Professional Documents
Culture Documents
STEEL CHIMNEYS
THE CICIND CHIMNEY STANDARD
CICIND
Model Code for Steel Chimneys
TABLE OF CONTENTS
FOREWORD .................................................................................................................................. 1
0. INTRODUCTION ..................................................................................................................... 1
0.1 General .................................................................................................................................. 1
0.2 Appendices and Commentaries ............................................................................................. 1
0.3 Philosophy............................................................................................................................. 2
1. SCOPE........................................................................................................................................ 2
2. FIELD OF APPLICATION ..................................................................................................... 2
3. REFERENCES .......................................................................................................................... 2
4. NOTATIONS, UNITS AND DEFINITIONS .......................................................................... 3
4.1 General .................................................................................................................................. 3
4.2 Subscripts-Superscripts ......................................................................................................... 4
4.3 Units ...................................................................................................................................... 4
4.4 Definitions............................................................................................................................. 4
5. BASIS OF DESIGN AND SAFETY FACTORS .................................................................... 4
5.1 General .................................................................................................................................. 4
5.2 Reliability differentiation ...................................................................................................... 4
5.3 Partial Safety Factors ............................................................................................................ 5
5.4 Cross-wind Effects ................................................................................................................ 5
6. MATERIALS ............................................................................................................................. 5
6.1 General .................................................................................................................................. 5
6.2 Structural Steels .................................................................................................................... 5
6.3 Stainless and Alloy Steels ..................................................................................................... 8
7. ACTIONS (EXTERNAL AND INTERNAL) ......................................................................... 8
7.1 Permanent Load .................................................................................................................... 8
7.1.1 Dust load (temporary load) .......................................................................................................... 8
7. 2 Wind ..................................................................................................................................... 8
7.2.1 General ......................................................................................................................................... 8
7.2.2 Wind speed................................................................................................................................... 9
7.2.3 Wind load in the direction of the wind ....................................................................................... 13
7.2.4 Vortex shedding ......................................................................................................................... 16
7.2.5 Ovalling ..................................................................................................................................... 20
7.2.6 Increase of wind effects by nearby structures ............................................................................ 21
7.2.7 Damping ratio ............................................................................................................................ 22
7.2.8 First and second natural frequencies .......................................................................................... 23
7.2.9 Passive dynamic control............................................................................................................. 23
7.2.10 Special chimney designs for damping ...................................................................................... 24
7.3 Earthquake loading ............................................................................................................. 24
7.4 Thermal Effects ................................................................................................................... 24
7.5 Explosions ........................................................................................................................... 24
7.5.1 External explosions .................................................................................................................... 24
7.5.2 Internal explosions ..................................................................................................................... 24
7.6 Internal Effects governing the Chimney Design ................................................................. 24
7.6.1 High temperature flue gases ....................................................................................................... 24
7.6.2 Fire ............................................................................................................................................. 25
7.6.3 Chemical effects......................................................................................................................... 25
page ii CICIND Steel Chimney Model Code
DISCLAIMER
This CICIND Model Code is presented to the best of the knowledge of its members as a guide only.
CIC1ND is not, nor are any of its members, to be held responsible for any failure alleged or proved to be
due to adherence to recommendations or acceptance of information published by a Model Code or in any
other way.
FOREWORD
When it was formed in 1973, the “Comité International des Cheminées Industrielles” (CICIND) adopted as
a major goal the harmonization of national codes for the design of industrial chimneys. As a means to this
end, a subcommittee was appointed in 1981, charged with drafting a proposal for a model code for steel
chimneys which reflected the current “state-of-the-art” and a consensus of views, internationally. This
document was published in 1988, with Commentaries being published the following year.
Since 1988, the science and technology of chimneys has advanced and in 1995, CIC1ND appointed a
committee to revise the Model Code, recognizing current best international practice and knowledge.
The 2010 revision of the Model Code refers to the wind loads, both in along-wind and across-wind
direction. The revision resulted from the findings, that the surface roughness of the chimney site and its
surroundings should be taken into account. Corresponding amendment had been introduced in the revised
Model Code.
0. INTRODUCTION
0.1 General
Chimneys are required to carry vertically and discharge to the atmosphere, gaseous products of
combustion, chemical waste gases, or exhaust air or for the combustion (flaring off) of industrial waste
gases.
This Model Code contains guide-lines which reflect the current state of art in the design and construction of
steel chimneys. Nevertheless, the design, fabrication and erection of steel chimneys require a thorough
knowledge of these structures, the properties of the materials used, the actions occurring upon the structure
and the recognized rules of the relevant technologies. The design of steel chimneys should therefore only
be entrusted to appropriately qualified and experienced engineers. The construction and erection should be
carried out by firms competent in this class of work. At all times the work should be under the direction of
appropriately qualified supervisors.
CIC1ND will continue to try to improve the understanding of the behavior of chimneys. Further revisions
of this Model Code will therefore be published from time to time.
e) Change of the meaning of certain regulations of the Model Code where these are falsely expressed
or obviously wrong.
f) Definition of the meaning of certain regulations of the Model Code which are so badly formulated
that they could easily be misinterpreted even by experts.
Certain information from the Model Code is repeated in the commentaries when this simplifies the
presentation of the ideas.
0.3 Philosophy
One of the main objectives of any code governing construction is the creation of a model which resembles
as far as possible, the real situation. The model should be sufficiently “safe, simple and true”. It is very
rarely that simplicity and truth are compatible, so a model must be used which provides an optimum
compromise between truth, simplicity, safety and economy.
