

ages increase the T250 temperature, indicating that
calcium and magnesium tend to raise ash viscosity and
fusion temperature. Increasing amounts of the other
base constituents (iron, sodium and potassium) tend
to lower the T250 temperature.
Taken together, these trends indicate higher melt-
ing temperatures and higher viscosities at a given tem-
perature for ash that is predominately composed of
either silicon and aluminum or calcium and magne-
sium. Lower melting temperatures result from inter-
mediate mixtures of these elements. However, in all
combinations, iron, sodium and potassium act to flux
the ash and increase the slagging potential.
As previously noted, the fluxing strength of iron is
related to its state of oxidation. Metallic iron (Fe) and Fig. 9 Basic content and dolomite percentage of ash versus temper-
ferrous iron (FeO) are stronger fluxes than Fe2O3 and ature for 250 poise viscosity.

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-9

 The Babcock & Wilcox Company

The effect of ferric percentage on slag viscosity for ing index for bituminous coals which used the total
a typical bituminous ash is shown in Fig. 10. Note that alkali content in the coal to predict fouling potential.
the T250 temperature can vary over a wide range de- Because ASTM ash produced in the laboratory
pending on the degree of iron oxidation. Experience could not be expected to represent the physical and
has shown that slag from boiler furnaces operating chemical properties of flyash produced by full scale
under normal conditions with 15 to 20% excess air has combustion, sintering strength testing required actual
a ferric percentage of approximately 20%. The curves flyash samples aspirated from the flue gas in operat-
in Fig. 8 are based on this value. ing boilers. This meant full scale tests under steady-
state conditions with a consistent coal supply, which
Influence of alkalies on fouling became increasingly more difficult as unit size in-
The alkali metals, sodium and potassium, have long creased. To improve the efficiency and accuracy of ob-
been associated with the fouling tendencies of coal ash. taining data, a small laboratory ashing furnace (LAF)
Volatile forms of these elements are vaporized in the was constructed to burn pulverized coal at controlled
furnace at combustion temperatures. Subsequent re- conditions similar to those in a commercial boiler.
actions with sulfur in the flue gas and other elements Subsequent tests on flyash produced in the LAF
in the ash form compounds that contribute to the forma- from a wide variety of bituminous coals demonstrated
tion of bonded deposits on convection heating surface. that sodium was the most important single factor af-
Research conducted by B&W dating back to the fecting ash fouling. Potassium, which had been in-
1950s identified a relationship between the total al- cluded in the previous alkali fouling indices, was
kali content in bituminous coals and fouling potential. found to make no significant contribution to sinter-
The specific laboratory procedure developed to estab- ing strength. Additionally, it was found that water
lish this relationship, called the sintering strength test, soluble sodium, which was related to the more readily
is described in detail later in this chapter. Basically, vaporized forms of sodium, had a major effect on sin-
the test involves measuring the compressive strength tered strength. This result was obtained by washing
of flyash pellets heated in air for a period of time at coals with hot condensate in the laboratory to remove
temperatures of 1500 to 1800F (816 to 982C). The ap- the water soluble sodium. The washed coals were
plication of this method, combined with observations ashed in the LAF and sintered at various tempera-
of fouling conditions in operating boilers, showed that tures. Results for a high fouling Illinois coal are shown
high fouling coals produced flyash with high sintered in Table 4. Water washing decreased the sodium con-
strength. Conversely, low strength flyash was associ- tent in the ash by approximately 70%, while the po-
ated with low fouling coals. Correlation of standard tassium content, which was initially higher than the
ASTM ash analysis data with the sintering test results sodium content, decreased by only 4%. Removing the
indicated a significant relationship (Fig. 11) between soluble sodium resulted in a reduction in sintering
total alkali content (Na 2O and K 2O, expressed as strength at 1700F (927C) from 17,300 psi (119.3 MPa)
equivalent total Na2O) and flyash sintered strength. for the raw coal to 550 psi (3.8 MPa) for the washed
These correlations formed the basis for the first foul- coal. Because the coal had a high chlorine content, it was
concluded that most of the volatile sodium was probably
in the form of NaCl. The insoluble potassium was likely
associated with clay minerals or feldspar which would
not readily decompose and vaporize during combustion.
The relationship of sintering strength to the per-
centage of soluble sodium in the ash was also found
to be a function of the base to acid ratio, as shown in

Fig. 10 Viscosity-temperature plots of a typical slag showing effect

of ferric percentage. Fig. 11 Effect of alkali content in coal.

21-10 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

As previously noted, in low rank coals, a major por-

Table 4 tion of the alkali and alkaline earth metals can be or-
Effect of Soluble Sodium on Sintered Strength ganically bound in the coal. Because they are inti-
Ash Analysis Raw Coal Washed Coal
mately mixed with the coal, it is believed that alkalies
in this form are readily vaporized during combustion
SiO2 45.0 49.8 and play a dominating role in fouling. The organically
Al2O3 18.0 20.9 associated elements occur in the form of cations chemi-
Fe2O3 21.0 22.9 cally bonded to the organic structure of the coal. Ion
TiO2 0.8 1.0 exchange techniques have been developed to remove
CaO 8.8 1.6 the cations from the coal for measurement. The
MgO 0.9 1.0
Na2O 1.6 0.5
method employed by B&W uses an ammonium acetate
K2O 2.4 2.3 solution to provide a source of NH4+ ions which extract
the ion-exchangeable cations. The laboratory proce-
Ash sintered dure is described later in this chapter. Ion exchange
strength, psi 17,300 550
data for a high fouling North Dakota lignite and a
(MPa) (119.3) (3.8)
severe fouling Montana subbituminous coal are
shown in Table 5. The data show that essentially all
Fig. 12. The combination of high sodium and high base of the sodium in both coals is organically bound. In
to acid ratios resulted in the highest sintering the lignite, the ion-exchangeable sodium actually
strengths. Low ratios and sodium contents resulted in exceeded the total sodium measured in ASTM ash.
reduced flyash strength at the same sintering tem- The difference most likely results from a loss of sodium
perature. Similar trends were noted for variations in due to vaporization during the high temperature
sintering strength as a function of base to acid ratio ashing procedure. The relatively low percentages of
and total Na2O in the ash. Statistical evaluations of ion-exchangeable K2O indicate that most of the potas-
these relationships were used to develop the fouling sium exists in stable mineral forms.
index currently used for coals with bituminous ash.
Similar tests on the sintering characteristics of lig- Viscosity-temperature relationship
nitic ash indicated that the sintering criteria associ-
ated with fouling for bituminous ash did not apply to of coal ash
lignitic ash with high alkaline (CaO, MgO) contents. The characteristics of slag deposits which form on
However, sintering strength was found to be directly furnace walls and other radiant surface are a func-
proportional to the total sodium content in the ash tion of deposit temperature and deposit composition.
shown in Fig. 13. Full scale and pilot scale tests con-
ducted by the U.S. Bureau of Mines at the Grand
Forks Coal Research Laboratory in North Dakota also
established a correlation between fouling rate and so-
dium content for coals with lignitic ash. As shown in
Fig. 14, deposition rates were found to increase sharply
as the Na2O content increased up to approximately 6%
and then level off at higher percentages of sodium.

Fig. 12 Bituminous (Eastern) ash fouling effect of base to acid ratio

and soluble sodium on sintered strength. Fig. 13 Effect of Na2O on sintering strength (North Dakota lignite ash).

