1995UGP3-V3 - User S - Guide - National - Building - Code - Part - 4 PDF

Copyright © NRC 1941 - 2019 World Rights Reserved © CNRC 1941-2019 Droits réservés pour tous pays
User’s Guide – NBC 1995

Structural Commentaries
(Part 4)
Issued by the
Canadian Commission on Building and Fire Codes
National Research Council of Canada
First Edition 1996
ISBN 0-660-16275-X
© National Research Council of Canada 1996

Ottawa
World Rights Reserved
NRCC 38826
Printed in Canada
Third Printing
Includes First Errata of July 1996 and
Second Revisions of April 2002
Aussi disponible en français :

Guide de l’utilisateur — CNB 1995 Commentaires sur le calcul des structures (Partie 4)
Table of Contents
Introduction ...............................................v
Commentary A Serviceability Criteria for

Deflections and Vibrations...1
Commentary B Wind Loads ........................... 9
Commentary C Structural Integrity.............43
Commentary D Effects of Deformations in

Building Components ......... 45
Commentary E Load Combinations.............49
Commentary F Limit States Design ............ 51
Commentary G Tributary Area.....................55
Commentary H Snow Loads ........................ 57
Commentary I Rain Loads .......................... 73
Commentary J Effects of Earthquakes.......75
Commentary K Application of NBC Part 4 for

the Structural Evaluation and
Upgrading of Existing
Buildings ............................. 93
Commentary L Foundations ...................... 103
Commentary M Structural Integrity of

Firewalls ........................... 131
Appendix A List of Referenced

Standards..........................135
iii
iv
Introduction
The purpose of these Commentaries is to make the following members: R. Halsall (Chair),
available to the designer detailed design informa- R.H. DeVall, V.C. Fenton, J. Khan, D. Mitchell and
tion which will assist in the use of the National E.Y. Uzumeri.
Building Code. The Commentaries are provided as
background information and, in some cases, as sug- Commentary L (Foundations) was prepared with
gested approaches to certain design questions, but the assistance of a task group appointed by the
not as mandatory requirements. Standing Committee on Structural Design; the task
group consisted of the following members:
Because the information provided in these Com- V. Milligan (Chair), L. Brzezinski, D. Klajnerman,
mentaries cannot cover all conditions or types of W.E. Lardner and E.Y. Uzumeri. The Task Group on
structures that occur in practice, and also because Limit States Design of Foundations assisted in the
new information may become available in the fu- revision of the Commentary to include limit states
ture, the designer should try to obtain the latest and design of foundations. This task group consisted of
most appropriate design information available. For L.D. Baikie (Chair), D. Becker, L. Eisenstein,
unusual types of structures, specialized information D.J.L. Kennedy, K.T. Law and A. Malhotra.
such as theoretical studies, model tests or wind
tunnel experiments may be required to provide ade- Commentary M (Structural Integrity of Firewalls)
quate design values. has been developed to provide guidance to the new
requirements in Sentence 4.1.10.4.(1) of the National
The commentaries were prepared with the assis- Building Code for the design of firewalls.
tance of the following:
D.E. Allen W.R. Schriever

D.A. Lutes A.G. Davenport
J.H. Rainer W.G. Plewes
W.A. Dalgliesh D.A. Taylor
J.G. MacGregor D.J.L. Kennedy
Commentary J (Effects of Earthquakes) was pre-

pared with the assistance of the Canadian National
Committee on Earthquake Engineering, which ad-
vises the Canadian Commission on Building and
Fire Codes on provisions for earthquake engineer-
ing in the Code.
Commentary K (Glass Design) has been deleted.

It is no longer considered necessary since it is out of
date and up to two-thirds of the material is covered
in the CGSB Standard 12.20-M89 “Structural Design
of Glass for Buildings,” referenced in Part 4 of the
NBC.
The new Commentary K (Application of NBC

Part 4 for the Structural Evaluation and Upgrading
of Existing Buildings) was prepared with the assis-
tance of a task group appointed by the Standing
Committee on Structural Design and consisting of
v
vi
Commentary A
Serviceability Criteria for Deflections
and Vibrations
1. The advent of stronger materials, lighter, and ponding. Cracks, besides being unsightly, may
more rigid cladding, smaller damping, longer spans transmit unwanted sound through partitions, or wa-
and more accurate strength calculations taking ac- ter and cold air through exterior surfaces, and thus
count of the interaction of components, means that promote corrosion. Control of cracking in structural
excessive deflections and vibrations now have a concrete is covered separately in CAN3-A23.3-M94,
greater influence in structural design than before. “Design of Concrete Structures for Buildings.”
Excessive deflections and vibrations are usually con-
trolled in codes by limiting the member deflection 3. A number of alternative design solutions can
under specified load to some ratio of the span (L), prevent problems caused by excessive deflection.
for example, L:360 (for cantilevers, L may be taken Partition cracking, for example, can be avoided ei-
as twice the length of the cantilever). Table A-1 ther by making the supporting structure stiff
summarizes deflection criteria in this form in vari- enough or by providing flexible joints in the parti-
ous standards pertinent to the National Building tions. Similarly, to avoid cracking, plastered ceilings
Code of Canada 1995. These deflection criteria de- should be hung from the floor beams, not rigidly at-
pend on the types of construction and materials and tached to them.
on the conditions of use. As an aid to the designer,
the problems associated with excessive deflection 4. The deflection criteria in Table A-1 apply to
and excessive vibration are briefly discussed and conventional forms of construction under conven-
references are given. tional conditions of use. The most severe deflection
requirement, 1:480, for members supporting plas-
Deflections tered ceilings or partitions,(1) may not be sufficient
to prevent cracking of plaster or rigid partitions.(3)
2. Excessive structural deflections can create a For new or unusual cases, more detailed deflection
variety of problems: cracks or crushing in non- criteria are suggested in Reference (2); case histories
structural components such as partitions, lack of fit of damage due to excessive deflections (including
for doors, walls out of plumb or eccentricity of also differential settlement and temperature move-
loading caused by rotation, unsightly droopiness ments) are given in References (4) to (7).
Table A-1
Summary of Maximum Deflection/Span Ratios in NBC 1995 and Pertinent CSA Standards(1)
CAN/CSA-O86.1-94 CAN3-A23.3-94 CAN/CSA-S16.1-94

Wood NBC 1995 Part 9
Concrete Structural Steel
Roof or floor members supporting — 1:480(2) or 1:240(2) 1:360 1:360
plastered ceilings, partitions, etc.
Floor members not supporting 1:360(3) or 1:180(3) 1:360(4) 1:300(5) 1:240(6) or 1:360
plastered ceilings, partitions, etc.
Roof members not supporting 1:360(3) or 1:180(3) 1:180(4) 1:180(7) or 1:240(7) 1:180(8) or 1:240
plastered ceilings, etc.
Wall members 1:180(3) — — —
Notes to Table A-1:

(1) Deflection under live load only unless otherwise noted.
User’s Guide – NBC 1995 Structural Commentaries (Part 4) 1

Commentary A
Table A-1 (Continued)
(2) Deflection which occurs after attachment of non-structural elements, including creep deflection due to sustained load plus immediate deflection
due to additional live load. The lower figure applies when non-structural elements are not likely to be damaged by large deflections.
(3) Modulus used for calculations based on short term test. 1:360 applies to deflection under sustained load.
(4) Immediate live-load deflection. Includes a warning on ponding for roof members.
(5) Includes a warning clause on vibrations.

(6) For bedrooms only.
(7) 1:180 applies to sheet metal or elastic membrane roof cover and 1:240 to asphaltic built-up roofs. Includes a warning clause on ponding.
(8) If there is no ceiling.
Vibrations Floor Vibrations in Assembly

Occupancies
5. Two types of vibration problems arise in
building construction: continuous vibrations and 8. NBC Sentence 4.1.10.5.(1) requires an investi-
transient vibrations. Continuous vibrations arise gation by means of a dynamic analysis for floor
due to the periodic forces of machinery or certain structures (including footbridges) supporting assem-
human activities such as dancing; these vibrations bly occupancies when the fundamental vibration
can be considerably amplified when the periodic frequency is less than 6 Hz. This requirement was
forces are synchronized with a building frequency – introduced because of problems with long-span
a condition called resonance. Transient vibrations floor structures used for rhythmic activities.(12-17) The
are caused by footsteps or other impact and decay following provides guidance for the designer in car-
at a rate which depends on the available damping. rying out a dynamic analysis for such cases, and
also suggests criteria to limit floor vibrations during
6. Transient vibrations in floor systems due to rhythmic activities to levels acceptable for human
foot impact may cause discomfort or annoyance to occupancy.
the occupants as a result of, for instance, china rat-
tling. In Table A-1 the deflection criteria of L:360 for
wood floors(1) and L:320 for steel floors which do
not support brittle materials attempt to control such
vibration effects. These criteria apply only to con-
ventional floors with spans less than approximately
6 m and frequencies greater than about 10 Hz. They
do not apply to long span floors, particularly for
those without partitions, or for floors for special
purposes; Reference (8) contains further information
and criteria on these cases. Reference (9) contains
criteria for footbridges. References (1) and (10) con-
tain further information for light residential floors
with wood decks.
7. The undesirable effects of continuous

vibrations caused by machines can be minimized by
special design provisions,(11) such as locating
machinery away from sensitive occupancies,
vibration isolation or alteration of the frequency of
the structure.
2 User’s Guide – NBC 1995 Structural Commentaries (Part 4)