While the judgments of ‘sufficiently true’ and ‘sufficiently simple’ are subjective, ‘sufficiently safe’ is
capable of rational judgment. This code interprets ‘sufficiently safe’ in terms of the social and economic
consequences of failure. It does this by comparing the probabilities of failure for given safety factors during
its design life with the failure probabilities required to satisfy accepted social and economic criteria. This
leads to the development of safety factors which ensure that a chimney will have a probability of failure
during its design lifetime between 10−3 and 10−4 , depending upon its reliability category.
CICIND has departed from generally accepted principles of steelwork design and construction only when
this was required by the philosophy outlined above or by specific chimney requirements.
1. SCOPE
This Model Code relates to the structural design and construction of steel chimneys of circular cross-
section, with a minimum height of 15m, with or without linings, and to the design and application of linings
to such chimneys where required. It also relates to chimneys with a height less than 15m and a slenderness
ratio more than 16. The Model Code does not deal with architectural or thermal aspects of steel chimneys
nor with their foundations, except insofar as they affect the chimney’s structural design. The Model Code
does not deal with those aspects of the design and construction of steelwork, refractories and insulation
which are not peculiar to chimneys.
2. FIELD OF APPLICATION
The Model Code is valid for all steel chimneys of circular cross- section. However, the design rules have
been formulated for self supporting chimneys taller than 15m. For other chimneys simplification may be
acceptable. Additional information is given in the Appendices and Commentaries.
3. REFERENCES
[1] “CICIND model code for concrete chimneys — Fart A, The Shell “, August 1998 CICIND,
Zurich, Switzerland.
[2] “Eurocode 3.2: Design of Steel Chimneys” ENV 1993-3-2: 1997
[3] Thom, H.C.S.: “Distribution of extreme winds over oceans” Journal of the Waterways, Harbors
and Coastal Engineering Division. Proc. of the American Society of Civil Engineers, February
1973.
[4] Vickery, B.J: “Wind loads and design for chimneys “, CICIND REPORT, Vol. 14, No. 2, 1998
CICIND Steel Chimney Model Code page 3
[5] Eurocode 1 — Basis of Design and actions on structures — Fart 2—4: Actions on structures —
Wind Actions ENV 1991-2-4: 1995
[6] Van Koten, H: “A calculation method for the fatigue life of steel chimneys subject to cross-wind
oscillations “, CICIND REPORT, Vol. 14, No. 2, 1998
[7] Ruscheweyh, H.: “Experience with Vortex Excited Oscillations of Steel Chimneys “, CICIND
REPORT, Vol.11, No. 2, 1995
[8] Ole Hansen, S: “Vortex — induced vibrations of line-like structures “, CICIND REPORT , Vol.
14, No. 2, 1998
[9] Van Koten, H: ‘Structural damping”, HERON report no.4, 1977, Delft. The Netherlands
[10] Berger, G : “Measured damping decrements of steel chimneys and their estimation taking account
of their type “, CICIND REPORT, Vol. 15, No. 1, 1999
[11] Turner J.G.: “Wind load stresses in steel chimneys “, CICIND REPORT, Vol. 12, No. 2, 1996
[12] Hirsch, G.& Jozsa, M.: “Optimum control of chimney vibration”, CICII’JD REPORT, Vol. 10,
No. 1,
[13] Bierrum,N.R.: “Mis-tuned Mass Dampers”, CICII’JD REPORT, Vol. 10, No.2, 1994
[14] Warren, R.M. & Reid, S.L. “Shell to Flue Impact Damping for Dual Wall and Multi-Flue
Chimneys” CICIND REPORT Vol. 10,N0. 1,1994
[15] Ruscheweyh, H., Kammel, C. & Verwiebe, C. “Vibration Control by Passive Dampers CICND
REPORT Vol. 12, No. 2, 1996
[16] Bunz, G., Diepenberg, H. and Rendie, A.: “Influence offuel oil characteristics and combustion
conditions of flue gas properties in W T boilers” Journal of the Institute of Fuel, Sept.1967
[17] Lech and Lewandowski: ‘Prevention of cold end corrosion in industrial boilers” Corrosion,
March 1979, Atlanta, U.S.A.
[18] Henseler, F.: “Desulphurisation Systems and their Effect on Operational Conditions in Chimneys
“, CICIND REPORT, Vol. 3, No. 2, 1987.
[19] “CICIND chimney protective coatings manual”, CICIND, Zurich, Switzerland
[20] Schulz, U.: “Die Stabilität axialer Zylinderschalen mit Mantelöffnungen”, Bauingenieur 5 1,1976.
[21] ‘European Recommendations for Steel Construction: Buckling of Cylinders” ECCS 1984
[22] Bouwman, E.P.: “Bolted connections dynamically loaded in tension “. Proceedings ASCE,
Journal of the Structural Division, 59,1982.
[23] “CICIND Model Code for Concrete Chimneys — Part C, Steel Liners “, December 1995
CIC1ND, Switzerland
V wind-speed (m/s)
w wind-force (N/m)
Cross-sectional forces
M bending moment (Nm)
e eccentricity (m)
Dimensions
h height (m)
z height above ground level (m)
d diameter (m)
t wall thickness (m)
4.2 Subscripts-Superscripts
y yield limit
k characteristic value
* stress multiplied by load factor
cr critical
4.3 Units
Generally, the units of the SI system are used.
Examples:
• m (meter) and mm (millimeter) for dimensions
• MN (Mega Newton) and N (Newton) for forces
• MPa for stresses
In those cases where other units are used, the relevant references are given.
4.4 Definitions
The common names of parts of a steel chimney are explained in commentary 1.
Two classes of reliability related to the consequences of structural failure are used — Normal and Critical,
as defined below. The choice of reliability category shall be decided by the chimney owner and relevant
statutory authorities. Most chimneys will, however, be regarded as of Normal reliability.
Critical chimneys: Chimneys erected in strategic locations, such as nuclear power plants or in densely
populated urban locations. Major chimneys in industrial sites where the economic and/or social
consequences of their failure would be very high.