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-11

 The Babcock & Wilcox Company

observation led to an extensive study of the relation-

ship between ash viscosity and potential slagging ten-
dency. Viscosity measurements that had previously
been used to determine flow characteristics for wet-
bottom furnace applications were extended to higher
viscosity ranges to define the temperature range
where a given ash would exhibit plastic characteristics.
As liquid ash is cooled, the logarithm of its viscosity
increases linearly with decreasing temperature as
shown in Fig. 15. At some point, the progression devi-
ates from the linear relationship, and viscosity begins
to increase more rapidly as the temperature contin-
ues to decrease. This transition into the plastic region
is caused by the selective separation of solid material
from the liquid, resulting from crystallization of the
higher melting point constituents of the ash. The tem-
perature at which this deviation takes place is called
the temperature of critical viscosity (Tcv). Tcv varies de-
pending on ash composition but normally occurs in a
range between 100 and 500 poise. The end of the plas-
tic region is the point of solidification, or freeze point,
of the slag. The freeze point typically occurs at a vis-
Fig. 14 Effect of Na2O on deposition rate. cosity of approximately 10,000 poise. For convenience
in comparing the viscosity-temperature relationship
Deposit composition, in turn, is a function of the local of various ashes, the viscosity range of 250 to 10,000
atmosphere, particularly for ash with a significant poise has been defined as the plastic region.
iron content. Relationships between these factors de- The temperature at which the plastic region begins
termine the physical state of the deposit, which can and the range of temperature over which the ash is
range from a dry solid to plastic or even a viscous liq- plastic provide an indication of the slagging tendency.
uid if temperatures are sufficiently high. Dry depos- The lower the temperature within this range and the
its are usually not troublesome; they tend to be loosely wider the range, the greater the potential for slagging.
bonded to the tube surface and relatively easy to re- Viscosity-temperature curves for a high slagging Illi-
move by sootblowing. If deposits are allowed to build nois coal and a low slagging east Kentucky coal, shown
in thickness, the temperature increases and the sur- in Fig. 16, illustrate this effect. The plastic range for
face of the deposit can become semi-molten or plastic. the Illinois coal begins at a relatively low temperature
The plastic slag traps other transient ash particles and and extends over a wide temperature range. In con-
continues to build more and more rapidly as the sur- trast, the east Kentucky coal has a very narrow plas-
face temperature continues to increase. Ultimately, the tic range which begins at a much higher temperature.
deposit reaches an equilibrium state as the slag be- In comparison to the Illinois coal, the Kentucky coal
gins to flow. ash would be expected to cool quickly below the tem-
Field experience has shown that plastic slag tends perature where the ash is plastic, exposing much less
to form large deposits that are highly resistant to re- of the furnace to potential deposition.
moval by conventional ash cleaning equipment. This As previously noted, the iron content of coal ash and
its degree of oxidation have a significant influence on
Table 5
Ion Exchange Data ⎯ High and Severe Fouling Coals
Source: North Dakota Montana
Rank: Lignite Subbituminous
Ash, dry basis, % 11.2 5.4
Total alkali, dry
coal basis, % Na2O 4.25 6.74
K2O 0.37 0.65
Ion exchangeable
alkali, dry coal
basis, % Na2O 4.52 6.37
K2O 0.10 0.13
Relative ion
exchange alkali, % Na2O 106% 95%
K2O 27% 20%
Fig. 15 Viscosity increase with decreasing temperature.

21-12 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

the viscosity of the ash. This effect is illustrated in Fig.

17 which shows the viscosity-temperature relationship
for the high iron Illinois coal under both oxidizing and
reducing conditions. Under reducing conditions, the
viscosity at a given temperature is significantly lower
and the ash remains plastic over a much wider tem-
perature range.

Ash reflectivity
Ash from certain coals produces furnace deposits
that have reflective rather than insulating properties.
This is particularly true of low sulfur, low sodium coals
found in the western U.S., from the Powder River Ba-
sin in Wyoming and Montana. Reflective deposits can
significantly reduce furnace heat absorption and in-
crease furnace exit gas temperature even when only
a very thin deposit is present. This can result in ex-
cessive radiant superheater slagging and fouling of
convection surfaces. Experience has shown that re-
flective ash deposits can be difficult to remove and re-
quire special considerations in selection of ash clean-
ing equipment and media. (See Chapter 24.) Propri-
etary methods based upon field experience and labo-
ratory studies are used to evaluate the potential for
reflective ash formation, and to address the impact on Fig. 17 Ash viscosity comparison – oxidizing and reducing conditions.
furnace design and boiler performance.
surface. Deposition characteristics are generally classi-
Ash characterization methods fied into four categories: low, medium, high and severe.
Several slagging and fouling indices have been de- For the most part, the indices described below are
veloped by B&W to provide criteria for various aspects based on readily available ASTM ash analysis and fus-
of boiler design. Slagging indices establish design cri- ibility data. In actual practice, when evaluating coals,
teria for the furnace and other radiant surface while designers take into account full scale experience on
fouling indices establish design criteria for convective similar fuels and results of non-routine testing which
can, in some cases, modify the classification. These in-
dices can also be used on a comparative basis to rank
coals with respect to their slagging and fouling potential
when evaluating a new coal supply for an existing unit.
Ash classification
Because the characteristics of bituminous and lig-
nitic ash vary significantly, the first step in calculat-
ing slagging and fouling indices is the determination
of ash type. In accordance with the criteria previously
described, ash is classified as bituminous when:
Fe2 O3 > CaO + MgO (7)
Ash is classified as lignitic when:
Fe2O3 < CaO + MgO (8)
Slagging index – bituminous ash (Rs) Calculation of
the slagging index (Rs) for bituminous ash takes into
account the base to acid ratio and the weight percent,
on a dry basis, of the sulfur in the coal. The base to
acid ratio indicates the tendency of the ash to form com-
pounds with low melting temperatures. The sulfur con-
tent provides an indication of the amount of iron that
is present as pyrite. The calculation is as follows:
B
Fig. 16 Ash viscosity comparison for a high slagging and low slagging Rs = ×S (9)
coal (oxidizing atmosphere). A

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-13

 The Babcock & Wilcox Company

where Fouling index – bituminous ash (Rf ) The fouling in-

dex for bituminous ash is derived from sintering
B = CaO + MgO + Fe2O3 + Na2O + K2O
strength characteristics using the sodium content of
A = SiO2 + Al2O3 + TiO2
the coal ash and the base to acid ratio as follows:
S = weight % sulfur, on a dry coal basis
Classification of slagging potential using Rs is as follows: B
Rf = × Na2O (12)
A
Rs < 0.6 = low
0.6 < Rs < 2.0 = medium where
2.0 < Rs < 2.6 = high
2.6 < Rs = severe B = CaO + MgO + Fe2O3 + Na2O + K2O
A = SiO2 + Al2O3 + TiO2
Slagging index – lignitic ash (Rs*) The slagging in- Na2O = weight % from analysis of coal ash
dex for lignitic ash (Rs*) is based on ASTM ash fus- Classification of fouling potential using Rf is as follows:
ibility temperatures. As previously noted, fusibility
temperatures indicate the temperature range where Rf < 0.2 = low
plastic slag is likely to exist. The index is a weighted 0.2 < Rf < 0.5 = medium
average of the maximum hemispherical temperature 0.5 < Rf < 1.0 = high
(HT) and the minimum initial deformation tempera- 1.0 < Rf = severe
ture (IT) as follows: Fouling index – lignitic ash The fouling classifica-
tion for lignitic ash coals is based on the sodium con-
Rs * =
( Max HT ) + 4 ( Min IT ) tent in the ash as follows:
(10)
5 When CaO + MgO + Fe2O3 > 20% by weight of coal ash
where Na2O < 3 = low to medium
Max HT = higher of the reducing or oxidizing 3.0 < Na2O < 6 = high
hemispherical softening temperatures, F Na2O > 6 = severe
Min IT = lower of the reducing or oxidizing ini- When CaO + MgO + Fe2O3 < 20% by weight of coal ash
tial deformation temperatures, F
Na2O < 1.2 = low to medium
Classification of slagging potential using Rs* is as follows: 1.2 < Na2O < 3 = high
2450 < Rs* = low Na2O > 3 = severe
2250 < Rs* < 2450 = medium
2100 < Rs* < 2250 = high Coal ash effects on boiler design
Rs* < 2100 = severe
Slagging index – viscosity (Rvs) As previously noted, Furnace design
B&W’s most accurate method for predicting slagging The key to a successful overall gas-side design is
potential is based on the viscosity-temperature rela- proper sizing and arrangement of the furnace. As a
tionship of the coal ash. This index (Rvs) is applicable first priority, the furnace must be designed to mini-
to both bituminous and lignitic ash coals; however, mize slagging and to provide effective control of slag
measured ash viscosities are required. where and when it does form.
Ash deposition in the furnace can cause a number
Rvs =
(T 250 oxid ) − (T10 ,000 red ) of problems. Slag deposits reduce furnace heat absorp-
(11) tion and raise gas temperature levels at the furnace
97.5 ( fs ) exit. This, in turn, can cause slagging and can aggra-
vate fouling in the convection banks where ash de-
where posits become increasingly more difficult to control as
T250 oxid = temperature, F, corresponding to a viscos- gas temperatures increase. The shift in heat absorp-
ity of 250 poise in an oxidizing atmosphere tion from the furnace to the superheater and reheater
T10,000 red = temperature, F, corresponding to a viscosity results in increased attemperator spray flow for con-
of 10,000 poise in a reducing atmosphere trol of steam temperatures, reducing cycle efficiency.
Slag buildup at the top of a tall furnace is dangerous.
and fs is a correlation factor based on the average of Large deposits can become dislodged and fall, caus-
the oxidizing and reducing temperatures (Tfs) corre- ing failures of furnace hopper tubes and loss of avail-
sponding to a viscosity of 2000 poise. Values for fs as ability. Excessive slagging in the lower furnace can
a function of Tfs are provided in Fig. 18. interfere with ash removal.
Classification of slagging potential using Rvs is as follows: Experience has shown that several interrelated fur-
nace design parameters are critical for slagging con-
Rvs < 0.5 = low trol. These parameters focus on keeping ash particles
0.5 < Rvs < 1.0 = medium in suspension and away from furnace surfaces, dis-
1.0 < Rvs < 2.0 = high tributing heat evenly to avoid high localized tempera-
2.0 < Rvs = severe tures, and removing enough heat to achieve tempera-