Commentary A
Overloading and Fatigue dancing, concerts

9. Dancing, foot stamping, jumping exercises αWp
and marching are rhythmic activities that create pe- Wt
riodic forces with rhythmic frequency in the range 1 Wp
to 4 Hz; for rhythmic events involving a group of
people, the most critical range is 1.5 to 3 Hz. Typical

loading cases are shown in Figure A-1. For many
Load
rhythmic activities, such as dancing, the periodic
forces can be approximated by a sinusoidal dynamic
load causing vibration at the rhythmic frequency. In
the case of jumping exercises, however, the periodic
forces shown in Figure A-1 can also create signifi-
cant sinusoidal load at double the rhythmic
frequency, and some sinusoidal load at triple the Time
rhythmic frequency. The dynamic load for any sinu-
soidal component can be represented by wpsin2ft,
where f is the forcing frequency, wp is the weight of
participants in kPa, t is time and is a dynamic
jumping exercises
load factor which depends on the activity and, for
jumping, the harmonic multiple of the rhythmic fre-
quency. Table A-2 gives estimated values of forcing
frequencies and dynamic load based on an estima-
Wt Wp
tion of density of participants and dynamic load
factor, , for typical rhythmic events.(15)(16) If the fun-
damental natural frequency of the floor structure, fo,
Load
is large compared to the forcing frequency, f, the dy-

namic load has the same effect as a static load of the
same magnitude, but if the structural frequency ap-
proaches the forcing frequency, the dynamic effect
increases with each cycle of vibration to a maximum
(Figure A-2) whose ratio to the static effect is given
by(11) Time
Figure A-1
Load during rhythmic event
(1)
Table A-2
Estimated Loading During Rhythmic Events
Activity Forcing Frequency f, Weight of Dynamic Load Dynamic Load wp,

Hz Participants(1) wp, kPa Factor,(2) kPa
Dancing 1.5 – 3 0.6 (2.5 m2/couple) 0.5 0.3
Lively concert or sports event 1.5 – 3 1.5 (0.5 m2/person) 0.25 0.4
Jumping exercises
First harmonic 2 – 2.75 0.2 (3.5 m2/person) 1.5 0.3
Second harmonic 4 – 5.5 0.2 (3.5 m2/person) 0.6 0.12
Third harmonic 6 – 8.25 0.2 (3.5 m2/person) 0.1 0.02
Notes to Table A-2:

(1) Density of participants represents maxima for commonly encountered conditions. For special events the density of participants can be greater.
(2) Values of based on commonly encountered events with a minimum of 20 participants. Values of should be increased for well coordinated
events with fewer than 20.

Commentary A
8% g
6s
Figure A-2
Resonance during rock and roll (precast stands, fo = 2.6 Hz)
For a floor with many people on it, the damping where

ratio, , is about 0.04 for a solid concrete floor, 0.06
for a concrete and steel floor and 0.12 for a wood 1 = (3)
floor, and about half these values with few people.
Damping ratios vary from these suggested values, 1 B = the elastic deflection of the beam due to
depending on the influence of non-structural com- bending and shear,
ponents such as partitions. The dynamic response
factor, , is shown in Figure A-3. 1 G = the elastic deflection of the girder due to
bending and shear, and
5 1 S = the elastic shortening of the column due

to axial strain.
β=0
β = 0.06 Each deflection is the elastic component resulting
4
β = 0.12 from the total weight supported by the member, in-
cluding people, and is relative to its supports. Both
3
supports are considered and the most flexible one
used. For cantilever and two-way slabs, the factor
0.77 is replaced by 0.65. In the case of beams and
ρ
dynamic effect girders continuous over supports, the elastic deflec-

2
1 1
tion, B or G, should be determined by assuming
that adjacent spans deflect in opposite directions
static effect with no change in slope over the supports and that
1
the weight supported by each span always acts in
the direction of deflection.
0 11. The total structural effect of a rhythmic ac-
0 0.5 1.0 1.5 2.0 2.5 3.0
tivity can, therefore, be represented by the static
fo/f effect of the load wt + wp, where wt is the total
weight supported. In the case of jumping forces, the
Figure A-3 dynamic loading component wp is replaced by
Dynamic response factor, Equation (1) , where i and i are the dynamic load and
response factors for each harmonic multiple, i, of
the rhythmic frequency. Overloading occurs if the
total load, including static and dynamic compo-
nents, is greater than the total specified load that
10. The fundamental structural frequency, fo,
the structure is allowed to carry. Potential for fa-
should be determined from the dynamic properties
tigue damage can be assessed by estimating the
of the floor structure, taking into account the flexi-
stress range corresponding to 2iiwp and the total
bility of supports. An approximate determination
number of cycles of dynamic load iwp expected
for simply-supported beams on girders supported
during the life of the structure for each harmonic.
by columns is obtained from:
Human Reaction
12. Floor vibration is much more likely to an-
noy people than to cause overloading or fatigue. An
(2) acceptable level of vertical vibration depends very

Commentary A
strongly on the activity of the people who feel the rhythmic event become very large, usually greater
vibration. People in offices or residences become than the recommended criterion. Generally, there-
annoyed when accelerations from continuous vibra- fore, the fundamental natural frequency of the floor
tion exceed approximately 0.5% gravity, whereas structure should be greater than the highest signifi-
people participating in rhythmic activities will ac- cant harmonic forcing frequency. The following
cept approximately 5% gravity. When a floor criterion is recommended(15)
structure is shared with a more sensitive occupancy,

then the limit should be based on that occupancy.
People such as diners who share a floor structure
with dancing, or weight lifters who share a floor
structure with aerobics will accept approximately (6)
2% gravity. Other factors besides occupancy affect
the acceptability of vibration, in particular the re-
moteness of the source of vibration from the people where
affected. For this reason, Table A-3 shows a range of
acceleration limits for each occupancy. The lower f = forcing frequency (i times the jumping
value is generally recommended for design. frequency for jumping exercises),
ao/g = acceleration limit,

Table A-3
Recommended Acceleration Limits for Vibrations due to K = 1.3, except for jumping exercises, where
Rhythmic Activities K = 2.0 is recommended.(16)
Acceleration Limit, 15. Table A-4 contains examples of the applica-

Occupancies Affected by the Vibration
% gravity tion of Equation (6) to typical floor structures using
the acceleration limits recommended above. If the
Office and residential 0.4 – 0.7
rhythmic activity is restricted to a portion of the
Dining and weightlifting 1.5 – 2.5 floor span, then an appropriate reduction can be
Rhythmic activity only 4–7 made to wp, the weight of participants effectively
acting on the floor span.(16) Similarly, extra mass
(other than the floor) supported by the vibrating
13. The maximum acceleration of a floor struc- structure, including the supports, can be taken into
ture during a rhythmic event can be determined account in determining the total weight, wt.(16)
from(15)
16. Measures to avoid or correct unacceptable
vibration include:
(a) provision of sufficient stiffness, Equation (6),
(4) (b) relocation of the rhythmic activity or the sensi-
tive occupancy,
(c) prevention of transmission of floor vibration to
where the approximation is valid away from the sensitive occupancies, for example, by placing
resonance frequency f = fo. The symbols are defined or altering partitions,
in Paragraph 9. In the case of jumping, the maxi- (d) increasing the damping sufficiently to reduce
mum acceleration can be determined from(16) resonance response, or
(e) increasing the span to where the fundamental
natural frequency becomes less than a signifi-
(5)
cant rhythmic frequency; this would require a
special investigation.
where ai is the maximum acceleration for the ith
harmonic loading component, which is determined Case histories of problems are described in
from Equation (4) by setting f equal to i times the References (14) and (17), including a case where un-
jumping frequency. acceptable aerobics vibration in a tall office building
occurred due to vertical spring action of the
14. If a natural frequency of the floor structure columns. If an existing floor is intended for a rhyth-
corresponds to a harmonic forcing frequency, reso- mic activity, a performance test should be carried
nance occurs and the accelerations during a out before alterations are made.