Normal Chimneys: All normal chimneys at industrial sites or other locations. (Typically chimneys in
industrial sites, power plants or chimneys less than 100m tall in urban locations, where any domestic
dwelling is outside the falling radius of the chimney).
6. MATERIALS
6.1 General
The materials generally used for steel chimneys are described in the
CICIND METALLIC MATERIALS MANUAL.
Table 6.1 shown below is a copy of Table 8.1 of this manual.
Special steels can be used provided they are precisely specified and their characteristics, such as yield
stress, tensile strength, ductility and weldability, enable the Model Code to be put into application.
fy
fk = (6.1)
1.1
Characteristic values of the yield stress of structural and high alloy steels at ambient and higher
temperatures are given in Table 6.1.
Characteristic values of the modulus of elasticity at temperatures up to 600°C are given in Tables 6.2,
and characteristic values of the thermal expansion coefficients are given in Table 6.3.
Values at intermediate temperatures can be obtained by linear interpolation.
Note: The conditions downstream of a flue gas scrubber or the presence of chlorides in the condensate will
radically increase the corrosion rate, possibly rendering these stainless steels unsuitable for these
applications.
Ordinary stainless steels are not suitable for use in contact with flue gases containing alkalis.
In cases where it is not possible to avoid high chemical load on the internal face of the structural shell, see
paragraph 7.6.3, the use of a protective coating may be considered (see ref. [19]). Alternatively, a steel liner
or liners, possibly of titanium or high nickel alloy, is a possible solution. See section 10 on Steel Liners.
Low copper alloy steels have good resistance to atmospheric corrosion, except in a marine environment or
other environment where chlorides are present. These steels also show some corrosion improvement over
carbon steel when in contact with flue gases where acid condensation of SO2/SO3 (not of HCL
condensation) is intermittent only (e.g. during shutdowns of a stack in intermittent service, its metal
temperature being normally above acid dew point).
When the metal temperature is below acid dew point for prolonged periods, the performance of low copper
alloy steels in contact with flue gases is similar to that of carbon steel.
Where stainless or alloy steel components are connected to carbon steel, bolted connections are preferred.
In order to avoid accelerated corrosion due to galvanic action, such connections should include insulating
gaskets. Welded connections are permitted, provided specialist metallurgical control is exercised with
regard to weld procedures, electrode selection, etc. Care should be taken to use the correct coefficient of
expansion for the grade and temperature of the steel being considered.
7. 2 Wind
7.2.1 General
The wind load on a chimney depends in the first instance upon the magnitude of the wind speeds in the area
in which the chimney is to be erected and their variation with height. Apart from that the wind loads, in the
direction of the wind or perpendicular to that direction, will be influenced by some or all of the following:
a) local topography
b) level of turbulence
c) presence of nearby structures, including chimneys
d) air density
CICIND Steel Chimney Model Code page 9
where:
V(z) = 10-minute mean design wind speed at elevation z (m/s)
z = height above ground level (m)
k(z) = height factor defined in expression (7.2)
kt = topographical factor, see 7.2.2.3
b
z
k(z) = a ⋅ for z ≥ z min
10 (7.2)
= k(z min ) for z < z min
z min = height below which the wind speed (and turbulence intensity) is assumed to remain the same,
see Table 7.1
Vb = basic 10-minute mean wind speed (m/s)
The height factor depends on the terrain roughness of the area in which the chimney is to be erected. The
Model Code distinguishes three terrain categories given in Table 7.1.
page 10 CICIND Steel Chimney Model Code
I Sea, lakes or flat and horizontal area with negligible vegetation and without
0.01 2
obstacles
II Area with low vegetation such as grass and isolated obstacles (trees,
0.05 4
buildings) with separations of at least 20 obstacle heights
III Area with regular cover of vegetation or buildings or with isolated obstacles
with separations of maximum 20 obstacles heights (industrial area, suburban 0.3 8
terrain, permanent forest)
3) If the suitability of a different formulation of the height factor can be proved (together with an
appropriate formulation of the turbulence intensity, expression (7.3)), it may be used (see
Commentary C3.1.3).
The exposure coefficients a, b and c, used in expressions (7.2) and (7.3) are given in Table 7.2.
Terrain I II III
The transition between terrain categories has to be considered when selecting the value of z0 . If the
upstream distance to a terrain with lower roughness is smaller than the larger of 5km or 100h, where h is
the height of the chimney, then the lower value of the terrain roughness should be used.
For terrains with a wind direction dependent roughness the (mean) terrain roughness can be determined by
considering the terrain roughness in angular sectors of say 30o multiplied by an appropriate direction factor
derived from the local directional wind statistics. When there is a choice between two or more terrain
categories in a given sector, then the lowest terrain category should be used.
The turbulence intensity, used in the formulation of the gust factor, see 7.2.3.3, and in the formulation of
cross-wind load, see 7.2.4, is given by:
−b
z
Iν (z) = c ⋅ for z ≥ z min
10 , (7.3)
= Iν (z min ) for z < z min
CICIND Steel Chimney Model Code page 11
where the coefficients b and c are defined in Table 7.2 and z min in Table 7.1.
where:
Φ = upwind slope H/L in the wind direction, see Figures 7.1 & 7.2
s = factor obtained from Figures 7.1 & 7.2
H = height of hill or escarpment
x = distance of chimney from crest
z = height of considered position in chimney
Le = effective length of the upwind slope, defined in Table 7.3
Lu = actual length of upwind slope in the wind direction
Ld = actual length of downwind slope in wind direction
Le = Lu Le = H / 0.3
where:
wm(z) = 10-minute mean wind load per unit height, see 7.2.3.2.1
w g (z) = static equivalent of the wind load per unit height due to gusts, see 7.2.3.3.1
where:
ρa = density of air, see 7.2.3.2.2 (kg/m3)
Momentary variations in the density due to atmospheric changes need not be taken into account.