21-14 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

surface is located in a relatively high gas temperature

zone and subject to ash particle impaction, the side
spacing must be sufficient to limit the potential for
bridging and provide a degree of self-cleaning. Typi-
cal side spacing between platen sections is 4 to 5 ft (1.2
to 1.5 m). When platen superheater surface is used,
the slagging classification of the coal establishes the
upper limit on platen inlet gas temperature, in addi-
tion to limiting the FEGT.
An alternate method of controlling furnace exit gas
temperature that has been widely used is gas temper-
ing by flue gas recirculation. In this method, relatively
cool gas from the economizer outlet is mixed with hot
furnace gas near the furnace exit. Gas tempering of-
fers a number of advantages. The FEGT can be lim-
ited with less furnace surface while the increased gas
weight improves the thermal head for heat transfer,
reducing the surface requirements in the convection
Fig. 18 Slagging index correction factor fs. pass. Proper introduction of the tempering flue gas
provides a flat temperature profile at the furnace exit,
reducing the possibility of localized slagging and foul-
tures at the furnace exit that will minimize deposition ing. Once the choice of gas recirculation is made, the
on convection surface. system can also be used to control reheat steam
In the context of gas-side design, the furnace basi- temperature at partial loads. For this purpose, flue gas
cally serves three functions. It must provide sufficient from the economizer outlet is introduced into the fur-
volume to completely burn the fuel, provide sufficient nace through the furnace hopper opening. The cool
heat transfer surface to cool the flue gas and ash par- gas reduces furnace heat absorption and makes more
ticles to a temperature suitable for admission to the heat available to the reheater which offsets its natu-
convection surface, and minimize the formation of NOx ral characteristic of decreasing outlet steam tempera-
emissions (see Chapter 34). In general, for a coal-fired ture at partial loads.
unit, it is the second criterion that determines the The major disadvantages of gas recirculation are
minimum furnace size. fan maintenance and power requirements. Fan ero-
The slagging classification of the coal establishes sion can be minimized to some extent by proper de-
the upper limit on furnace exit gas temperature sign and operation of a mechanical dust collector ahead
(FEGT) required to minimize the potential for slag- of the fan. Extracting the recirculated gas after a hot
ging both in the radiant superheater and the close- precipitator offers the best potential for a relatively
spaced convection surface. As described in Chapter 22, clean recirculated gas source.
furnace exit gas temperature is a function of furnace In addition to having sufficient volume and heat-
heat release rate. Limiting the FEGT, therefore, lim- ing surface, the furnace also must be correctly propor-
its the heat release rate, resulting in lower average tioned with respect to width, depth and height to mini-
temperatures in the furnace. FEGT limits and corre- mize slagging. A significant design parameter in this
sponding heat release rates have been established by regard is heat input from fuel to the furnace per unit
experience for different types of coal. In general, units of furnace plan area at the burners. Maximum limits
using coals with low or medium slagging tendencies on plan area heat release rate are a function of the
can have higher heat release rates and higher FEGTs. slagging potential of the coal. Limits typically range
Units firing coals with high or severe slagging poten- from 1.5 to 1.8 × 106 Btu/h ft2 (4.7 to 5.7 MWt/m2) for
tial require lower heat release rates and lower FEGTs. severe slagging and low slagging coals respectively.
Ideally, the furnace would be an open box, sized The furnace must also be designed to limit the po-
with sufficient wall surface to cool the furnace gas and tential for ash particle impaction on furnace surfaces.
ash particles to the desired temperature before they Ample clearance must be provided between the burn-
reached any superheater surface. However, thermo- ers and furnace walls as well as the furnace hopper
dynamic considerations in modern high pressure and and arch. These critical dimensions have been estab-
high temperature cycles require that a significant por- lished by operating experience and keyed to the slag-
tion of the total heat absorption be accomplished in ging classification of the coal.
the superheater and reheater. This requirement places The slagging classification also determines the lo-
a practical limit on the amount of furnace wall sur- cations, quantity and spacing of furnace wall blow-
face which, in a drum boiler, is dedicated to generat- ers and long retractable sootblowers in the pendant
ing saturated steam. In order to achieve the required radiant surface. (See Chapter 24.) These allow con-
FEGT it becomes necessary to replace water-cooled trol of the deposition that inevitably occurs and are
furnace wall surface with steam-cooled superheater essential for maintaining furnace surface effectiveness
surface. These surfaces are generally in the form of and furnace exit gas temperature within the range
widely spaced platens (see Chapter 19) located in the provided for in the design. Some degree of slagging
upper radiant zone of the furnace. Because platen may be permitted above the burner zone but only to

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-15

 The Babcock & Wilcox Company

the extent that it can be controlled by selective opera- Convection pass design
tion of the wall blowers. The control of these deposits The key to successfully preparing a design that will
can help maintain steam temperature at reduced loads. control convection pass fouling reverts back to a fur-
Slag deposits on furnace walls must be avoided below nace design that will maintain the furnace exit gas
and between burners, however, where they can not temperature at predicted levels. Temperature excur-
be controlled by sootblowers. sions at the furnace exit result in corresponding
higher temperature levels throughout the convection
Effect of slagging potential on furnace sizing pass which can cause deposition problems even with
Referring to Fig. 19, three large utility boilers are coals which normally would be considered to have a
shown sized for 660 MW at maximum continuous load. low or moderate fouling tendency.
The boilers are assumed to have the same width for In general, convective heating surface, both pen-
purposes of illustration, with the boiler setting height dant and horizontal, is arranged to minimize the po-
and furnace depth varied to accommodate the slag- tential for bridging and obstruction of the gas lanes
ging characteristics of the different fuels. Boiler (a) is between adjacent sections. The minimum clear side
designed to fire a bituminous coal having a low to spacing (measured perpendicular to the gas flow) be-
medium slagging potential. The slightly larger boiler tween sections in a bank varies as a function of the
(b) is designed to fire a subbituminous coal classified average flue gas temperature entering the bank. The
as having a high slagging potential. The difference widest spacing is required in the superheater banks
in size can be attributed primarily to the difference in which are in close proximity to the furnace exit, where
slagging potential. The furnace (b) depth has been the gas temperature and fouling potential are high.
increased to control slagging by reducing the input per As the flue gas temperature is reduced, the side spac-
plan area. The input and gas weight are higher for ing in succeeding banks can also be reduced. The spe-
the subbituminous coal due to its higher moisture con- cific side space dimensions at a given temperature
tent and resulting lower boiler efficiency. This increases entering the bank depend on the fouling classification
the required furnace surface and the furnace exit area of the coal. Severe fouling coals require the widest
to maintain acceptable gas velocities entering the con- spacing. Adequate side spacing must be maintained
vection pass. Comparing boiler (c), firing a severe slag- even in low temperature horizontal banks such as
ging lignite, to boiler (b), the furnace depth has again economizers. While these surfaces are not normally
been increased due to the increased slagging poten- subject to bonded deposits, sufficient clear space must
tial. The furnace surface has also been increased to be maintained between sections to ensure that accu-
reduce the gas temperature leaving the furnace. The size mulations of ash dislodged from upstream surfaces will
differential of the three units is quantified in Table 6. not bridge and plug the gas lanes. (See Chapter 20.)
This table shows the proportionate differences or in- Bank depths (measured parallel to the direction of
creases using boiler (a) as a base. Boiler (a) is assigned gas flow) are established as a function of fouling po-
a size factor of 1.0 for the various parameters shown. tential, clear side spacing and the temperature enter-

Fig. 19 Influence of slagging potential on furnace size. (See Table 6.)