Commentary A
Table A-4
Application of Structural Criteria for Human Reaction
Minimum Structural
Activity and Construction Forcing Frequency f, Effective Weight of

Total Weight wt, kPa Frequency fo, Hz
Hz Participants wp, kPa
Equation (6)
Dancing and Dining ao/g = 0.02
Solid concrete 5 kPa 3 0.6 5.6 6.4
Steel joist 2.5 kPa 3 0.6 3.1 8.1
Wood 0.7 kPa 3 0.6 1.3 12.0
Lively Concert or Sports Events ao/g = 0.05
Solid concrete 5 kPa 3 1.5 6.5 4.8
Steel joist 2.5 kPa 3 1.5 4.0 5.7
Wood 0.7 kPa 3 1.5 2.2 7.2
Jumping Exercises only ao/g = 0.06
Solid concrete 5 kPa 8.25(1) 0.2 5.2 8.8(2)
Steel joist 2.5 kPa 8.25(1) 0.2 2.7 9.2
Wood 0.7 kPa 5.5(1) 0.2 0.9 12.8
Jumping Exercises Shared with Weight Lifting ao/g = 0.02
Solid concrete 5 kPa 8.25(1) 0.12 5.12 9.2
Steel joist 2.5 kPa 5.5(1) 0.12 2.62 10.6
Wood 0.7 kPa 5.5(1) 0.12 0.82 17.2
Notes to Table A-4:

(1) Equation (6) is applied to three harmonics (i.e., f = 2.75 Hz, 5.5 Hz and 8.25 Hz) and the governing harmonic is used.
(2) This can be reduced if, according to Equation (4), damping times mass is sufficient to reduce third harmonic resonance to an acceptable level.
References (6) A.W. Skempton and D.H. MacDonald, The

Allowable Settlements of Buildings. Proc., In-
(1) W.A. Russell, Deflection Characteristics of stitution of Civil Engineers, Vol. 5, Part III,
Residential Wood-Joist Floor Systems. Hous- 1956, p. 727.
ing and Home Finance Agency, Housing (7) F.R. Khan and M. Fintel, Effects of Column
Research Paper 30, Washington, D.C., April Exposure in Tall Structures – Design Consid-
1954. erations and Field Observations of Buildings.
(2) Allowable Deflections. Subcommittee 1, ACI Journal Am. Concrete Inst., Vol. 65, No. 2,
Committee 435. Journal Am. Concrete Inst., February 1968, p. 99.
Vol. 65, No. 6, June 1968, p. 433. (8) CSA Standard CAN3-S16.1-M89. Steel Struc-
(3) W.G. Plewes and G.K. Garden, Deflections of tures for Buildings – Limit States Design.
Horizontal Structural Members. Canadian Appendix Guide on Floor Vibrations.
Building Digest No. 54, Division of Building (9) Ontario Highway Bridge Design Code 1984.
Research, National Research Council of Ontario Ministry of Transportation and Com-
Canada, Ottawa, June 1964. munications.
(4) H. Mayer and H. Rüsch, Bauschäden als Folge (10) D.M. Onysko, Performance of Wood-Joist
der Durchbiegung von Stahlbeton-Bauteilen Floor Systems. Forest Products Laboratory In-
(Building Damage Caused by Deflection of formation Report OP-X-24, Canadian Forestry
Reinforced Concrete Building Components). Service, Department of the Environment,
Deutscher Ausschuss für Stahlbeton, Heft 193, Ottawa, January 1970.
Berlin 1967. National Research Council of (11) W.J. Smith, Vibrations of Structures; Applica-
Canada Technical Translation TT1412, 1970. tions in Civil Engineering Design. Chapman
(5) O. Pfeffermann, Les Fissures dans les Con- and Hall, London, 1988.
structions Conséquences de Phénomènes (12) Pop Concert Shock for Loading Code. New
Physiques Naturels. Annales de l’Institut Civil Engineer International, May 1981, p. 18.
Technique du Bâtiment et des Travaux
Publics, No. 250, October 1968.

Commentary A
(13) G. Pernica, Dynamic Live Loads at a Rock

Concert. Can. J. Civ. Eng., June 1983, pp. 185-
191.
(14) H. Bachmann and W. Ammann, Vibrations in
Structures Induced by Man and Machines.
Structural Engineering Document 3e. Interna-
tional Association for Bridge and Structural

Engineering, Zurich, 1987.
(15) D.E. Allen, J.H. Rainer and G. Pernica, Vibra-
tion Criteria for Assembly Occupancies. Can.
J. Civ. Eng., Vol. 12, No. 3, September, 1985.
(16) D.E. Allen, Floor Vibrations from Aerobics.
Can. J. Civ. Eng., October 1990.
(17) D.E. Allen, Vibrations from Human Activities.
Concrete International – Design and
Construction, American Concrete Institute,
June 1990.

8
Commentary B
Wind Loads
1. Three different approaches to the problem of involving (a) the intensity of wind turbulence for the
determining design wind loads on buildings are site as a function of height and of the surface rough-
mentioned in Subsection 4.1.8., “Live Loads Due to ness of the surrounding terrain, and (b) properties
Wind” of the 1995 edition of the National Building of the building such as height, width, natural fre-
Code.(1) quency of vibration and damping. The end-product
of the calculations is the gust effect factor, Cg, which
2. The first approach, the “simple procedure,” is is multiplied by the reference wind pressure, q, the
appropriate for use with the majority of wind load- exposure factor, Ce, and the pressure coefficient, Cp,
ing applications, including the structure and to give that static design pressure which is expected
cladding of low and medium rise buildings and the to produce the same peak load effect as the dy-
cladding design of high rise buildings. These are sit- namic resonant response to the actual turbulent
uations where the structure is relatively rigid. Thus, wind. The format of the simple procedure in the
dynamic actions of the wind do not require detailed NBC has been arranged to permit an easy transition
knowledge of the dynamic properties of the build- to this more detailed consideration of wind effects.
ings and can be dealt with by equivalent static
loads. Numerical values for all the factors involved Reference Wind Speed, V̄, and
are provided in the NBC except for pressure coeffi- Pressure, q
cients given in this Commentary.
6. In both the simple and detailed procedures
3. The two other approaches to wind load anal- the reference wind speed, V̄, is determined by ex-
ysis are referred to in Article 4.1.8.2. of the 1995 treme value analysis of meteorological observations
NBC, and are required whenever the building is of hourly mean wind speeds, taken at sites (usually
likely to be susceptible to wind-induced vibration. airports) chosen in most cases to be representative
This may be true, for example, of tall and slender of a height of 10 m in an open exposure. The refer-
structures or doubly cantilevered canopies for which ence wind pressure, q, is determined from V̄ by the
wind loading plays a major role in the structural de- following equation:
sign. Here the designer is required to use either (a)
special wind tunnel tests or other experimental
methods, or (b) a dynamic approach to the action of (1)
wind gusts to be called the “detailed procedure.”
4. Wind tunnel testing is appropriate when 7. The factor C depends on the atmospheric
more exact definition of dynamic response is pressure and the air temperature. The atmospheric
needed and for determining exterior pressure coeffi- pressure in turn is influenced mainly by elevation
cients for cladding design on buildings whose above sea level, but also varies somewhat in accor-
geometry deviates markedly from more common dance with changes in the weather.
shapes for which information is already available.
Information on modern wind tunnel techniques can 8. The following value of C is chosen to repre-
be found in References (2), (3) and (4). sent Canadian conditions:
5. The “detailed procedure” is intended primar- if V̄ is in kilometres per hour, C = 50 10-6
ily for determining the overall wind loading and
amplified resonant response of tall buildings and if V̄ is in metres per second, C = 650 10-6.
slender structures, but not cladding and compo-
nents.(5) It consists of a series of calculations 9. Appendix C of the 1995 NBC contains a de-
scription of the procedures followed in obtaining
In Appendix C of the National Building Code 1995, the veloc- the reference wind pressures, q, for three different
ity pressure, P, and the design wind speed, V, are meterological levels of probability of being exceeded per year (1
terms, equivalent to the reference wind pressure, q, and the refer- in 10, 1 in 30 and 1 in 100), the values commonly re-
ence wind speed, V̄, which are engineering terms used in this ferred to as having return periods of 10, 30 and 100
Commentary. years. These values of q are tabulated for many

Commentary B
Canadian locations along with other climatic design Exposure A (open or standard exposure): open
data. A reference giving more detail on the choice of level terrain with only scattered buildings, trees or
the conversion factor, C, from wind speed to pres- other obstructions, open water or shorelines thereof.
sure and a table for converting from pressure in This is the exposure on which the reference wind
kilonewtons per square metre to speed in metres speeds are based.
per second are also supplied in Appendix C.
Exposure Factor, Ce
Simple Procedure (2)
10. The exposure factor, Ce, reflects changes in

wind speed and height, and also the effects of varia- Exposure B: suburban and urban areas, wooded
tions in the surrounding terrain and topography. terrain or centres of large towns.
Hills can significantly amplify the wind speeds near
the ground and this should be reflected in the expo-
sure factor. Representative values for use with either
the simple or detailed procedure are presented in (3)
Paragraph 18.
11. For the simple procedure, Ce is based on the Exposure C: centres of large cities with heavy
1/5 power law which is appropriate for wind gust concentrations of tall buildings. At least 50% of the
pressure in open terrain (1/10 power law for wind buildings should exceed 4 storeys.
gust speeds). The wind gust referred to lasts about
3 to 5 s and represents a “parcel” of wind which is
assumed effective over the whole of most ordinary
buildings.
(4)
12. The reference height which is used in deter-
mining the exposure factor is related to the manner
In Equations (2) to (4), Z is the height above
in which the pressure coefficient, Cp, is defined. In ground in metres.
this Commentary, the reference height applicable to
low rise structures is the mean height of the roof or
14. Exposure B or C should not be used unless
6 m, whichever is greater. The eave height may be
the appropriate terrain roughness persists in the up-
substituted for the mean height if the slope of the
wind direction for at least 1.5 km, and the exposure
roof is less than 10. For tall buildings, the reference
factor should be varied according to the terrain if
height for pressures on the windward face corre-
the roughness differs from one direction to another.
sponds to the actual height above ground; for
suctions anywhere on the leeward face of tall build-
Use of Exposure Factors
ings, the reference height is half the height of the
structure. Where required in this Commentary, the
15. Exposure factors can be calculated from
definition of the reference height is given along with
Equations (2) to (4) or obtained directly from the
that of Cp. In instances where the reference height is
graphs in Figure B-1. The exposure factor is needed
not specified, it should be taken as equal to the
in three different capacities in the detailed proce-
height above ground of the element considered.
dure. Firstly, the square root of CeH, the value of Ce
Detailed Procedure at the top of the building, H, is needed to determine
the hourly mean wind speed at the top of the struc-
13. For the detailed procedure, the exposure fac- ture being designed, VH
tor, Ce, is based on the mean wind speed profile,
which varies considerably depending on the general
roughness of the terrain over which the wind has (5)
been blowing before it reaches the building. This
dependence on terrain is much more significant
than is the case for the gust speed profile (variation The reference wind speed, V̄, can be obtained
of gust speed with height) and hence three cate- from the reference wind pressure and the conversion
gories have been established as follows: table in Appendix C or by applying Equation (1).