The air density relevant to a chimney site at an altitude h1 (m) can be found from the expression:
h1
ρa = 1.25 − (kg/m3) (7.8)
8000
in which:
ki = interference factor accounting for the influence of nearby structures, see 7.2.6.1;
ka = end-effect factor for chimneys with an aspect ratio below 20, see expression (7.12);
C D,0 = basic shape factor as a function of the Reynolds number and reflecting the influence of
turbulence and the surface roughness of the stack, see expression (7.11) and Figure 7.3.
The basic shape factor is given by, see Figure 7.3:
page 14 CICIND Steel Chimney Model Code
The (basic) shape factor is measured in wind tunnel experiments. Most experiments are so-called 2D, i.e.
the cylinder is extending over the full width or height of the wind tunnel. In this arrangement the influence
of the flow near the free end of a cantilevered cylinder on the shape factor is absent. The resulting mean
shape factor (averaged over the height) of a cantilevered cylinder is smaller than its 2D-value. This
reduction decreases for larger aspect ratios. This reduction, called the end effect factor is given by
expression (7.12) and Figure 7.4:
h h
k a = 0.60 + 0.129 ⋅ log for 1 ≤ ≤6
d d
h h
= 0.70 + 0.574 ⋅ log − 0.778 for 6 ≤ ≤ 20 (7.12)
d d
h
=1 for > 20
d
CICIND Steel Chimney Model Code page 15
where:
G = gust factor, see 7.2.3.3.2
h = height of the chimney above ground level
z = height above ground level
w m (z) = 10-minute mean wind load per unit height at height z, see 7.2.3.2.1
G = 1 + 2 ⋅ g ⋅ Iν (zs ) ⋅ B2 + R 2 (7.14)
where:
0.577
g = peak factor = 2 ⋅ log e (νT) + (7.15)
2 ⋅ log e (νT)
R2
νT = 600 ⋅ f1 ⋅ ; νT ≥ 48 (7.16)
B + R2
2
1
B2 = background factor = 0.63
(7.17)
b+h
1 + 0.9 ⋅
L ( zs )
page 16 CICIND Steel Chimney Model Code
π
R 2 = resonance response factor = ⋅ SL ⋅ R h ⋅ R d (7.18)
4⋅ζ
6.8 ⋅ f L
SL = power spectral density = (7.19)
(1 + 10.2 ⋅ fL )5/3
f1 ⋅ L(zs )
fL = (7.20)
V(z s )
β
z
L ( z ) = turbulent length scale = 300 ⋅ for z min ≤ z ≤ 300m
300
(7.21)
β
z
L ( z ) = 300 ⋅ min for z < z min
300
The exponent β depends on the terrain category and is given in Table 7.4.
Terrain I II III
1 1
Rx = − ⋅ {1 − exp ( −2 ⋅ ηx )} ; Rx = 1 for ηx = 0 (7.22)
ηx 2 ⋅ η2x
with
4.6 ⋅ f L
ηx = ⋅x for x = d or x = h (7.23)
L(zs )
ζ = structural plus aerodynamic damping (plus the damping due to an external damping device if
applied) expressed as a fraction of critical damping, see 7.2.7
f1 = natural frequency of the chimney oscillating in its first mode (s-1)
h = height of chimney (m)
fn ⋅ d
Vcr,n = (7.25)
St
where the index n refers to the first, second or higher natural frequency (mode).
CICIND Steel Chimney Model Code page 17
In most cases the structural movements are dominated by the first natural frequency, but the second and
sometimes even the third natural frequency can be important.
The structural movements can substantially be increased by nearby chimneys in a group or row
arrangement. The influence is discussed in Section 7.2.6.
The cross-wind movements depend strongly on the mass and the damping of the chimney and on the
motion induced aerodynamic forces. The latter may generate a negative aerodynamic damping which
opposes the positive structural damping. Excessively large cross-wind movements occur when the total
damping is close to zero. The cross-wind movements are to a large extent determined by the ratio of two
dimensionless parameters, i.e. the Scruton number Sc and the aerodynamic damping parameter K a . Both
parameters depend on the mode shapes associated with the natural frequencies.
The cross-wind movements are small and (mainly) determined by the classical lift force when
Sc n
1 (7.26)
4 ⋅ π ⋅ K a,n
where Sc n and K a,n are given by, respectively, expression (7.34) and (7.36).
They are large and determined by the motion induced aerodynamic forces when this ratio is smaller than
one.
Cross-wind vibrations may cause fatigue. Fatigue is determined by a combination of the number of load
cycles during the operational life time of the chimney and the stress range occurring; see Section 8.5 for a
detailed discussion.
The effect of vortex shedding need not be investigated when:
Vcr,n > 1.25 ⋅ V(h) (7.27)
where V(h) is the design wind speed at the top of the chimney.
where
2
σ y,n = d1 ⋅ c1,n + c1,n +c2,n = nth mode standard deviation of deflection (7.29)
Scn
4
k p,n = 2 ⋅ 1 + 1.2 ⋅ arctan 0.75 ⋅ = nth mode peak factor (7.30)
4 ⋅ π ⋅ K a,n
u n (z) = mode shape of nth natural frequency, with u n (z) = 1 at the point of maximum deflection.