21-16 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

Secondary air imbalances can be minimized by ad-

justing individual burner flows to provide a flat O2 pro-
Table 6 file at the economizer outlet. Care must be exercised
Boiler Size Versus Slagging Classification
to avoid burner adjustments that cause flame im-
Boiler pingement on furnace walls. On the fuel side, burner
line resistances should be balanced to maintain uni-
(a) (b) (c) form coal flow to each burner. Coal feeders should be
Coal Rank Bituminous Subbituminous Lignite calibrated and adjusted to provide uniform coal flow
to each pulverizer.
Slagging Low/Med High Severe Low pulverizer fineness (see Chapter 13) can also
Furnace plan cause problems associated with delayed combustion.
area 1.0 1.11 1.24 Coarse particles require longer residence times for
burnout and can cause slagging in the lower furnace.
Furnace surface 1.0 1.18 1.50 Excess air has a tempering effect on average tem-
No. of furnce peratures within the furnace and on furnace exit tem-
wall blowers 30 36 70 perature. Excess air also reduces the potential for lo-
calized reducing conditions in the furnace when it is
introduced through the burners. Air infiltration into
the furnace or convection pass is far less beneficial and
ing the bank. Cavities between the banks provide lo- should be corrected or taken into account when estab-
cations for long retractable sootblowers. At high gas lishing excess air requirements. While there is an as-
temperatures, shallow bank depths are required to sociated efficiency loss, raising excess air above nor-
ensure adequate sootblower effectiveness. Sootblower mal design levels is usually an effective tool for con-
jet penetration increases as temperatures are reduced trolling deposition problems. In some cases, high ex-
and bank depths can be increased incrementally in cess air may also upset superheater/reheater absorp-
cooler areas. tion and steam temperatures.
Sootblowers (see Chapter 24) are the primary
Flyash erosion means of dealing directly with furnace wall slagging
The metal loss on convection pass tubes due to flyash and convection pass fouling. The most important fun-
erosion is proportional to the total ash quantity pass- damental requirement is to use this equipment in a
ing through the boiler and is an exponential function preventive, rather than corrective, manner.
of flue gas velocity. While with a given fuel there is Sootblowers are most effective in controlling dry,
no control of the ash quantity, erosion problems can loosely bonded deposits which typically occur in the
be eased by reducing flue gas velocities. Velocity lim- early stages of deposition. If furnace slag is allowed
its are determined based on the ash quantity on a to accumulate to the point that it becomes plastic or
pounds per million Btu (kg/MWt) basis and the rela- wet, or if convection pass deposits are allowed to build
tive proportion of abrasive constituents in the ash. and sinter for long periods of time, removal becomes
Typical limits range from 65 ft/s (19.8 m/s) for relatively much more difficult. Sootblower sequencing require-
non-abrasive low ash coals to 45 ft/s (13.7 m/s) or less ments must be established by initial operating expe-
for coals with high ash quantities and/or abrasive ash. rience and updated when required, especially when
fuel characteristics change. Boiler diagnostic systems,
which are discussed in the following section, can as-
Effect of operating variables sist in optimizing sootblower operation.
Although the predominant factors affecting depo- The least desirable operating technique for control-
sition are ash characteristics and boiler design, oper- ling deposition problems is load reduction. The most
ating variables can also have a significant impact on severe situations may require a permanent derate.
slagging and fouling. However, in many marginal situations, temporary
In general, operating variables associated with load reductions during off peak periods may provide
combustion optimization (see Chapter 14) tend to re- sufficient cooling to shed slag and allow sootblowers
duce the potential for deposition problems. These vari- to regain effectiveness.
ables include air distribution, fuel distribution, coal
fineness and excess air. Application of advanced diagnostic and
Air and fuel imbalances can result in high excess
air at some burners while others operate with less than control systems
theoretical air. This, in turn, results in localized reduc- Awareness of slagging and fouling conditions is
ing conditions in the burner zone which can aggra- critical to achieving reliability and availability on a
vate slagging, especially with coals having high iron coal-fired utility boiler. However, boiler surface clean-
content. High coal/air ratios can also delay combus- liness has been, traditionally, one of the most difficult
tion and upset heat distribution, resulting in elevated operating variables to quantify. Typical indications of
temperatures in the upper furnace and at the furnace surface fouling appear to the operator indirectly in the
exit. Long burnout times also increase the potential form of steam temperatures, spray attemperation
for burning particles to contact furnace walls and flows and draft losses (gas resistance). In some cases,
other heat transfer surfaces. experienced operators who are familiar with the op-

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-17

 The Babcock & Wilcox Company

erating characteristics of a unit can make judgments Additional discussion of the application of control
on slagging and fouling conditions based on operat- systems to local sootblower cleaning requirements is
ing conditions, but these secondary indications can be provided in Chapter 24.
misleading. For example, the furnace can be slagged,
causing undesirably high gas temperatures entering
the convection surface. However, the steam tempera- Non-routine ash evaluation methods
tures and spray attemperation may be normal if the The following describes the laboratory equipment
convection surfaces are also fouled. and test procedures, referenced earlier, that are used
Another indication of surface cleanliness is draft to supplement the standard ASTM coal ash charac-
loss. By watching draft loss across a bank, an alert terization methods.
operator can determine that sootblowing is probably
required. Usually, however, by the time a change in Laboratory ashing furnace
draft loss is detected across widely spaced pendant As noted, the ASTM ashing procedure does not
sections, the banks are already bridged and it may be duplicate the ashing process that actually occurs in a
too late for removal by the sootblowers. boiler. A laboratory ashing furnace (LAF) provides a
Visual observation is frequently used to further means to obtain flyash and deposit samples that are
quantify cleanliness conditions. In many instances, comparable to those obtained from full scale installa-
however, access is limited and subjective evaluations tions operating under similar conditions.
can leave considerable room for error. Advanced meth- B&W’s LAF, shown in Fig. 20, is designed to fire
ods have been developed to overcome these shortcom- pulverized coal at rates typically between 5 and 10 lb/
ings and to improve upon traditional time-based h (2.3 and 4.5 kg/h). The facility consists of a fuel feed
sootblowing control. system, pulverized coal burner and a refractory lined
Computer based performance monitoring systems chamber. The combustion chamber is surrounded by
can provide a direct and quantitative assessment of an electrically heated guard furnace which controls
furnace and convective surface cleanliness. B&W’s the rate of heat removal from the chamber to simu-
Heat Transfer Manager™ (HTM) program is based on late full scale furnace temperatures. The firing rate
the heat transfer analysis program developed over is established to approximate full scale furnace resi-
many years for boiler design and validated by exten- dence time. A deposition section located at the furnace
sive empirical data. The HTM program is configured exit contains air- or water-cooled probes. The surface
on a boiler-specific basis, taking into account the ar- temperature of the probes can be adjusted to simulate
rangement of the furnace and all convective surface. furnace and superheater tube operating temperatures.
Measurements of temperatures, pressures, flows, and
gas analysis data are used to perform heat transfer
analysis in the furnace and convective section on a
bank by bank basis.
Advanced intelligent sootblowing systems have also
been developed to combine this real time assessment
of furnace and convective surface cleanliness with
closed loop control of the cleaning equipment. B&W’s
Powerclean™ system automatically determines where
and when sootblowing should occur in the furnace and
convection pass. Powerclean uses cleanliness data
from the HTM program in an expert decision making
structure that dictates when blowers should be cycled.
Intelligent systems such as Powerclean recognize
problem areas early in their development, so that se-
lective sootblowing can be directed at a specific prob-
lem area and ash cleaning equipment is operated
based on need. Intelligent sootblowing systems can
optimize blowing medium use and improve perfor-
mance while reducing tube damage and providing
consistency to boiler operations.
Slagging can be particularly troublesome in local-
ized areas of the furnace. To help optimize wall clean-
ing, heat flux sensors can be used. These sensors are
installed in the waterwalls of the furnace and provide
a differential temperature across the wall which
changes in proportion to the amount of deposition.
Sensor data is integrated into the overall intelligent
sootblowing system so that cleanliness can be optimized
in the furnace region. If an array of sensors is installed,
the furnace can be broken into regions for better control
of wall cleaning equipment and to optimize operation. Fig. 20 Schematic of laboratory ashing furnace (LAF).