Commentary B
400
sponds to that for the height, Z, to the point in
300
question and therefore increases with height in
200
accordance with Equations (2), (3) or (4) as is appro-
priate. For the leeward face, Ce is evaluated at half
the height, H, of the building.
100
80
Speed-Up over Hills and Escarpments
60
50
18. Buildings on a hill (with a maximum slope
40
exposure C B A > 1 in 10), particularly near a crest, may be subject
Height above ground, m
30
to significantly higher wind speeds than buildings
on level ground. The exposure factor at height z
20 above the local ground elevation is then equal to
that over flat terrain multiplied by a factor (1 +
1 1
S(z))2, where S(z) is the “speed-up factor” for
10 the mean wind speed. This can be applied in both
8 the simple and detailed procedures. This effect is il-
6 lustrated in Figure B-2. Near the crest, and within a
5 distance |x| < kL, the exposure factor is modified to
4
2
(6)
1
0.1 0.2 0.3 0.4 0.6 0.8 1 2 3 4 5 6 8 10 where
Exposure factor, Ce
Ce(z) = the exposure factor over flat terrain

Figure B-1 given in Equation (2), (3) or (4),
Exposure factor as a function of terrain roughness and
height above ground = the corresponding modified value for
use on the hill,
1Smax = the relative speed-up factor at the crest

16. Secondly, CeH appears in Equation (8) used near the surface, and
for calculating the gust effect factor, Cs. Here again,
CeH is the value of Ce evaluated at the top of the = a decay coefficient for the decrease in
structure. speed-up with height.
17. Thirdly, Ce is used in the calculation of pres- 1

The values of and Smax depend on the shape
sures for the windward and leeward faces of tall, and steepness of the hill. Representative values for
slender buildings. For the windward face, Ce corre- these parameters are given in Table B-1.
Table B-1
Parameters for Maximum Speed-Up Over Low Hills
k
Hill Shape 1S max
(1)
x<0 x>0
2-dimensional ridges (or valleys with 2.2 H/L 3 1.5 1.5
H negative)
2-dimensional escarpments 1.3 H/L 2.5 1.5 4
3-dimensional axisymmetrical hills 1.6 H/L 4 1.5 1.5
Notes to Table B-1:

(1) For H/L > 0.5, assume in the formulae that H/L = 0.5 and substitute 2H for L in Equation (6).

Commentary B
effect of the loading. The dynamic response in-

“speed up” cludes the action of
(a) random wind gusts acting for short durations
V(z) ∆ SV(z) over all or part of the structure,
z
(b) fluctuating pressures induced by the wake of
x
H the structure, including “vortex shedding
2 forces,” and
V(z) H
z L H (c) fluctuating forces induced by the motion of the
2
structure itself through the wind.
Hill
These forces act on the external surfaces of the
structure as a whole or on cladding components
and may also affect internal surfaces. They may act
longitudinally, laterally or torsionally and further
they may be amplified by resonance of the structure
V(z) ∆ SV(z)
z at one or other of its natural frequencies.
x
All structures are affected to some degree by
V(z) these forces. The total response may be considered
z L
as a superposition of a “background component,”
Escarpment
which acts quasi-statically without any structural
dynamic magnification, and a “resonant” compo-
Figure B-2 nent due to excitation close to a natural frequency.
Definitions for wind speed-up over hills and escarpments For the majority of structures, the resonant compo-
nent is small and the dynamic factor can be
simplified by considering the background compo-
The definitions of the hill height H and length L, nent only and treated using normal static methods.
shown in Figure B-2, are as follows: H is the height For structures that are particularly tall, long, slen-
of the hill or the difference in elevation between the der, light-weight, flexible or lightly damped, the
crest of the hill and that of the terrain surrounding resonant component may be dominant.
the hill upstream; L is the distance upwind of the
crest to where the ground elevation is half the The majority of structures can be treated using
height of the hill. The maximum slope of the hill is the “simple procedure.”
roughly H/2L for rounded hill shapes. In these ex-
pressions, it is assumed that the wind approaches A general expression for the maximum or peak
the hill along the direction of maximum slope, the loading effect, denoted Wp, is
direction giving the greatest speed-up near the crest.
(7)
These formulae suggest that hills with slopes less
than 1 in 10 are unlikely to produce significant
speed-up of the wind. A more extended discussion where
of this question and other simplied models for
three-dimensional hills are given in Reference (6). = the mean loading effect,
Background material may be found in References
(7) and (8). Wind tunnel tests and computational
= the “root-mean square” loading effect,
and
methods may be used to obtain design information
in other cases. gp = a statistical peak factor for the loading
effect.
The speed-up principally affects the mean wind
speed and not the amplitude of the turbulent fluctu- According to this expression a dynamic factor,
ations. This leads to a correction in the gust effect equal to the ratio of the peak loading to the mean
factor referred to below. loading, can be identified as
Dynamic Response and Gust Effect
Factor, Cg (8)
General
The form of the fluctuating wind loading effect,
19. In this section, procedures are recommended , varies with the excitation, whether due to gusts,
for determining the dynamic response and “gust ef- wake pressures or motion induced forces. For a
fect factor” referred to in Sentence 4.1.8.1.(6) of the large class of smaller structures, only the added
NBC. This factor, denoted Cg, is defined as the ratio loading due to gusts must be dealt with and simpli-
of the maximum effect of the loading to the mean fied methods are adequate.

Commentary B
Simple Procedure 22. Suggested values for must be based

mainly on experiments on real structures. Expressed
20. For small structures or structures and as a fraction of critical damping, values commonly
components having a relatively high rigidity, a sim- used in the design of buildings with steel frames
plified set of dynamic gust factors is and concrete frames are 0.01 and 0.02, respectively.
Masts and stacks, on the other hand, may have
Cg = 2.5 for cladding elements and small much lower inherent or structural damping. Aero-
structural components, dynamic damping in the along-wind direction
= 2.0 for structural systems, including an- becomes significant at high wind speeds, but plays
chorages to foundations. no useful role in limiting cross-wind motion due to
vortex shedding. Spread footings on soft or medium
For some structures, peak pressure coefficients stiff soil provide additional damping in comparison
have been determined directly from wind tunnel to piled foundations or spread footings on stiff soil
tests, and composite values of (CpCg) are obtained and rock. Measured damping values from more
incorporating the aerodynamic shape factors. than 20 stacks are tabulated in Reference (9) and re-
sults from 5 more stacks are given in Reference (10).
Detailed Procedure The logarithmic decrement mentioned therein is 2
times the critical damping ratio. Sachs(9) concludes
21. In the detailed procedure, the value of / by stating a range of 0.0016 to 0.0080 for for the
can be expressed total damping of closed circular, unlined welded
steel stacks, and suggests that the minimum value
be used in design. Corresponding ranges for lined
welded steel stacks and for unlined reinforced con-
(9) crete stacks are given as 0.0048 to 0.0095 and 0.0095
to 0.0191, respectively.
where 23. The peak factor, gp, in Equation (7) gives the
number of standard deviations by which the peak
K = a factor related to the surface roughness load effect is expected to exceed the mean load ef-
coefficient of the terrain, fect, and is given in Figure B-6 as a function of the
= 0.08 for Exposure A, average fluctuation rate. The average fluctuation
= 0.10 for Exposure B, rate, v, can be estimated as follows:
= 0.14 for Exposure C,
CeH = exposure factor at the top of the building,

H, evaluated according to Paragraph 13 v
(10)
or Figure B-1. Over hills the value
should be used (see Equation (6)).
using the variables defined for Equation (9).
B = background turbulence factor obtained
from Figure B-3 as a function of W/H,
W = width of windward face of the building,
H = height of windward face of the building,
s = size reduction factor obtained from

Figure B-4 as a function of W/H and the
reduced frequency noH/VH,
no = natural frequency of vibration, Hz,
VH = mean wind speed (m/s) at the top of

structure, H,
F = gust energy ratio at the natural fre-

quency of the structure obtained from
Figure B-5 as a function of the wave
number, no/VH , and
= critical damping ratio.