For the first natural frequency a parabolic mode shape may be used, see Remark below.
page 18 CICIND Steel Chimney Model Code
2
Scn
c1,n = 0.5 ⋅ a L,n ⋅ 1- (7.31)
4⋅π⋅K
a,n
a 2L,n ⋅ Ca,n
c 2,n = (7.32)
K a,n
0.4 ⋅ f1
a L,n = = limiting deflection amplitude as a fraction of d1 (7.33)
fn
4 ⋅ π ⋅ m o,n ⋅ ζ s
Sc n = = Scruton number of mode n (7.34)
ρa ⋅ d12
ρa = air density , see 7.2.3.2.2
∫ m(z) ⋅ u n (z) ⋅ dz
2
0
m o,n = h
= equivalent mass per unit length of mode n (7.35)
∫ u 2n (z) ⋅ dz
0
The turbulence intensity given by expression (7.3) is for equilibrium meteorological conditions. Under
certain meteorological conditions, giving very cold and stable air, the turbulence intensity can be
suppressed to a much lower value. If this situation can occur it is recommended investigating the
sensitivity of the cross-wind deflection for zero turbulence intensity. Typically, this is relevant for
critical wind speeds of up to 10 m/s.
The difference between the approximations and the full expressions can increase the deflections
by up to 25% or more. Though the deflection itself might be small.
3) It is recommended to apply the full expressions of Commentary C3.3 if expression (7.42) applies.
The standard deviation of the deflection, expression (7.25), as a function of the wind velocity is used to
compute the partial and the total Palmgren-Miner sum in a fatigue analysis.
page 20 CICIND Steel Chimney Model Code
7.2.5 Ovalling
In most cases, a suitably sized stiffening ring at the top of a chimney will eliminate problems associated
with ovalling.
where w 5sec is the wind pressure at height (z) averaged over 5 seconds (N/m).
Note: The assumption that a 5 sec gust wind speed at height z = 1.4·V(z) is safe at all heights.
0.5 ⋅ t E
f o,1 = 2
⋅ (7.46)
d ρs
where:
E = Young’s Modulus of the steel shell (Pa)
t = average shell thickness over the top third (m)
d = shell diameter (m)
ρs = density of shell material (kg/m3)
Substituting typical values of E and ρs, the associated critical wind speed is then:
t
Vr = 6500 ⋅ (m/s) (7.47)
d
These vibrations can be reduced sufficiently by stiffening rings. The distance between stiffeners L shall not
exceed the smaller of the following two expressions:
L d L 10.9
< 0.56 and < (7.48)
d t d z 0.06
The associated moment of inertia of the stiffening ring section (together with the participating length of
shell) about its centroid, see Figure 7.5, must be larger than:
p0 ⋅ d 4 ⋅ ( d + 6 ⋅ b r )
I r ≥ 0.060 ⋅ (m4) (7.49)
E⋅t
where:
p0 = wind pressure = 0.5 ⋅ρa ⋅ V 2
b r = width of the stiffening ring
The participating length of the shell = d ⋅ t but its area must not exceed that of the stiffener ring, see
Figure 7.5.
CICIND Steel Chimney Model Code page 21
When the distance is smaller than 4d the shape factor of the upwind chimney can be substantially larger
than for a single chimney. For these small distances expert advice is required.
7.2.6.2 Increase in cross-wind response
When an approximately cylindrical structure (e.g. another chimney) is upwind and within 15 diameters of a
chimney of similar or smaller height, aerodynamic “Wake Interference” effects can considerably increase
the downwind chimney’s cross-wind response (the diameter concerned being that of the interfering
structure). The increase is not yet fully understood, but is thought to be due to increases in both lift
coefficient and negative aerodynamic damping. Note that aerodynamic stabilizers (e.g. helical spoilers) are
ineffective in controlling response in cases of wake interference. For a spacing ratio (a/d) greater than 10,
the magnification factor kc, applied to the response amplitude, calculated per expression (7.24), may be
estimated as follows:
a
k c = 1.0 for ≥ 15
d
(7.50)
a
= 1.5 for = 10
d
for intermediate values linear interpolation is permitted.
page 22 CICIND Steel Chimney Model Code
For a spacing ratio (a/d) less than 10 there is a risk of very large increases in amplitude. In these
circumstances the chimney’s structural damping should be increased (e.g. by the use of a tuned mass
damper) to ensure that the Scruton Number exceeds 25. At this value of Scruton Number, the amplitude of
response is expected to be minimal.
The associated critical wind speed and value of “c2” in expression (7.28) increase with decreasing values of
a/d due to a reduction in the value of the Strouhal Number. This can be important in the design of a tuned
mass damper. Figure 7.6 shows the relationship between Strouhal Number and a/d.
When the interfering structure or chimney is less than 2 diameters away, “Interference Galloping” can
cause even greater increases in the chimney’s response. Probably the best solutions in this case would be
either to fit tuned mass dampers, or to connect structurally, the chimney to the interfering structure, using
an energy absorbing connection system.
λ < 26 0.006
λ > 28 0.002
Notes:
1) If rotation of foundation decreases the first natural frequency more than about 10% the foundation
is considered to be soft and the damping ratio may be increased by 0.0005.
2) λ = liner length / liner diameter
3) In order to ensure impact damping the gap between the liner and its restraint should not be greater
than 50mm.
The damping for wind loading in wind direction can be increased by the aerodynamic damping:
V
ζ a = 2.7 ⋅10−6 ⋅ (7.51)
f1 ⋅ t
in which:
V = design wind speed V(z) at the top of the chimney for wind loading in wind direction, see 7.2.2.2
V = 0 for cross-wind loading
f1 = fundamental natural frequency, see 7.2.8
t = thickness of the wall in the top third. Where chimneys are lined, t = total mass per square meter
over the top third (kg/m2) divided by 7850 kg/m3
in which:
µ = mass of the shell per unit length (kg / m)
Most such dampers are mounted near the top of the chimney. Because of their profile and small size, the
associated increase in wind drag is minimized. The use of damping devices, therefore, has been proved to
be beneficial in the design of steel chimneys and they can be safely retro-fitted without incurring significant
increase in wind drag loads.