21-18 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

The probes are instrumented to allow measurement

of metal temperatures, cooling fluid flow rates, and
cooling fluid inlet and outlet temperatures. These data
permit calculation of the total heat flux from the flue
gas through the deposit and into the probe. The depo-
sition section is also fitted with sootblowers to evalu-
ate the effectiveness of ash removal equipment. Fig.
21 shows ash particles impacting a simulated super-
heater tube during a deposition test. The B&W LAF
was used to develop extensive data which has now
been correlated to the ash characteristics. These cor-
relations have largely replaced the need for routine
laboratory testing.
Measurement of ash viscosity
Viscosity of coal ash is measured in a high tempera-
ture rotating-bob viscometer (Fig. 22). The ash under
study is contained in a cylindrical platinum-rhodium
crucible, and a cylindrical bob is rotated in the liquid
at a constant speed through a calibrated suspension
wire. The torque or amount of twist produced in the
suspension wire is proportional to the viscosity. The
amount of twist is measured and recorded as the in-
terval between impulses from light beams reflected
from mirrors attached to the ends of the wire. The
suspension wires are calibrated against viscosity stan-
dard oils obtained from the Bureau of Standards. Fig. 22 Section through furnace of high temperature viscometer.
The electrically heated furnace is of the Globar tube
type with temperature regulation provided through
a controlling type potentiometer actuated by a ther- mined steps and the viscosity of the ash is measured
mocouple located in the furnace adjacent to the sample at each temperature.
crucible. A thermocouple imbedded in the ash crucible
support indicates sample temperature. Provision is Ash sintering strength
made for controlling the atmosphere within the fur- The sintering strength test is performed on a flyash
nace. Ash is introduced into the crucible at an elevated sample prepared in the LAF under a standard set of
temperature [2600 to 2800F (1427 to 1538C)] and firing conditions. The flyash is passed through a 60
held at that temperature until it becomes uniformly mesh (U.S. standard) (250 micron) screen to remove
fluid. The temperature is then decreased in predeter- any particles of slag and then ignited to constant
weight at 900F (482C) to remove any carbon that
might be present. The ignited ash is then reduced to
a minus 100 mesh size and at least 24 cylindrical speci-
mens [0.6 in. (15.2 mm) diameter by 0.85 in. (21.6 mm)
long] are formed in a hand press at a pressure of 150 psi
(1034 kPa). At least six specimens are heated in air, usu-
ally at each of four temperature levels [1500, 1600, 1700
and 1800F (816, 871, 927 and 982C)] for 15 hours.
After the specimens have cooled slowly in the fur-
nace, they are removed, measured and then crushed
in a standard metallurgical testing machine. The sin-
tered or compression strength is then computed from
the applied force and the cross-sectional area of the
sintered specimen. The average strength of six speci-
mens is used as the strength of the sintered flyash at
a particular sintering temperature.
Measurement of ion exchangeable cations in coal
Twenty grams of an air-dried minus 60 mesh coal
sample are mixed with 100 ml of 1 N ammonium ac-
etate in a 300 ml three-neck round bottom flask. A
thermometer is inserted into the slurry. The slurry is
stirred constantly and heated to 60 ±5C. The coal
slurry sample is refluxed for 18 hours. The sample is
Fig. 21 Deposit formation on simulated superheater tube. filtered through a cellulosic filter media with 0.45 μ

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-19

 The Babcock & Wilcox Company

average pore size and washed twice with 25 ml of 1 N heaters and reheaters of several boilers burning low
ammonium acetate solution. to medium alkali coals. Where there was no corrosion,
The above procedure is repeated on the filtered coal the complex sulfates were either absent or the tube
except that the time is shortened to three hours. The metal temperatures were moderate [less than 1100F
combined filtrates are acidified by adding 2% by vol- (593C)]. The general conclusions drawn from this
ume of glacial acetic acid and stored for inductive survey of corrosion were:
coupled plasma atomic emission spectrometric
1. All bituminous coals contain enough sulfur and
(ICPAES) analysis of Na, K, Ca and Mg.
alkali metals to produce corrosive ash deposits on
superheaters and reheaters, and those containing
Coal ash corrosion more than 3.5% sulfur and 0.25% chlorine may
be particularly troublesome.
Serious external wastage or corrosion of high tem-
2. Experience has shown that corrosion rate is af-
perature superheater and reheater tubes was first
fected by both tube metal temperature and gas
encountered in coal-fired boilers in 1955. Tube fail-
temperature. Fig. 24, which is used as a guide in
ures resulting from excessive thinning of the tube
design, indicates stable and corrosive zones of fuel ash
walls, as shown in Fig. 23, occurred almost simulta-
corrosion as a function of gas and metal temperatures.
neously in the reheater of a dry ash furnace boiler and
the secondary superheater of a slag-tap furnace unit. Based on this information, B&W modified the de-
Corrosion was confined to the outlet tube sections of sign of its boilers to greatly reduce the corrosion of
the reheater and the secondary superheater, which superheaters and reheaters. These modifications in-
were made from chromium ferritic and stainless steel cluded changes in furnace geometry, burner configu-
alloys, respectively. ration, superheater arrangement and the use of gas
Significantly, these boilers were among the first to tempering, all of which reduced metal and gas tem-
be designed for 1050F (566C) main and reheat steam peratures and reduced temperature imbalances. Ex-
temperatures; also, both units burned high sulfur, perience from these installations has shown that it is
high alkali Central and Southern Illinois coals, which possible to operate boilers with main and reheat steam
were causing chronic ash fouling problems at the time. temperatures up to 1050F (566C) with little, if any,
Early investigations showed that corrosion occurred corrosion from most coals.
where complex alkali sulfates concentrated on tube Meanwhile, there was a gradual return to the
surfaces beneath bulky layers of ash and slag. When 1000F (538C) steam conditions, due primarily to eco-
dry, the complex sulfates were relatively innocuous; but nomic factors and secondarily to coal ash corrosion.
when semi-molten [1100 to 1350F (593 to 732C)], they This temperature level has permitted the use of lower
corroded the alloy steels used in superheater construc- cost alloys in the boiler, steam piping and turbine, with
tion, and also other normally corrosion resistant alloys. substantial savings in investment costs; it also has
At first, it appeared that coal ash corrosion might provided a greater margin of safety to avoid corrosion.
be confined to boilers burning high alkali coals, but Steam temperatures remained on the 1000F (538C)
complex sulfate corrosion was soon found on super- plateau for several decades. However, improved alloy
creep and corrosion resistance and the potential for

Fig. 23 Typical corroded 18Cr-8Ni tube from secondary superheater. Fig. 24 Coal ash corrosion – stable and corrosive zones.

21-20 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

cycle efficiency improvement have led to increased

steam temperatures in the newest generation of util-
ity boilers. Some recent boilers have been designed
with main and reheat steam temperatures at and above
1100F (593C), and temperatures of 1200F (640C) and
above are envisioned.
General characteristics of corrosion
External corrosion of superheaters and reheaters
is concentrated on the upstream side of the tube, as
shown in Fig. 25. The greatest metal loss usually oc-
curs on the 10 and 2 o’clock sectors of the tubes, and
it tapers off to little or none on the back side of the
tubes. The corroded surface of the tube is highly sculp-
tured by a shallow macropitting type of attack. The
amount of corrosion, as measured by reduction in tube
wall thickness, varies considerably along the length
of the tube, depending on local conditions, i.e., the
position of the tube in the bank or platen, the prox-
imity of sootblowers, the composition of ash deposits and,
most importantly, the gas and metal temperatures. Fig. 26 Effect of temperature on corrosion rate.
The corrosion rate is a nonlinear function of metal
temperature (Fig. 26). Typically, the corrosion of both
chromium ferritic and 18Cr-8Ni stainless steels in- to that of flyash. Though often hard and brittle, this
creases sharply above a temperature of 1150F (621C), layer is a porous structure through which gases may
passes through a broad maximum between 1250 and diffuse. Innocuous by itself, it plays an important part
1350F (677 and 732C) and then decreases rapidly at in the formation of an intermediate layer that contains
still higher temperatures. However, the corrosion be- the corrosive agents.
havior varies depending on gas and ash chemistries The intermediate layer, frequently called the white
and other factors, as well as temperatures. layer, is a white to yellow colored material which var-
The highest corrosion rates are generally found on ies in thickness from 0.03 to 0.25 in. (0.76 to 6.35 mm).
the outlet tubes of radiant superheater or reheater It usually has a chalky texture where corrosion is mild
platens opposite retractable sootblowers. Values rang- or nonexistent, but is fused and semi-glossy where cor-
ing from 50 to 250 mils/yr (1.27 to 6.35 mm/yr) have rosion is severe. In the latter condition this layer is dif-
been observed on 18Cr-8Ni stainless steel tubes un- ficult to remove as it is so firmly bonded to the corroded
der these adverse conditions. When similar high tem- surface beneath.
perature surfaces [1100 to 1175F (593 to 635C)] are
arranged in convection tube banks so they are
shielded from direct furnace radiation and sootblower
action, corrosion rates are much lower, ranging be-
tween 5 and 20 mils/yr (0.13 to 0.51 mm/yr).
Corrosive ash deposits
Corrosion is rarely found on superheater or reheater
tubes having only dusty deposits. It is nearly always
associated with sintered or slag type deposits that are
strongly bonded to the tubes. Such deposits consist of
at least three distinct layers. The outer layer, shown
diagrammatically in Fig. 27, constitutes the bulk of
the deposit and has an elemental composition similar

Fig. 25 Transverse sections of corroded tubes from secondary

superheater platens. Fig. 27 Analyses of typical ash deposit from 18Cr-8Ni superheater tube.