Commentary B
400
300
200
100
80
Height of structure, m
60
50
40
W/H
30
0
0.
0.
0.
2
3
1
20
0
3. 1. 0 .5
0 2. 5 1. .7
0 0
5.0
10
8
6
5
4
914/H
2
1 1 x
B = 4/3 xH
1+ 1 + xW (1 + x 2) 4/3 dx
457 122
0
1
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
Background turbulence factor, B
Figure B-3
Background turbulence factor as a function of width and height of structure
5.0
4.0
3.0
VH
Reduced frequency, no H
2.0
W/H = 2.0 1.0 0.5 0.2 0.1 0

1.0
0.7
0.5
0.3
s= π
1 1
3 8n0H 10n0W
0.2 1+ 1+
3VH VH
0.1
0.001 0.002 0.004 0.007 0.01 0.02 0.03 0.05 0.1 0.2 0.3 0.5 1.0
Size reduction factor, s
Figure B-4
Size reduction factor as a function of width, height and reduced frequency of structure

Commentary B
1.00
0.80
0.60
0.40
Gust energy ratio = F
0.30
0.20
0.10
0.08
0.06
2
0.04 x0
F=
0.03 2
(1 + x 0 ) 4/3
0.02
x = (1220 n0/VH)
0
0.01
0.0001 0.001 0.01 0.1 1.0
Wave number, waves/m, n0/VH
Figure B-5
Gust energy ratio as a function of wave number
Explanatory Notes Regarding / and gp shape of a rather broad hump (Figure B-5). The ef-
fect of the aerodynamic admittance is to reduce the
24. The response of a tall, slender building to a ordinates of the curve to the right of the hump more
randomly fluctuating force can be evaluated rather and more as the frequency increases. This is partly a
simply by treating it as a rigid, spring-mounted can- reflection of the reduced effectiveness of small gusts
tilever whose dynamic properties are specified by a in loading a large area. The effect of the dynamic
single natural frequency and an appropriate damp- load magnification factor or mechanical admittance
ing value. The variance of the output quantity or is to create a new peak or hump centred at the nat-
loading effect is the area under the spectrum of the ural frequency of the structure, usually well to the
input quantity (the forcing function) after it has right of the broad peak, which represents the maxi-
been multiplied by the transfer function. The trans- mum density of input power of the wind.
fer function is the square of the well-known
dynamic load magnification factor for a one-degree-
27. The area under the loading effect spectrum,
of-freedom oscillating mechanical system. the square root of which is the coefficient of varia-
25. In the case of wind as the random input, the tion /, is taken as the sum of two components:
spectrum of the wind speed must first be multiplied the area under the broad hump, which must be in-
tegrated numerically for each structure, and the area
by another transfer function called the “aerody-
namic admittance function,” which in effect under the resonance peak, for which a single ana-
describes how the turbulence in the wind is modi- lytic expression is available. These components are
represented in Equation (9) by B and sF/, respec-
fied by its encounter with the building, at least
insofar as its ability to produce a loading effect on tively. The factor K/CeH can be thought of as scaling
the structure is concerned. the result for the appropriate input turbulence level.
If resonance effects are small, then sF/ will be
26. For the purposes of calculating /, the small compared to the background turbulence B,
spectrum of the wind speed is represented by an al- and vice versa.
gebraic expression based on observations of real
wind. The aerodynamic admittance function is also 28. The peak factor, gp, depends on the average
an algebraic expression, computed on the basis of number of times the mean value of the loading ef-
somewhat simplified assumptions but appearing to fect is crossed during the averaging time of 1 h
be in reasonable agreement with the limited experi- (3 600 s). The functional relationship in Figure B-6
mental evidence at present available. The spectrum holds when the probability distribution of the mean
of wind speed is a function of frequency having the loading effect is normal (Gaussian).(11)

Commentary B
6.0
5.0
4.0
gp,Peak factor
3.0
2.0 gp = 2 loge +
0.577
νT
2 loge νT
T = 3 600 s
1.0
0
0.02 0.04 0.06 0.1 0.2 0.4 0.6 0.8 1 2 4
ν, Average fluctuation rate, cycles/second
Figure B-6
Peak factor as a function of average fluctuation rate
Sample Calculation of Cg noH/VH = reduced frequency for calculation of s

= 0.2 183/37.8
29. To illustrate the calculation of a gust effect = 0.968
factor the following sample problem will be worked
in detail: Step 2: Calculate /, using Equation (9)
Required: To obtain the gust effect factor for a K = 0.10 for Exposure B
building with the following properties: B = 0.62 (from Figure B-3)
Height, H 183 m s = 0.11 (from Figure B-4)

Width, W 30.5 m F = 0.28 (from Figure B-5)
Depth, D 30.5 m
= 0.015 (given)
Fundamental natural frequency, no 0.2 Hz
Critical damping ratio, 0.015
Terrain for site Exposure B
Reference wind speed, V̄,
at 10 m open terrain 27.4 m/s
Step 1: Calculate required parameters

CeH = 1.90 (from Figure B-1)
VH = (Equation (5)) Step 3: Calculate v, using Equation (10)
= 27.4 no = 0.2 Hz (given)
= 37.8 m/s
W/H = aspect ratio
= 30.5/183
= 0.17
no/VH = wave number for calculation of F
= 0.2/37.8
= 0.0053

Commentary B
Step 4: Obtain peak factor gp: if nD2 ≥ 0.75 m2/s,

gp = 3.75 (from Figure B-6).
(11c)
Step 5: Calculate Cg, using Equation (8)
Cg = 1 + gp(/) The Reynolds number is given by:
= 1 + 3.75(0.375)
= 2.41
Vortex Shedding Equation (11a) applies when Re < 2 x 105 and S =
1/6. Equation(11b) covers an intermediate region
30. Slender free standing cylindrical structures where, for computational convenience, S is taken to
such as chimneys, observation towers and in some increase approximately linearly as Re increases to
cases, high rise buildings, should be designed to re- 2.5 x 105. Equation (11c) usually governs, in which
sist the dynamic effect of vortex shedding. When Re > 2.5 x 105 and S = 1/5.
the wind blows across a slender prismatic or cylin-
drical body, vortices are shed alternately from one 32. The dynamic effects of vortex shedding
side and then the other, giving rise to a fluctuating from a cylindrical structure can be approximated by
force acting at right angles to the wind direction a static force acting over the top one-third. The
along the length of the body. A structure may be equivalent static force per unit height, FL, is given by
considered slender in this context if the ratio of
height to diameter exceeds 5. The wind speed, VH,
at the top of the structure when the frequency of
vortex shedding equals the natural frequency, n, of (12)
the structure is given by:
where
(11) = the critical damping ratio as defined for
Equation (9),
where = aspect ratio (H/D),
n = frequency in Hz, H = height of structure,
S = Strouhal Number, qH = the velocity pressure corresponding to
VH, where VH is in m/s, ≈ 0.6VH2 in Pa,
VH = the mean speed at the top of the struc-
ture in m/s, and M = average mass per unit length over the
top one-third of the structure (kg/m),
D = diameter in metres.
and
31. For circular and near-circular cylinders, the
= density of air ≈ 1.2 kg/m3.
Strouhal number is a function of the Reynolds num-
ber, Re. Although the Reynolds number is a For most situations
function of VH, a trial-and-error approach to finding C1 = 3 for > 16
the critical mean speed can be avoided by examin- C1 = 3()0.5/4 for < 16
ing the product, nD2, and using the appropriate C2 = 0.6.
version of Equation (11) as follows:
If then large amplitude motions up to
1 diam may result. Amplitude predictions for this
if nD2 ≤ 0.5 m2/s, case are discussed in Reference (12), which is sug-
gested as a general reference on the subject of
(11a) vortex shedding on chimneys.
If VH is low, temperature gradients may produce
if 0.5 m2/s < nD2 < 0.75 m2/s, very low turbulence levels, and in such cases vortex
induced motions are significantly increased, particu-
larly for very slender structures. If VH is less than
(11b) 10 m/s and greater than 12, then
C1 = 6, and
C2 = 1.2.

Commentary B
33. For tapered structures some reduction in the local pressure coefficients, . In Figures B-7 to B-
vortex shedding forces can be made. However, if the 13, dealing with low rise structures, values of the
variation in the diameter over the top third is less product CpCg are given; this is the form in which
than 10% of the average diameter of the top third, they are used, and of the basic wind tunnel data
then the recommendations of Paragraph 32 apply. from which they were derived.
If the diameter variation exceeds 10%, then the Internal Pressures