Tuned mass dampers provide an extra mass, coupled to the chimney by an energy absorbing medium,
which absorbs the wind induced energy. Tuned mass dampers have proven effective in reducing self-
generated along wind and cross-wind vibrations and also the effect of nearby chimneys or structures.
Other chimney damping devices such as hanging chains have also been successfully used.
7.5 Explosions
Consideration must be given to the effects of oxidation when the material being used is close to its
temperature limit. This is especially so with gas turbine exhausts, where levels of excess air can be greater
than those normally experienced. This problem may not be solved solely by an increase in corrosion
allowance as the environment may be polluted by the corrosion product. Expert advice should be sought on
the choice of suitable material.
7.6.2 Fire
The risk of a chimney fire should be assessed. Chimney fires can be caused by ignition of:
• Unburned fuel carried over from the associated boiler or furnace.
• Where the associated furnace is in petrochemical service, unburned hydrocarbon carryover
following a furnace tube rupture.
• Soot, sulphur and other deposits.
During chimney fires, the radiant heat loss to atmosphere from a bare steel chimney is often sufficient to
maintain its temperature at a reasonable level. By contrast an externally insulated steel chimney or a bare
steel chimney close to a reflective surface will quickly buckle during a fire. In such cases, if the risk of
internal fire is significant, a refractory concrete internal liner should be installed to provide a degree of fire
protection. Typically, a castable refractory lining following the requirements of Appendix 3 will provide
sufficient fire protection for most situations.
Degree of Operating hours per year when temperature of the surface in contact with flue
chemical load gases is below estimated acid dew point +10°C
Low < 25
Medium 25 - 100
Table 7.6 ─ Degree of chemical load for gases containing sulphur oxides.
Notes:
1) The operating hours in Table 7.5 are valid for an SO3 content of 15 ppm. For different values of
SO3 content, the hours given vary inversely with SO3 content. When the SO3 content is not known,
chimney design should be based upon a minimum SO3 content amounting to 2% of the SO2
content in the flue gas.
2) In assessing the number of hours during which a chimney is subject to chemical load, account
should be taken of start-up and shut-down periods when the flue gas temperature is below its acid
dew point.
3) While a steel chimney may generally be at a temperature above acid dew point, care should be
taken to prevent small areas being subject to local cooling and therefore being at risk of localized
acid corrosion. Local cooling may be due to:
• air leaks
• fin cooling of flanges, spoilers or other attachments
• cooling through support points
• downdraft effects at top of the chimney
4) The presence of chlorides or fluorides in the flue gas condensate can radically increase corrosion
rates. Estimation of the corrosion rate in these circumstances depends upon a number of complex
factors and would require the advice of a corrosion expert in each individual case. However, in the
page 26 CICIND Steel Chimney Model Code
absence of such advice, provided the concentrations of HC1 < 30mg/m3 or of HF < 5mg/m3 and
if the operating time below acid dew point does not exceed 25 hours per year, the degree of
chemical load may be regarded as “low”.
5) Regardless of temperatures, chemical load shall be considered “high” if halogen concentrations
exceed the following limits:
• hydrogen fluoride: 0.025% by weight (300 mg/m3 at 20°C and 1bar pressure)
• elementary chlorine: 0.1% by weight (1300 mg/m3 at 20°C and 1 bar pressure)
• hydrogen chloride: 0.1% by weight (1300 mg/m3 at 20°C and 1 bar pressure)
6) Saturated flue gas conditions downstream a desulphurization system: “High” chemical load”.
The carrying capacity check shall prove that the forces resulting from the working loads multiplied by the
load factors do not exceed the resistance of the shell. The check should comprise both the strength and
stability proof. The calculations shall be carried out for the corroded thickness of the steel (without
corrosion allowance). The serviceability shall be checked under working loads without load factors
A fatigue check shall be carried out if movement due to shedding is expected (see 7.2.4).
For unstiffened chimneys with a ratio of L/R> 50 (where L height of chimney and R radius), stresses may
be safely calculated assuming beam theory, flexural stresses being added vectorially ovalling stresses. For
unstiffened chimneys (i.e. chimneys without stiffening rings or substantial flanged joints) having L/R < 50,
shell theory or finite element modeling should be used, considering flexural and ovalling stresses
simultaneously. This will lead to reduction in compression stress at the chimney base or immediately above
changes in chimney diameter, but will increase compression stresses elsewhere. Similarly, this will lead to
increases in tensile stresses at the base and immediately above a change in chimney diameter, which will be
important in deriving bolt tensions.
The increase in tensile stress in these regions may be approximated by the expression:-
where:
N
β = h⋅ (8.3)
E⋅I
and
h = height of the chimney (m)
N = total axial load at the base of the shell (without load factor) (N)
E·I = stiffness of the cross section at the base of the chimney (Nm2)
The second order moment M1I is approximately determined from:
β2
M1I = M I ⋅ 1 + (8.4)
8
where MI is the wind moment at any particular level.
This simplified approximation may only be used when β <0.8 and Nh / N < 0.1, where Nh is the design
value of the total vertical load at the top of the shell.
σ*2 *2 * *
x + σy − σx ⋅ σy + 3 ⋅ τ
*2
≤ fk (8.5)
Note: The ovalling stresses are both negative and positive and the maximum value of expression (8.5)
occurs when σ*x and σ*y are of opposite signs.