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-21

 The Babcock & Wilcox Company

Upon heating in the air, the intermediate layer and

melts around 1000F (538C) and slowly discolors and
hardens into a hard mass resembling rust. Chemical K 2 SO4 + Al 2 O3 + 3SO3 → 2KAl ( SO4 )2 (13)
analyses of this layer show that it contains higher con-
centrations of potassium, sodium and sulfur than does Similar reactions occur with sodium sulfate (Na2SO4),
the parent coal ash. A large part of this deposit is water although the complex sodium sulfates are less apt to form
soluble and the water soluble fraction is always acidic. at high temperatures because of their lower stability.
The most common compounds found are Na3Fe(SO4)3 Work at B&W has shown that SO3 concentrations
and KAl(SO4)2. in ash deposits must be very high (1000 to 1500 ppm)
Complex alkali sulfates, when molten, rapidly cor- compared to the level in the flue gas (10 to 25 ppm) in
rode most, if not all, superheater alloys. Corrosion order to form the complex alkali sulfates in the inter-
begins between 1000 and 1150F (538 and 621C), de- mediate layer. Therefore, the bulk of the SO3 must
pending on the relative amounts of complex sodium come from the catalytic oxidation of SO2 in the outer
and potassium sulfates present and whether these are layer of the deposit.
predominantly iron or aluminum base compounds. When the SO3 produced in the outer deposit exceeds
Corrosion usually begins at the lower temperature the partial pressure of SO3 necessary for stability, the
where the sodium-iron-sulfate system is the major part complex sulfates form through the above reactions.
of the intermediate layer, but corrosion is more severe When the opposite is true, the complex sulfates begin
and persists into a higher temperature range when to decompose according to the reverse of these reac-
the potassium-aluminum-sulfate system dominates. tions until a new equilibrium is reached. Because the
If the intermediate layer is carefully removed, a formation of SO 3 is temperature dependent, the
black, glassy inner layer is revealed, which appears reversibility of these reactions is also temperature
to have replaced the normally protective oxide on the dependent. As shown in Fig. 26 the corrosion rate in-
tube. This layer is composed primarily of corrosion creases with temperature, passes through a maximum
products, i.e., oxides, sulfides, and sulfates of iron and between 1250 and 1350F (677 and 732C), and then falls
other alloying constituents in the tube metal. It sel- to a comparatively low level at higher temperatures.
dom exceeds 0.063 in. (1.59 mm) thickness on corroded The temperature range of this rapid liquid-phase
18Cr-8Ni stainless steel tubes, probably because of its attack is bracketed by: 1) the melting temperature of
strong tendency to spall when the tubes cool. The layer the mixture of complex alkali sulfates present, and 2)
containing corrosion products from chromium ferritic their thermal stability limits. The extreme width of this
steels often reaches 0.125 in. (3.18 mm) thickness and temperature band is approximately 400F (222C); cor-
exhibits little tendency to spall as the tube cools. rosion due to the complex alkali sulfates may range
from as low as 1000F (538C) to a maximum of 1400F
Corrosion mechanisms (760C), depending on the species present in the in-
termediate layer.
The elements in coal ash corrosion (sodium, potas-
sium, aluminum, sulfur and iron) are derived from the Corrective measures
mineral matter in coal. The minerals supplying these Various methods of combating corrosion of super-
elements include shales, clays and pyrites which are heater and reheater tubes have been used or sug-
commonly found in all coals. gested, including the following:
During the combustion of coal, these minerals are 1. the use of stainless steel shields to protect the most
exposed to high temperatures and strongly reducing vulnerable tubes,
effects of carbon for very short periods of time. Al- 2. coal selection,
though comparatively stable, the mineral matter un- 3. improvement of combustion conditions, i.e., pro-
dergoes rapid decomposition under these conditions. viding proper coal fineness, fast ignition, good mix-
Some of the alkalies are released or volatilized as rela- ing and proper excess air, and
tively simple compounds, which have dew points in 4. the use of more corrosion resistant alloys and al-
the 1000 to 1300F (538 to 704C) range. Furthermore, loy cladding on the most vulnerable superheater
sulfur in the coal is oxidized, releasing SO2 with the and reheater tubes. (See Chapter 19.)
formation of a small amount of SO3, leaving a residue
of iron oxide (Fe2O3). New high chromium content stainless steels are
By far the largest portion of the mineral matter or generally more resistant to coal ash corrosion, but
its derived species react to form the glassy particulates added chromium tends to degrade other metallurgi-
of flyash. The flyash and volatile species in the flue cal properties. New modified 25% chromium 20%
gases deposit on the tube surfaces. Slowly, over a pe- nickel stainless steels offer improved performance for
riod of weeks, the alkalis and the sulfur oxides diffuse moderate conditions. Higher chromium nickel-based
through the layer of flyash toward the tube surface. alloys and cladding offer maximum resistance for the
In the lower temperature zone of the ash deposit, most severe conditions.
chemical reactions between the alkalis, the sulfur ox-
ides, and the iron and aluminum components of the Fuel oil ash
flyash form complex alkali sulfates as follows: The ash content of residual fuel oil seldom exceeds
0.2%, an exceedingly small amount compared to that
3K 2 SO4 + Fe2 O3 + 3SO3 → 2K 3 Fe ( SO4 )3 in coal. Nevertheless, even this small quantity of ash

21-22 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

is capable of causing severe problems of deposition and residual fuel oil or distillate residue. Virtually all me-
corrosion in boilers. Of the many elements that may tallic compounds and a large part of the sulfur com-
appear in oil ash deposits, the most important are va- pounds are concentrated in the distillation residue.
nadium, sodium and sulfur. Compounds of these ele- Where low sulfur residual fuel oils are required, they
ments are found in almost every deposit in boilers fired are obtained by blending with suitable stocks, includ-
by residual fuel oil and often constitute the major por- ing both heavy distillates and distillation from low
tion of these deposits. sulfur crudes. This procedure is also used occasionally
if a residual fuel oil must meet specifications such as
Origin of ash vanadium or ash content.
As with coal, some of the ash-forming constituents
in the crude oil had their origin in animal and veg- Release of ash during combustion
etable matter from which the oil was derived. The Residual fuel oil is preheated and atomized to pro-
remainder is extraneous material, resulting from con- vide enough reactive surface so that it will burn com-
tact of the crude oil with rock structures and salt pletely within the boiler furnace. (See Chapter 11.)
brines, or is picked up during refining processes, stor- The atomized fuel oil burns in two stages. In the first
age and transportation. (See Chapter 9.) stage the volatile portion burns and leaves a porous
In general, the ash content increases with increas- coke residue and in the second stage the coke residue
ing asphaltic constituents in which the sulfur acts burns. In general, the rate of combustion of the coke
largely as a bridge between aromatic rings. Elemen- residue is inversely proportional to the square of its
tal sulfur and hydrogen sulfide have been identified diameter, which in turn is related to the droplet diam-
in crude oil, and simpler sulfur compounds are found eter. Therefore, small fuel droplets give rise to coke
in the distillates of crude oil including thioesters, dis- residues that burn very rapidly, and the ash forming
ulfides, thiophenes and mercaptans. constituents are exposed to the highest temperatures
Vanadium, iron, sodium, nickel and calcium in the in the flame envelope. The ash forming constituents
fuel oil were probably derived from the rock strata but in the larger coke residues from the larger fuel drop-
some elements, such as vanadium, nickel, zinc and lets are heated more slowly, partly in association with
copper, probably came from organic matter from which carbon. Release of the ash from these residues is de-
the petroleum was derived. Vanadium and nickel es- termined by the rate of oxidation of the carbon.
pecially are known to be present in organo-metallic During combustion, the organic vanadium com-
compounds known as porphyrins which are charac- pounds in the residual fuel oil thermally decompose
teristic of certain forms of animal life. Table 7 indicates and oxidize in the gas stream to V2O3, V2O4 and finally
the amounts of vanadium, nickel and sodium present V2O5. Although complete oxidation may not occur and
in residual fuel oils from various crudes. there may be some dissociation, a large part of the
Crude oil as such is not normally used as a fuel but vanadium originally present in the oil exists as vapor
is further processed to yield a wide range of more valu- phase V 2O 5 in the flue gas. The sodium, usually
able products. For example, in a modern U.S. refin- present as chloride in the oil, vaporizes and reacts
ery, 92.4% of the crude is converted to lighter fraction with sulfur oxides either in the gas stream or after
fuels and products such as gasoline, leaving 7.6% of deposition on tube surfaces.
Subsequently, reactions take place between the va-
nadium and sodium compounds, with the formation
Table 7 of complex vanadates having melting points lower
Vanadium, Nickel and Sodium than those of the parent compounds; for example:
Content of Residual Fuel Oils
(ppm by wt)
Na2 SO4 + V2O5 → 2Na VO3 + SO3 ↑
(14)
Source of Melting points: 1625F 1275F 1165F
Crude Oil Vanadium Nickel Sodium
Excess sodium or vanadium in the ash deposit, above
Africa: that necessary for the formation of the sodium vana-
1 5.5 5 22 dates (or vanadylvanadates), may be present as
2 1 5 − Na2SO4 and V2O5, respectively.
Middle East: The sulfur in residual fuel oil is progressively re-
3 7 − 1 leased during combustion and is promptly oxidized to
4 173 51 −
5 47 10 8
SO2. A small amount of SO2 is further oxidized to SO3
U.S.: by a small amount of atomic oxygen present in the
6 13 − 350 hottest part of the flame. Also, catalytic oxidation of
7 6 2.5 120 SO2 to SO3 may occur as the flue gases pass over va-
8 11 − 84 nadium rich ash deposits on high temperature super-
Venezuela: heater tubes and refractories. (See Chapter 35.)
9 − 6 480
10 57 13 72 Oil slag formation and deposits
11 380 60 70
The deposition of oil ash constituents on the furnace
12 113 21 49
13 93 − 38 walls and superheater surfaces can be a serious prob-
lem. This deposition, coupled with corrosion of super-