effective static load need be applied only over that
part of the structure over which the diameter is 37. The internal pressure coefficient, Cpi, defines
within 10% of the average for that part. the effect of wind on the air pressure inside the
building and is important in the design of both
For tapered structures with a diameter variation cladding elements and the primary structure. The
exceeding 10% over the top third magnitude of this coefficient depends on the distri-
C1 = 3, and bution and size of the leakage paths and openings
C2 = 0.6 which vent the internal air space to the external
and no increase in these coefficients is required for wind pressures. If the leakage is slow, through small
low values of VH. cracks and pores in the building envelope, and is
uniformly distributed, the internal pressures will
34. The recommendations of Paragraphs 32 and equilibrate to a pressure approximately equal to the
33 apply to free standing structures vibrating in the average external pressure over the exposed surface
fundamental mode. For other mode shapes a dy- of the building. With very small cracks and pores,
namic analysis would be appropriate, although the influence of gusts will be attenuated. If the
Equation (11) may be used to determine the critical openings are larger and more significant, on the
speed. The application of Equation (12) will yield a scale of doors or windows, the internal pressure
rough estimate of the effects of vortex shedding if will move closer to that prevailing externally at the
the load is applied over a length of up to one-third largest dominant opening and gust pressures will be
of the total length and in a position centred upon felt within the interior. Because of the changeability
the location of the maximum displacement for the and uncertainty of the size and distribution of these
mode of vibration in question. Slender structures openings, internal pressure coefficients can be influ-
with cross-sections other than circular may also give ential and wide ranging.
rise to vortex shedding, but data are limited and
other forms of across-wind motion may develop if In the face of these uncertainties, an appropriate
the wind speed, VH, is greater than about 7nD, treatment of internal pressure coefficients for both
where D is the across-wind breadth. In such cases high and low structures is to use the coefficients in
wind tunnel tests provide the most satisfactory Figure B-15 in conjunction with the formula in NBC
method of estimating the likely response. Sentence 4.1.8.1.(3) and a Cg of 1.0 or 2.0 as appro-
priate. This choice depends on whether there are
Pressure Coefficients significant openings and whether small openings
producing background leakage are uniformly dis-
35. Pressure coefficients are the non- tributed. In this context, a large or significant
dimensional ratios of wind-induced pressures on a opening is a single opening or a combination of
building to the dynamic pressure (velocity pressure) openings on any one wall which offers a passage to
of the wind speed at the reference height. Pressures the wind and whose area exceeds the leakage area
on the surfaces of structures vary considerably with of the remaining building surface, including the
the shape, wind direction and profile of the wind roof. Such a significant opening may be provided by
velocity. Pressure coefficients are usually deter- main doors, shipping doors, windows and ventila-
mined from wind tunnel experiments on small-scale tors if they are open during a storm, either through
models, although in a few recent instances measure- expected usage or through damage.
ments on full-scale buildings have been used
directly. In most cases these pressures must be mea- To handle the range of circumstances which may
sured in a wind tunnel in which the natural velocity prevail, three basic design categories are provided
profile and turbulence are simulated; experiments in as follows:
uniform flow can be highly misleading.(13)(14)
Category 1 (Cpi = 0.0 to – 0.3; Cg = 1.0)
36. The information on external and internal
pressure coefficients given in Figures B-7 to B-27 This category includes buildings without large or
covers the requirements for the design of the significant openings, but having small uniformly
cladding and the structure as a whole for a variety distributed openings (accumulating to less than
of simple building geometries. With the exception of 0.1% of total surface area). The value of Cpi should
B-7 to B-13, the values of the pressure coefficients be – 0.3 as given in Figure B-15, except where such
are given as either time- and spatially-averaged openings alleviate an external load, when Cpi = 0
pressure coefficients, Cp, or simply as time-averaged should be used. Internal pressure fluctuates even

Commentary B
within buildings having small distributed openings, different inside and outside air temperatures. Under
and the pressure fluctuations occasionally reach Cpi normal operations, mechanical systems create a dif-
= 0. Such buildings would include most high-rise ferential across walls of somewhat less than 0.1 kPa,
buildings that are nominally sealed and ventilated but the stack effect for differences in temperature of
mechanically, and exceptional lower buildings, such 40C could amount to 0.2 kPa per 100 m of building
as windowless warehouses, with door systems not height (Reference (30)).
prone to storm damage.

38. Figures B-7 to B-13 refer to low buildings
Category 2 (Cpi = 0.7 to – 0.7; Cg = 1.0) and present data obtained from systematic bound-
This category includes buildings in which signifi- ary layer wind tunnel studies. In several instances
cant openings, if any, can be relied on to be closed these data have been verified against full scale mea-
in storms but in which background leakage may not surements. The coefficients are based on the
be uniformly distributed. Most low buildings fall maximum gust pressures lasting approximately 1 s
into this category provided that all elements (ship- and, consequently, include an allowance for the gust
ping doors especially) are designed to be fully wind factor, Cg. The coefficients, therefore, represent the
resistant. In this category, the building should be product CpCg. These figures refer to the tributary
designed for the full range of Cpi (± 0.7) given in area associated with the particular element or mem-
Figure B-15. ber over which the wind pressure is assumed to act.
In all cases these coefficients should be combined
Category 3 (Cpi = 0.7 to – 0.7; Cg = 2.0) with the appropriate internal pressures. Figures B-7
to B-10 are most appropriate for buildings with
This category includes buildings with large or widths greater than twice their heights and a refer-
significant openings through which gusts are trans- ence height that does not exceed 10 m. In the
mitted to the interior. Normally the range of Cpi = absence of more appropriate data they may also be
± 0.7 should be used as given in Figure B-15, in used for buildings with H/W < 1 and reference
conjunction with Cg = 2.0. Such buildings would in- height less than 20 m. Beyond these extended limits,
clude, for example, sheds with one or more open Figure B-14 should be used. Further details of the
sides, and industrial buildings with shipping doors, work on which these results are based are given in
ventilators or the like, having a high probability of References (31) and (32).
being open during a storm or not fully resistant to
design wind loads. 39. Figure B-7 presents values of CpCg applica-
ble to those primary structural actions affected by
An ever-present threat in severe storms is the wind pressures on more than one surface, such as in
breakage of large unprotected glass areas (and other framed buildings. These simplified load distribu-
vulnerable components) by flying debris. Structures tions were developed to yield as closely as possible
required in post-disaster services should be capable the structural actions (horizontal thrust, uplift and
of withstanding all the consequences of failure of frame moments) determined directly from experi-
glass and be included in Category 3. For other ment. These results make allowance for the partial
structures, in which the glass is designed for wind loading of gusts referred to in NBC Sentence
and there is adequate protection against roof uplift, 4.1.8.3.(1).
the contingency of glass damage due to debris is
covered by normal load factors for wind.
40. Figures B-8 to B-10 are intended for those
In most cases, there is no need to consider non- actions influenced mainly by wind acting over sin-
uniform internal pressures except in the design of gle surfaces, such as the design of cladding and
internal partitions, see Clause 4.1.8.4.(1)(a). Thus, for secondary structural members. They should also be
most structural design, the two limiting values of used for design of primary structural elements of
internal pressure can be considered separately. Ex- single surfaces, for example, roofs for which mo-
ceptions might arise if interior compartments of the ment connections are not provided at the roof/wall
building are well sealed and wind damage or the intersection. In this case, the edge region loads need
like could expose one area of the building to Cate- not be included around the entire perimeter of the
gory 3 conditions while the rest of the building roof, but only adjacent to the windward edges. For
remains Category 1 or 2, resulting in unbalanced in- roofs exceeding 10 where edge regions are also
ternal pressures. specified along the ridge, these increased loads need
only be included on the downstream side. The loads
Internal pressures are also affected by mechanical on other edge regions can revert to the values speci-
ventilation systems and by the stack effect due to fied for the interior regions.

Commentary B
Load case A: winds generally perpendicular to ridge

2.5 H(9)
3
3E 2
4
4E 2E
roof
H slope 1
Reference
1E height (h)(5)
B
y(7)
wind direction
range
Building surfaces
Roof slope
1 1E 2 2E 3 3E 4 4E
0° to 5° 0.75 1.15 –1.3 –2.0 –0.7 –1.0 –0.55 –0.8
20° 1.0 1.5 –1.3 –2.0 –0.9 –1.3 –0.8 –1.2
30° to 45° 1.05 1.3 0.4 0.5 –0.8 –1.0 –0.7 –0.9
90° 1.05 1.3 1.05 1.3 –0.7 –0.9 –0.7 –0.9
Load case B: winds generally parallel to ridge
3 6
3E 6E
2
4
2E
4E
roof
slope 1
5 1E
5E y(7)
Z(8)
wind
direction
range
Building surfaces
Roof slope
1 1E 2 2E 3 3E 4 4E 5 5E 6 6E
0° to 90° –0.85 –0.9 –1.3 –2.0 –0.7 –1.0 –0.85 –0.9 0.75 1.15 –0.55 –0.8
Figure B-7
External peak pressure coefficients, CpCg, for primary structural actions arising from wind load acting simultaneously on all
surfaces.
Notes to Figure B-7:

(1) The building must be designed for all wind directions. Each corner must be considered in turn as the windward corner shown in the sketches.
For all roof slopes, Case A and Case B are required as two separate loading conditions to generate the wind actions, including torsion, to be
resisted by the structural system.
(2) For values of roof slope not shown, the coefficient (CpCg) may be interpolated linearly.
(3) Positive coefficients denote forces toward the surface, whereas negative coefficients denote forces away from the surface.
(4) Interior pressure coefficients Cpi are given in Figure B-15.
(5) The reference height, h, for pressures is mid-height of the roof or 6 m, whichever is greater. The eave height, H, may be substituted for the mean

height if the slope of the roof is less than 10 but still subject to the restriction of h > 6 m.
(6) For the design of foundations, exclusive of anchorages to the frame, only 70% of the effective load is to be considered.
(7) End zone width “y” should be the greater of 6 m or 2 z, where “z” is the gable wall end zone defined for Case B. Alternatively, for buildings with
frames, the end zone “y” may be the distance between the end and the first interior frame.
(8) End zone width “z” is the lesser of 10% of the least horizontal dimension or 40% of height, H, except that “z” must be at least 4% of the
horizontal dimension and at least 1 m.
(9) For B/H > 5 in Load Case A, the listed negative coefficients on surfaces 2 and 2E should only be applied on an area of width 2.5 H measured
from the windward eave. The pressures on the remainder of the windward roof should be reduced to the coefficients specified for the leeward
roof (i.e., those for 3 and 3E).