8.3.4 Stability
The proof of stability of the shell is given if the critical buckling stress divided by 1.1 is greater than the
sum of longitudinal stresses due to bending and compression:
σk
σ*N + σ*B ≤ (8.6)
γm
where:
σ*N , σ*B = normal and bending compressive stress at ultimate limit state
( )
= 1 − 0.412 ⋅ λ1.2 ⋅ f y for λ ≤ 2
fy (8.7)
= 0.75 ⋅ for λ > 2
λ2
fy = yield strength of steel at design temperature, see Table 6.2
fy
λ = (8.8)
α ⋅ σcr
page 28 CICIND Steel Chimney Model Code
E⋅t
σcr = critical elastic buckling stress = 0.605 ⋅ (8.9)
r
E = Young’s modulus of steel at design temperature, see Table 6.3
t = corroded plate thickness
r = radius of the structural shell of the chimney at section considered
α N ⋅ σ*N + α B ⋅ σ*B
α = (8.10)
σ*N + σ*B
The influence of the grade of steel as well as that of the σ min / σ max ratio is neglected.
To prevent fatigue occurring the stress must be smaller than the cut-off limit in the S/N diagram, see Figure
8.3. Generally, this requires a large Scruton number, typically 15 or larger, see expression (7.22). This may
be the case in a multiflue chimney or a chimney fitted with a refractory liner. However, in the case of an
unlined chimney, without a damping device to significantly reduce the cross-wind vibrations, this may
require a large wall thickness making the design uneconomic.
It is recommended to attach a damping device to chimneys with a Scruton number less than 15 and to make
sure the stress range is smaller than the cut-off limit in the S/N diagram.
For existing chimneys the Palmgren-Miner sum and the expected lifetime of the chimney can be estimated
using the expressions given in the next section.
{ }
N n = 3.15 ⋅107 ⋅ T ⋅ f n ⋅ exp(- V12 / Vo2 ) - exp(- V22 / Vo2 ) ≥ 200 ⋅ T (8.13)
where:
V1 and V2 are the lower and upper limit of the range of wind speeds in which cross-wind vibrations
occur.
This range is determined by the Scruton number, the aerodynamic damping parameter and the
turbulence intensity. A first order estimate is found from Figure C3.12 by the crossings of the
horizontal line given by K a,0 and the line for the respective turbulence intensity, where K a,0 is
given by:
Sc n
K a.0 = with Sc given by (7.30) and K a,n given by expression (7.32).
4 ⋅ π ⋅ K a,n
The Palmgren-Miner sum is used for the fatigue check, i.e. the chimney is expected to develop a crack,
which ultimately will result in a failure of the weakened section if the factored Palmgren-Miner sum, the
Palmgren-Miner sum multiplied by a partial safety factor, is equal to or larger than 1.
page 30 CICIND Steel Chimney Model Code
The Palmgren-Miner sum is given by the maximum value of the sum of the partial Palmgren-Miner sums
for each active cross-wind vibration mode. The partial Palmgren-Miner sum of mode n is given by:
f y ,T a
k
1 a a2
M n (z) = N n ⋅ ∫ N W,0 ⋅
a
⋅ 2 ⋅ exp - da
2 ⋅ σ 2 (z)
(8.15)
a min W,0 σa,n (z) a,n
where:
a = stress due to cross-wind oscillations
k = slope of the fatigue curves in the S/N diagram, Figure 8.3
σa,n (z) = standard deviation of the cross-wind stress of mode n at height z, see expression (8.16)
a min = stress cut-off limit = 0.5 x {stress range cut-off limit in the S/N diagram}
f y,T = yield strength of the shaft’s steel at operation temperature T, see expression (6.2)
a W,o = 0.5 x {stress range of specified detail category at 5.106 load cycles}
h
2
d(z) ⋅ ∫ m(s) ⋅ ( 2 ⋅ π ⋅ f n ) ⋅ u n (s) ⋅ (s - z) ⋅ ds
z
σa,n (z) = ⋅ σ y,n (8.16)
2 ⋅ I(z)
Notes:
1) Due to the approximations applied to derive the expressions in Section 7.2.4.2, σ y,n is not a
function of the wind velocity. As a result the integration of the wind and stress distribution
functions are separated, see Commentary 4 for details.
2) For a detailed fatigue analysis the full response curve of the chimney as a function of the wind
velocity is needed, see Commentaries 3 and 4 for details.
with:
γ = partial safety factor for fatigue; the recommended value is 10.
If the maximum value of the factored Palmgren-Miner sum is less than one no crack is expected to occur
during the design life time. Nevertheless, occasionally the deflection amplitude may be sufficient to cause
alarm. In such cases the amplitude limitation of Section 5.4 may govern.
CICIND Steel Chimney Model Code page 31
1) Butt welds, when high quality has to be achieved and 4) Butt weld: welded one side only
verified:
5) T ─ joint by double-bevel butt
• developed root, cap pass counter welding
weld
• evenly machined surface in stress direction
2) 6) T ─ joint by double Y –butt weld
Butt weld: developed root, cap pass counter welding
with broad root face
3) Butt weld:
7) T ─joint with special quality
• welded one side only double fillet weld
• through-welding of seam root and plane surfaces
• secured on opposite side by auxiliary welding aid 8) T ─ joint double fillet weld
e.g. weld-pool backing ceramics or copper rail
CICIND Steel Chimney Model Code page 33
Note: The external corrosion allowances quoted in Table 8.1 are suitable for a normal environment. When
a chimney is sited in an aggressive environment, caused by industrial pollution, nearby chimneys or
close proximity to the sea, consideration should be given to increasing these allowances.
not applicable
<65°C low
(chemical load always “high”) 1)
not applicable
medium
(chemical load always “high”) 1)
medium 4mm 3)
medium 2mm
Table 8.2 ─ Internal corrosion allowance (CI) for carbon steel only.
For chimneys handling flue gases.