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-23

 The Babcock & Wilcox Company

heater and reheater tubes by deposits, was largely stituents, 3) melting and freezing temperatures of the
responsible for the break in the trend towards higher ash, and 4) the total sulfur content of the oil. Apply-
steam temperatures that occurred in the early 1960s. ing this information in boiler design is largely a mat-
Practically all boiler installations are typically de- ter of experience.
signed for steam temperatures in the 1000 to 1015F
(538 to 546C) range to minimize those problems and Boiler design
to avoid the higher capital costs of the more expen- Generally speaking, progressive fouling of furnaces
sive alloys required in the tubes, steam piping and tur- and superheaters should not occur as long as the ash
bine for 1050 to 1100F (566 to 593C) steam conditions. characteristics are not severe compared to the tube
There are many factors affecting oil ash deposition metal temperatures. If such trouble is encountered,
on boiler heat absorbing surfaces. These factors may the solution can usually be found in improving com-
be grouped into the following interrelated categories: bustion conditions in the furnace and/or modifying the
characteristics of the fuel oil, design of the boiler, and sootblowing procedures.
operation of the boiler. Studies on both laboratory and field installations
have shown that the rate of ash deposition is a func-
Characteristics of fuel oil ash tion of the velocity and temperature of the flue gases
Sodium, sulfur and vanadium are the most signifi- and the concentration of oil ash constituents in the flue
cant elements in the fuel oil because they can form gases. The geometry of the furnace and the spacing
complex compounds having low melting temperatures, of tubes in the convection banks are selected in the
480 to 1250F (249 to 677C), as shown in Table 8. Such design of a boiler to minimize the rate of deposition.
temperatures fall within the range of tube metal tem- It is common practice to use in-line tube arrangements
peratures generally encountered in furnace and su- with wider lateral spacings for tubes located in higher
perheater tube banks of many oil-fired boilers. How- gas temperature zones. This makes bridging of ash
ever, because of its complex chemical composition, fuel deposits between tubes less likely and facilitates clean-
oil ash seldom has a single sharp melting point, but ing of tube banks by the sootblowers.
rather softens and melts over a wide temperature
range. Boiler operation
An ash particle that is in a sticky, semi-molten state Poor atomization of the fuel oil results in longer
at the tube surface temperature may adhere to the flames and frequently increases the rate of slag
tube if it is brought into contact by the gas flow over buildup on furnace walls which, in turn, makes it more
the tube. Even a dry ash particle may adhere due to difficult to keep the convection sections of the boiler
mutual attraction or surface roughness. Such an ini- clean. Completing combustion before the gases pass
tial deposit layer will be at a higher temperature than over the first row of tubes is especially important.
that of the tube surface because of its relatively low
thermal conductivity. This increased temperature pro-
motes the formation of adherent deposits. Therefore,
fouling will continue until the deposit surface tem- Table 8
perature reaches a level at which all of the ash in the Melting Points of Some Oil Ash Constituents
gas stream is in a molten state, so that the surface is Compound Melting Point, F (C)
merely washed by the liquid without freezing and
continued buildup. Aluminum oxide, Al2O3 3720 (2049)
In experimental furnaces, it has been found that Aluminum sulfate, Al2(SO4)3 1420 * (771)
Calcium oxide, CaO 4662 (2572)
the initial rate of ash buildup was greatest when the Calcium sulfate, CaSO4 2640 (1449)
sodium-vanadium ratio in the fuel oil was 1:6, but an Ferric oxide, Fe2O3 2850 (1566)
equilibrium thickness of deposit [0.125 to 0.25 in. Ferric sulfate, Fe2(SO4)3 895 * (479)
(3.175 to 6.35 mm)] was reached in approximately 100 Nickel oxide, NiO 3795 (2091)
hours of operation. When the fuel oil contained more Nickel sulfate, NiSO4 1545 * (841)
refractory constituents, such as silica, alumina and Silicon dioxide, SiO2 3130 (1721)
iron oxide, in addition to sodium and vanadium, an Sodium sulfate, Na2SO4 1625 (885)
equilibrium condition was not reached and the tube Sodium bisulfate, NaHSO4 480 * (249)
banks ultimately plugged with ash deposits. However, Sodium pyrosulfate, Na2S2O7 750 * (399)
these ash deposits were less dense, i.e., more friable, Sodium ferric sulfate, Na3Fe(SO4)3 1000 (538)
Vanadium trioxide, V2O3 3580 (1971)
than the glassy slags encountered with a 1:6 sodium- Vanadium tetroxide, V2O4 3580 (1971)
vanadium fuel oil. Both the rate of ash buildup and Vanadium pentoxide, V2O5 1275 (691)
the ultimate thickness of the deposits are also influ- Sodium metavanadate,
enced by physical factors such as the velocity and tem- Na2O V2O5(NaVO3)
• 1165 (629)
perature of the flue gases and particularly the tube Sodium pyrovanadate, 2Na2O V2O5 • 1185 (641)
metal temperature. Sodium orthovanadate, 3Na2O V2O5 • 1560 (849)
In predicting the behavior of a residual oil insofar Sodium vanadylvanadates,
as slagging and tube bank fouling are concerned, sev- Na2O V2O4 V2O5
• • 1160 (627)
eral fuel variables are considered including: 1) ash 5Na2O V2O4 1V2O5
• • 995 (535)
content, 2) ash analysis, particularly the sodium and * Decomposes at a temperature around the melting point.
vanadium levels and the concentration of major con-

21-24 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

Relatively large carbonaceous particles have a far tubes. High chromium contents, greater than 30%,
greater tendency to impinge on the tubes than do the give added corrosion resistance but at the expense of
smaller ash particles. If these larger particles are in a physical properties; 25Cr-20Ni has been used as a tube
sticky state, they will adhere to the tubes where oxi- cladding but even this alloy has not provided complete
dation will proceed at a slow rate with consequent protection. High-nickel high-chromium alloys may be
formation of ash. Fouling from this cause is difficult more resistant to oil ash attack under oxidizing con-
to detect by inspection during boiler outages because ditions, but the higher material cost must be justified
the carbonaceous material has usually disappeared by longer life, which is not always predictable.
completely. It can generally be detected during opera-
tion because flames are usually long and smoky, and Low temperature corrosion
sparklers may be carried along in the flue gases. In oil-fired boilers, the problem of low temperature
Regular and thorough sootblowing can have a de- corrosion resulting from the formation and condensa-
cisive effect on superheater and reheater fouling. (See tion of sulfuric acid from the flue gases is similar to
Chapter 24.) To be fully effective, however, that previously described for coal firing.
sootblowing cycles should be frequent enough so that Oil-fired boilers are more susceptible to low tem-
ash deposits can not build to a thickness where their perature corrosion than are most coal-fired units for
surfaces become semi-molten and difficult to remove. two reasons: 1) the vanadium in the oil ash deposits
In instances of extreme slagging, it is sometimes nec- is a good catalyst for the conversion of SO2 to SO3, and
essary to relocate sootblowers, to install additional 2) there is a smaller quantity of ash in the flue gases.
sootblowers to control deposition in a critical zone, or Ash particles in the flue gas react with and reduce the
to use additives. amount of SO3 vapor in the gas, and oil has consider-
The boiler load cycle can also have a significant ef- ably less ash than coal. Furthermore, coal ash is more
fect on the severity of slagging and superheater foul- basic than oil ash and more effectively neutralizes acid.
ing. A unit that is base loaded for long periods is more
apt to have fouling problems on a borderline fuel oil
than a unit that takes daily swings in load. In the lat- Methods of control
ter instance, the furnace generally remains cleaner The methods of control that have been used or pro-
due to periodic shedding of slag, with the result that posed to control fouling and corrosion in oil-fired boil-
the gas temperatures through the superheaters are ers are summarized in Table 9, but in every instance
appreciably lower. This eases the burden on the economics govern their applicability. There is no doubt
sootblowers and substantially controls ash deposit for- that reducing the amount of ash and sulfur entering
mation in the superheater-reheater tube banks. Over-
loading the boiler, even for an hour or two a day,
should be avoided, especially if excess air has to be
lowered to the point where some of the burners are
starved of air. The furnace is apt to become slagged
and ash deposition can creep into the superheater and
reheater tube banks.