Commentary B
e
w
α z
H e
w e e
reference
z z height (h)
z
–3.0
–2.0
e
w
–1.0
Cp Cg
1.0
e and w
2.0
0 1 2 5 10 20 50 100
Area, m2
Figure B-8
External peak pressure coefficients, CpCg, on individual walls for design of cladding and secondary structural members

(1) These coefficients apply for any roof slope, .
(2) The abscissa area in the graph is the design tributary area within the specified zone.
(3) z = 10% of least horizontal dimension or 40% of height, H, whichever is less. Also z ≥ 1 m, z ≥ 4% of least horizontal dimension.
(5) The reference height, h, for pressures is mid-height of the roof or 6 m, whichever is greater. The eave height may be substituted for the

mid-height if the slope of the roof is less than 10 but still subject to the restriction of h > 6 m.
(6) Pressure coefficients may generally apply for facades with architectural features; however, when vertical ribs deeper than 1 m are placed on a
facade, a local CpCg = −2.8 may apply to zone e.(15)(16)

Commentary B
roof with overhang z z

roof without overhang
–6.0 z
(7)
c s
–5.0
(1, 7) r r
oc
–4.0
(1)
Cp Cg
os or
z
–3.0 (2) c
s
α
reference
H
(2) height (h)
r
–2.0 0° < α ≤ 10°
(6)
*
–1.0
s r c
+1.0
0 1 2 5 10 20 50 100
Area, m2
Figure B-9(a)

External peak pressure coefficients, CpCg, on roofs of 10 slope or less for design of cladding and secondary structural
members
Notes to Figure B-9(a):

(1) Overhang coefficients have the prefix “o” and refer to the same areas as the corresponding symbol without a prefix. They include contributions
from both upper and lower surfaces. In the case of overhangs, the walls are inboard of the roof outline.(17)(18)
(2) s and r apply to both roofs and upper surfaces of canopies.
(6) For uplift on tributary areas in excess of 100 m2 on unobstructed nearly-flat roofs with low parapets where the centre of the tributary area is at
least two building heights from the nearest edge, the value of CpCg may be reduced to −1.1 at x/H = 2 and further reduced linearly to −0.6 at x/H
= 5, where x is distance to the nearest edge and H is building height.(19)
(7) For roofs having a perimeter parapet with a height of 1 m or greater, the corner coefficients CpCg for small tributary areas can be reduced from
−5.4 to −4.4.(20)(21)

Commentary B
h1
H
h2
W1 W2
W
h1
H
b b
h2
W1 W2 W3
W
Figure B-9(b)
External peak pressure coefficients, CpCg, on stepped roofs for design of cladding and secondary structural members
Notes to Figure B-9(b):

(1) All notes to and CpCg values of Figure B-9(a) apply.
(2) On lower level(s) of horizontal stepped roofs, positive pressure coefficients equal to those for the walls (Figure B-8) apply for a distance, b. On
the vertical sections of the step wall(s) pressure coefficients provided for walls (Figure B-8) apply.(22)(23)
(3) Figure B-9(b) only applies where 0.3H ≤ h1, h2 ≤ 0.7H, where 0.25W ≤ W1, W2, W3 ≤ 0.75W and h1 ≥ 3 m.
(4) The dimension “b” is equal to 1.5h1 but not greater than 30 m.

Commentary B
z zz z
c z
z z
z
s'
s r r r
s
c z
α reference α
height (h) h
H
10° < α ≤ 45° 10° < α ≤ 30°
roof with overhang

roof without overhang
-8.0 gable and hip roofs

-8.0
gable roofs
10° < α ≤ 30° 30° < α ≤ 45°
-7.0 -7.0
-6.0 -6.0
-5.0 oc -5.0
os'
-4.0 c -4.0 oc os
'
s'
CpCg
-3.0 -3.0
CpCg
s c s s' os
-2.0 os -2.0
r r
-1.0 -1.0
0.0 0.0
s r c
1.0 1.0 s r c
2.0 2.0
0.1 1 10 100 0.1 1 10 100
Area, m2 Area, m2
Figure B-10

External peak pressure coefficients, CpCg, on roofs of greater than 10 slope for design of cladding and secondary structural
members

(1) Overhang coefficients have the prefix “o” and refer to the same areas as the corresponding symbol without a prefix. They include contributions
from both upper and lower surfaces. They apply only where wind has access to both upper and lower surfaces, which means that the walls
would be inside the outline of the roof.(28)(37)

(5) For hip roofs with = 10 – 30 , edge/ridge strips and pressure coefficients for ridges of gabled roofs apply along each hip.(24)

Commentary B
41. Figures B-14 and B-15 are for use with taller, Icing
rectangular structures for which H/W is greater
than 1. The pressure coefficients given are not yet 46. In locations where the strongest winds and
multiplied by a gust effect factor, Cg, because they icing may occur simultaneously, forces on structural
are intended for use either with those factors given members, cables and ropes must be calculated as-
in NBC Sentence 4.1.8.1.(6) or with the detailed suming an ice covering based on climate and local
method in which Cg for the structure as a whole experience. For the iced condition, values of Cn
depends on its dynamic properties as well as turbu- given in Figure B-22 for thick wire cables for a
lence characteristics. A local pressure coefficient, “rough” surface should be used.
= –1.0, applicable to the design of small cladding
areas (about the size of a window), can occur almost Structural Members
anywhere at any elevation, and is not limited to
corners. The simple procedure is mandatory for cal- 47. In Figures B-23, B-24, B-26 and B-27 pressure
culating local cladding loads. coefficients with the subscript ∞ are used to indicate
that they apply to structural members of infinite
42. Figures B-16 to B-27 are based on wind tun- lengths and this is multiplied by a reduction factor,
nel experiments in which the correct velocity profile k, for finite lengths of members. If a member
and wind turbulence were not simulated and should projects from a large plate or wall, the reduction
therefore be regarded with caution. They are based factor, k, should be calculated for a slenderness
on the Swiss Association of Engineers and Archi- based on twice the actual length. If a member termi-
tects Standards, S.I.A., No. 160, published in 1956.(33) nates with both ends in large plates or walls, the
reduction factors for infinite length should be used.
Protected Membrane Roofs
Loads on Frames and Shielding
43. In the case of a protected membrane roof,
with insulation which is not bonded to the water- 48. For framing members that are located be-
proofing membrane, the insulation is not subjected hind each other in the direction of the wind, the
to the same uplift pressure as is applied through the shielding effect may be taken into account. The
depth of the entire roof assembly, because of air windward members and those parts of the leeward
leakage and partial pressure equalization between members that are not shielded should be designed
the top and bottom of the insulation boards. Exter- with the full pressure, q, whereas the shielded parts
nal pressure or uplift due to wind is, therefore, of the leeward members should be designed with
applied to the membrane, which acts as an air the reduced pressure, qx, according to Figure B-25.
barrier between the inside and the outside and pre-
vents pressure equalization. Further information can A detailed discussion of the loads on unclad
be found in References (34) and (35). building frameworks is given in Reference (36).
Rounded Structures
44. For rounded structures (in contrast to sharp-

edged structures) the pressures vary with the wind
velocity, depending on the Reynolds number, Re,
(defined following Equation (11)). In Figures B-18,
B-19, B-22 and B-27, which have been translated and
reproduced from the Swiss tables,(30) the Reynolds
number is expressed by , where d is the di-
ameter of the sphere or cylinder in metres and q is
the velocity pressure in kilopascals. To convert to
Re, multiply by 2.7 106.
45. The roughness of rounded structures may be

of considerable importance. Common well-laid
brickwork without parging can be considered as
having a “moderately smooth” surface (Figure B-
18). Surfaces with ribs projecting more than 2% of
the diameter are considered “very rough.” In case of
doubt, those Cn values which result in the greater
forces should be used. For cylindrical and spherical
objects with substantial stiffening ribs, supports and
attached structural members, the pressure coeffi-
cients depend on the type, location and relative
magnitude of these roughnesses.

Commentary B
z zz z
z
z
s'
s r r
α reference α reference
H height (h) H height (h)
10° < α ≤ 30° 30° < α ≤ 45°
c c
-5.0 -5.0
s' s
-4.0 -4.0 r
s
-3.0 r -3.0 s'
r
-2.0 -2.0
CpCg
CpCg
-1.0 -1.0
0.0 0.0
s r c
1.0 1.0
s r c
2.0 2.0
0.1 1 10 100 0.1 1 10 100
Area, m2 Area, m2
Figure B-11

External peak pressure coefficients, CpCg, on multi-gabled (folded) roofs of greater than 10 slope for design of cladding and
secondary structural members.(25)(26)


Commentary B
2z 2z 2z s
s' c 2z z
4z c' 4z c'
s' s
r r
s
c' c'
s' c z
z z s
α α
H reference reference
height (h) H height (h)
3° < α ≤ 10° 10° < α ≤ 30°
-6 -6
-5 -5
c' c'
-4 s' -4
c
s
-3 -3
CpCg
CpCg
s
-2 -2 r
r
-1 -1
0 0
1 All regions 1 All regions
0 2 10 50 0 2 10 50
1 5 20 100 1 5 20 100
Area, m2 Area, m2
Figure B-12
External peak pressure coefficients, CpCg, on monoslope roofs for design of cladding and secondary structural
members.(27)(28)

(3) Interior pressure coefficients, Cpi, are given in Figure B-15.

(4) Coefficients given in this figure apply when the slope is greater than some small value (3 ) below which Figure B-9(a) applies.