CICIND Steel Chimney Model Code page 37
Notes:
1) Provided acid concentration in the condensate is less than 5% and chloride concentration does not
exceed 30mg/m3, high molybdenum stainless steel (such as ASTM Type 316L) is suitable within
this temperature limit, using a corrosion allowance of 3mm for a 20 year life. These conditions
are, however, unlikely to be met in a chimney downstream of a FGD system, generating
condensing gases. In these circumstances great care is required in the protection of the gas face of
the chimney or its liner, e.g. by cladding with a suitable high nickel alloy or titanium or by the
application of a suitable organic coating. For further guidance, see:
CICIND Chimney Coatings Manual.
2) In conditions of low chemical load, “Corten” steel shows some improvement of corrosion
resistance over carbon steel, especially when contact with condensing SO2/SO3 is intermittent or
of short duration (e.g. during repeated shut-downs).
3) In these circumstances, ordinary stainless steels (including high molybdenum stainless steel) have
little better corrosion resistance than carbon steel and are, therefore not recommended. If carbon
steel is used in chimneys subject to high chemical load, it will require protection by an appropriate
coating. For further guidance, see:
CICIND Chimney Coatings Manual.
9. DESIGN DETAILS
9.1 Connections
9.1.2.1 Shear
The shear stresses multiplied by the load factors shall not exceed the limit shear stress divided by resistance
factor = 1.1:
τu
τ* < (9.1)
1.1
The values of limit shear stress are given in Table 9.1.
The design shear stress τ* relates to the gross area or to the nett area, depending on whether the shear plane
is in the unthreaded or threaded part of the bolt.
The design bearing stress σ1* relates to the area obtained by multiplying the diameter d of the shank by the
thickness of the connected part. Regardless of any preload, the limit stress σ1,u is valid for edge distances
greater or equal 2d in the direction of stress.
9.1.2.3 Tension
The limit state is described:
σ t,u
σ*t < = 0.73 ⋅ σ u,B (9.3)
1.1
Note: The stresses given in Tables 9.2 and 9.3 are for ambient temperatures. For stresses at elevated
temperatures refer to the factors in column 2 of Table 6.2.
If the external loading results in a combination of tensile stress σ*t and shear stress τ* in the bolt, the
carrying capacity shall be checked for the condition:
2 2
τ* σ*t
+ < 1.0 (9.4)
τu σ t,u
This check is not necessary if:
2) in the net section, the stress shall not exceed 80% of the tensile strength σu
Regardless of the direction of stress, the two design stresses σ*w and σ*s for fillet welds shall be checked:
σ w,u
• in the throat section a-a: σ*w < = 0.455 ⋅ σ u,E
1.1
σs,u
• in the contact section s-s: σ*s < = 0.636 ⋅ f y
1.1
where σ u,E is the guaranteed minimum value of the tensile strength of the weld metal and fy the yield stress
of the parent material.
Table 9.4. ─ Limit stresses σw,u and σs,u for fillet welds in MPa
The yield stress, tensile strength, strain at failure and notch toughness of the weld metal shall exceed
minimum values for parent material, and, failing a specific agreement, shall be at least equal to those of
Fe 510. σw,u values given in Table 9.4 are valid for electrodes with properties of steel Fe 510.
page 40 CICIND Steel Chimney Model Code
If the fatigue load Zf is greater than the fatigue strength divided by 1.10, a joint with contact areas shall be
used, see ref. [22] and Figure 9.2. The pretension of the bolts should provide a sufficient force ZA to
prevent the fatigue in the bolt material:
w
ZA = 0.73 ⋅ σ u,b ⋅ A n ⋅ ≥ Zf (9.8)
a
It should be noted that the change of the type of connection to one with profiled contact areas may reduce
the damping ratio used in estimating along and across-wind response. The fitting of gaskets to the flanges
of structural shells is not permitted.
9.3.2 Grouting
After the chimney has been erected and plumbed (with the use of steel shims which remain in position) the
space between the base plate and concrete foundation must be filled with no shrink grout. The compressive
strength of the grout must be equal to or greater than the compressive strength of the concrete.
11. CONSTRUCTION
11.1 General
The following will be observed during shop and site construction as appropriate.
page 42 CICIND Steel Chimney Model Code
11.6 Straightness
Adjoining cylinder sections shall be welded together straight in the longitudinal direction to a tolerance of
+/-12mm per 10 m of shell length.
Flanges shall be welded to the structural shell within a perpendicular tolerance of ±0.5°.
13. OPENINGS
The width of a single opening shall not exceed two-thirds of the diameter of the structural shell of the
chimney.
Where large apertures are cut in the shell plates, as for gas inlets or inspection panels, a structural analysis
of the stresses shall be made and compensating material provided, as required, to ensure that the stresses
specified in this Model Code are not exceeded. As a result, it may be necessary to incorporate stiffeners
around the opening. When longitudinal stiffeners are used, their design shall include the effects of
CICIND Steel Chimney Model Code page 43
circumferential bending stresses in the shell, above and below the opening. Also they shall be long enough
to distribute stresses into the main area of the shell without overstress. (Note: this may generally be deemed
to be satisfied if the stiffeners project above and below the opening a distance at least 0.5 times the spacing
of the stiffeners.). The ends of the longitudinal stiffeners should be tapered in a radial direction (see cases
16.1-3 in Figure 8.4).
Additional horizontal stiffeners may be used to absorb the circumferential bending stresses. These
stiffeners may be attached between the longitudinal stiffeners, at the hole’s edge and at the end of the
longitudinal stiffeners.
A suggestion for stiffeners is given in the Commentaries for this Model Code.
Smaller apertures in the shell plates, not equipped with stiffeners, shall have the corners radiused to a
minimum of 10 t, where t is the thickness of the plate.
The effect of openings upon the chimney’s stiffness should be taken into account when determining the
chimney’s natural frequencies.