Oil ash corrosion

High temperature corrosion
The sodium-vanadium complexes, usually found in
oil ash deposits, are corrosive when molten. A measur-
able corrosion rate can be observed over a wide range
of metal and gas temperatures, depending on the
amount and composition of the oil ash deposit. Fig. 28
shows the combined gas and metal temperature effects
on corrosion for a specific fuel oil composition of 150
ppm vanadium, 70 ppm sodium and 2.5% sulfur. As
the vanadium concentration of the fuel oil varies, the
amount of corrosion, compared to a 150 ppm vanadium
fuel, will increase or decrease according to the curve
shown in Fig. 29. The effect of the sodium level in the
fuel oil is not as clear. The sodium content does, how-
ever, definitely affect the minimum metal temperature
at which corrosion will be significant.
At the present time there appears to be no alloy that
is immune to oil ash corrosion. In general, the higher
the chromium content of the alloy, the more resistant
Fig. 28 Effect of gas and metal temperatures on corrosion of 304,
it is to attack. This is the main reason for the use of 316 and 321 alloys in a unit fired with oil containing 150 ppm vana-
18Cr-8Ni alloys for high temperature superheater dium, 70 ppm sodium and 2.5% sulfur. Test duration 100 hours.

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-25

 The Babcock & Wilcox Company

Low temperature sulfuric acid corrosion is reduced

by the formation of refractory sulfates by reaction
with the SO3 gas in the flue gas stream. By removing
the SO3 gas, the dew point of the flue gases is suffi-
ciently reduced to protect the metal surfaces. The sul-
fate compounds formed are relatively dry and easily
removed by the normal cleaning equipment.
In general, the amount of additive used should be
about equal to the ash content of the fuel oil. In some
instances, slightly different proportions may be re-
quired for best results, especially for a high tempera-
ture corrosion reduction, in which it is generally ac-
cepted that the additive should be used in weight ra-
tios of 2:1 or 3:1 (additive/ash), based on the vana-
dium content of the oil.
Several methods have been successfully used to in-
troduce the additive materials into the furnace. The
one in general use consists of metering a controlled
amount of an additive oil slurry in the burner supply
line. The additive material should be pulverized to
100% through a 325 mesh (44 micron) screen for good
dispersion and minimum atomizer wear.
For a boiler fired by a high pressure return flow oil
Fig. 29 Effect of vanadium concentration on oil ash corrosion. system (Chapter 11), it has been found advantageous
to introduce the additive powders by blowing them
the furnace is the surest means of control and that into the furnace at the desired locations. The powder
minimizing the effects of the ash constituents, once has to be reduced in size to 100% through a 325 mesh
they have deposited on the tubes, is the least reliable. (44 micron) screen for good dispersion.
Because the severity of fouling and corrosion depends The choice of a particular additive depends on its
not only on the fuel oil characteristics but also on boiler availability and cost to the individual plant and the
design and operating variables, a generalized solution method of application chosen. For example, alumina
to these problems can not be prescribed. causes greater sprayer plate wear than the other ma-
terials when used in an oil slurry.
Fuel oil supply The quantity of deposit formed is, of course, an im-
Although fuel selection and blending are practiced to portant consideration for each unit from the aspect of
some extent in the U.S., the common purpose is to pro- cleaning. A comparison of the amounts of deposit
vide safe and reliable handling and storage at the power
plant rather than to avoid fouling difficulties. Because
the threshold limits of sodium, sulfur and vanadium are
not well defined for either fouling or corrosion, use of Table 9
these means of control can not be fully exploited. Classification of Methods for Controlling Fouling
and Corrosion in Oil-Fired Boilers
Fuel oil additives
An approach that is effective where the fuel oil ash Fuel Oil Supply
is most troublesome involves adding, to the fuel or
furnace, small amounts of materials that change the Reduce amount of fuel ash Selection
constituents to the furnace Blending
character of the ash sufficiently to permit its removal Purification
by steam or air sootblowers or air lances.
Additives are effective in reducing the problems Design
associated with superheater fouling, high tempera-
Minimize amounts of fuel Furnace geometry
ture ash corrosion and low temperature sulfuric acid ash constituents reaching Tube bank arrangement
corrosion. Most effective are alumina, dolomite and heat transfer surfaces Metal temperature
magnesia. Kaolin is also a source of alumina. Analy- Gas temperature
ses of typical superheater deposits from a troublesome Sootblower arrangement
fuel oil, before and after treating it with alumina or
dolomite, are shown in three bar graphs on the left of Operation
Fig. 30. The results for a different oil treated with Minimize effects of bonding Load cycle
magnesia are shown in the bar graph on the right. and corrosive compounds in Sootblowing schedule
The reduction of fouling and high temperature cor- ash deposits Combustion ⎯ excess air
rosion is accomplished basically by producing a high Additives
melting point ash deposit that is powdery or friable Water washing
and easily removed by sootblowers or lances. When
the ash is dry, corrosion is considerably reduced.

21-26 Steam 41 / Fuel Ash Effects on Boiler Design and Operation

 The Babcock & Wilcox Company

Fig. 31 Effect of low excess air combustion on high temperature oil

ash corrosion.

proximately 0.5%, which is generally acceptable for

Fig. 30 Effect of fuel oil additives on composition of oil ash deposit. electric utility and industrial practice.
A number of large industrial boilers both in the U.S.
formed with different additives shows that dolomite and in Europe have been operating with low excess
produces the greatest quantity because of its sulfat- air for several years. As a result, the benefits of re-
ing ability, magnesia is intermediate, and alumina and ducing low temperature corrosion are well established.
kaolin form the least. However, when adequate clean- However, the benefits on high temperature slagging
ing facilities are available, the deposits are easily removed and corrosion are not wholly conclusive. In any event,
and the quantities formed should not be a problem. great care must be exercised to distribute the air and
fuel oil equally to the burners, and combustion condi-
Excess air control tions must be continuously monitored to assure that
As mentioned previously the problems encountered combustion of the fuel is complete before the combus-
in the combustion of residual fuel oils – high tempera- tion gases enter the convection tube banks.
ture deposits (fouling), high temperature corrosion and
low temperature sulfuric acid corrosion – all arise from
the presence of vanadium and sulfur in their highest
states of oxidation. By reducing the excess air from
7% to 1 or 2%, it is possible to avoid the formation of
fully oxidized vanadium and sulfur compounds and,
thereby, reduce boiler fouling and corrosion problems.
In a series of tests on an experimental boiler, it was
found that the maximum corrosion rate of type 304
stainless steel superheater alloy held at 1250F (677C)
in 2100F (1149C) flue gas was reduced more than
75% (Fig. 31) when the excess air was reduced from
an average of 7% to a level of 1 to 2%. Moreover, the
ash deposits that formed on the superheater bank
were soft and powdery, in contrast to hard, dense de-
posits that adhered tenaciously to the tubes when the
excess air was around 7%. Also, the rate of ash buildup
was only half as great. Operation at the 1 to 2% ex-
cess air level practically eliminated low temperature
corrosion of carbon steel at all metal temperatures
above the dew point of the flue gases (Fig. 32). How-
ever, much of the beneficial effects of low excess air
combustion are lost if the excess air at the burner fluc-
tuates even for short periods of time to a level of about
5%. Carbon loss values for low excess air were ap- Fig. 32 Effect of excess air on low temperature corrosion of carbon steel.

Steam 41 / Fuel Ash Effects on Boiler Design and Operation 21-27

 The Babcock & Wilcox Company

Two coal-fired boilers: one 685 MW pulverized coal unit and one 844 MW unit with CycloneTM furnaces.

21-28 Steam 41 / Fuel Ash Effects on Boiler Design and Operation