Commentary B
z z
2z z
c
s
s
2z z
z
α reference
H height (h)
w/2 w/2
10° < α < 30°
-8
c (span A)
-7 A B C D
-6
s
-5 c (Spans B, C, D)
r
-4
-3
CpCg
-2
-1
+1 r
+2 c
s
+3
1 2 5 10 20 50 100
Area, m2
Figure B-13

External peak pressure coefficients, CpCg, on sawtooth roofs of greater than 10 slope for design of cladding and secondary
structural members.(29)

(3) Interior pressure coefficients, Cpi, are given in Figure B-15.
(4) Negative coefficients on the corner zones of Span A differ from those on Spans B, C and D.

Commentary B
Cp = –1.0 Cp = –0.7
Cp* = –2.3
Cp = –0.5
Cp = 0.8 Cp* = –1.5
0.2D
0.1D
Cp* = –1.0 0.2D

W
pressure
pressure distribution
distribution
(7)
Cp = –0.7
H
0.1D
D
Plan view of building

Z
Interior pressures Cpi

H1
1. Openings mainly in windward wall +0.7
2. Openings mainly in leewardwall –0.5
3. Openings mainly in walls –0.7

D parallel to wind direction
Elevation of building 4. Openings uniformly distributed –0.3
in all 4 walls
Figure B-14
Flat roofed buildings greater in height than in width Figure B-15
End wall pressure coefficients, local suction maxima on the
Notes to Figure B-14: roof and interior pressures for use with Figures B-7 to B-11
(1) Wind perpendicular to one wall: for width, use the dimension
perpendicular to the wind direction.
(2) Wind at an angle to the wall: this condition produces high local
(1) Local maximum suctions: the coefficients C*p for the roof surface
suctions on the wall which is at a slight angle to the wind. The
occur for wind at an angle to one corner, and are used in the
coefficient C*p may occur anywhere over the wall area, but need
design of the roofing itself and its anchorage to the structure. C*p
not be considered in conjunction with the Cp for over-all loading.
are not to be added to Cp for determining total uplift on the roof.
The coefficients C*p for the roof are given in Figure B-15.
(2) End walls: the end walls are the ones parallel to the wind direction;
(3) End walls: pressure coefficients for end walls (parallel to wind
they have a uniform pressure distribution over the whole building
direction) are given in Figure B-15.
height, except for local maximum suction as indicated in Figure
(4) Interior pressure: coefficients Cpi for interior pressures are given in
B-14.
Figure B-15.
(3) Reference height for exposure factor: for the calculation of external
(5) Reference heights for exposure factor for the calculation of both
pressures on end walls use H, the total height of the building. For
spatially-averaged and local pressures:
the calculation of internal pressures, use one-half H; where there
• leeward walls — 0.5 H
are dominant openings in the windward wall, use Z, the height to
• roof and side walls — H
the highest such opening.
• in conjunction with C*p — H
(4) The value of C*p can be reduced from −2.3 to −2.0 for roofs with
• any area at height Z above ground on the windward wall —
perimeter parapets having heights > 1 m.(20)(21)
Z
(5) On lower level(s) of horizontal stepped roofs, positive pressure
(6) Height H1: the height to which Ce is constant is 10 m for the
coefficients equal to those for the walls (Figure B-14) apply for a
simplified method and exposure A, 12.7 m for exposure B and
distance, b (see Figure B-9(b)). On the vertical sections of the step
30 m for exposure C.
wall(s), pressure coefficients provided for walls (Figure B-14)
(7) Windward wall Cp: the pressure coefficient is 0.8 for the entire
apply.(22)(23)
height. The variation in the pressure distribution shown is due to
variation in exposure factor Ce.
(8) Pressure coefficients generally apply for facades with architectural
features excluding large vertical ribs. When vertical ribs deeper
than 1 m are placed on a facade, a local C*p = -1.4 may apply to
an edge zone 0.2 D wide.(15)(16)
Commentary B
49. As the shape of a structure may change dur- discussed in Paragraph 52 should also be examined
ing erection, the wind loads may be temporarily for combined loads for its influence on torsion, as
higher during erection than after completion of the specified in NBC Clause 4.1.8.3.(1)(d). Further dis-
structure.(37) These increased wind loads should be cussion of combined loading effects can be found in
taken into account using the appropriate coefficients References (38) and (39).
from Figures B-7 to B-27.
Lateral Deflection of Tall Buildings

50. For constructions made from circular under Wind Loading
sections with < 0.167 and As/A ≥ 0.3, the
shielding factors can be taken by approximation 54. Lateral deflection of tall buildings under
from Figure B-22. If > 0.167, the shielding ef- wind loading may require consideration from the
fect is small and for a solidity ratio As/A ≤ 0.3, it standpoints of serviceability or comfort. The general
can be taken into account by a constant shielding trend is toward more flexible structures, partly be-
factor kx = 0.95. cause adequate strength can now be achieved by
using higher strength materials that may not pro-
Unenclosed Parking Structures vide a corresponding increase in stiffness.
51. For multi-level, unenclosed parking struc- 55. One symptom of unserviceability may be
tures, the exposed exterior area is reduced the cracking of masonry and interior finishes. Un-
compared with enclosed structures. However, inter- less precautions are taken to permit movement of
nal parts of the structure, and vehicles parked there, interior partitions without damage, a maximum lat-
are subject to additional wind forces not present on eral deflection limitation of 1/250 to 1/1 000 of the
enclosed structures. In lieu of a detailed analysis of building height should be specified. According to
the specific structure under consideration, a reason- Sentence 4.1.1.5.(4) of the NBC, 1/500 should be
able assumption is to treat the unenclosed parking used unless a detailed analysis is made.
structure as though it were enclosed.
Wind-Induced Building Motion
Partial Loading and Torsional Loading
56. While the maximum lateral wind-loading
52. NBC Sentence 4.1.8.3.(1) requires all build- and deflection are generally in the direction parallel
ings to be designed for partial loading as well as with the wind (along-wind direction), the maximum
full loading, as illustrated in Figure B-28. Removal acceleration of a building leading to possible human
of up to 25% of the load prescribed by the Code perception of motion or even discomfort may occur
from any part of the structure is intended to reflect in the direction perpendicular to the wind (across-
the observed behaviour of pressure patterns in tur- wind direction). Across-wind accelerations are likely
bulent wind. Some structures, such as arch-type to exceed along-wind accelerations if the building is
roof systems, undergo larger stresses under partial slender about both axes, that is if is less
loading. Low buildings designed to the specifica- than one-third, where W and D are the across-wind
tions of Figure B-7 do not need further unbalanced and along-wind plan dimensions and H is the
loads (see Paragraph 39). Tall buildings should be height of the building.
checked against partial loadings that produce
torsional effects. In wind tunnel testing, torsional ef- 57. Although treatment of this subject is some-
fects have sometimes been even greater than those what tentative, the following guidelines may be of
afforded by a 25% removal of loads from selected assistance. A wide range of turbulent boundary
areas of the building.(14) Torsional effects are en- layer wind tunnel studies have demonstrated that
hanced when the centre of twist is eccentric from the peak acceleration in the across-wind direction at
the centre of gravity (inertial loading) or from the the top of the building can be found from the fol-
centre of area (wind loading, full or partial). lowing:
53. The pressure coefficients in Figures B-14 and
B-15 refer to the pressures acting along the principal
axes of rectangular building forms. They are most (13)
appropriate for buildings where H/W > 1 and H >
10 m. In some structural systems more severe effects
may be induced when the resultant wind pressures 58. In less slender structures or for lower wind
approach the building diagonal. To account for this speeds, the maximum acceleration may be in the
and also the additional tendency for structures to along-wind direction and can be found from the ex-
sway transversely to the wind direction, the struc- pression
ture should be capable of resisting 75% of the
maximum loads for each of the principal directions
applied jointly, as specified in NBC Sentence
4.1.8.3.(1). The influence of the 25% removal of load (14)

Commentary B
where 62. If a more rigorous analysis is not available,

the maximum deflection resulting from the equiva-
W, D = across-wind and along-wind building di- lent static wind loading can be related to the
mensions, m, fundamental building frequency using modal repre-
sentation of the building motion. The following
aW, aD = peak acceleration in across-wind and assumptions may be acceptable:
along-wind directions, m/s 2,
(1) Use first mode only, assumed linear
ar = , Pa,
(15)
B = average density of the building, kg/m3,
(2) Uniform distribution of building mass
W, D = fraction of critical damping in across-
wind and along-wind directions,
(16)
nW, nD = fundamental natural frequencies in
across-wind and along-wind directions,
Hz, As a consequence of modal representation
1 = maximum wind-induced lateral deflec-
tion at the top of the building in
along-wind direction, m,
(17)
g = acceleration due to gravity = 9.81 m/s2,
gp, K, s, F, Ce, Cg are as defined previously in
connection with Equation (9). (18)
Note that D = and nD = nO in terms of previ-

ous definitions.
59. Although many additional factors such as
(19)
visual cues, body positions and orientation and
state-of-mind influence human perception of mo-
tion, when the amplitude of acceleration is in the where
range of 0.5% to 1.5% of the acceleration due to
gravity, movement of the building becomes percep- (Z) = fundamental eigenvector,
tible to most people.(40)(41)(42)
r1, r2 = constants,
60. Based on this and other information, a tenta-
tive acceleration limitation of 1 to 3% of gravity m(Z) = distribution of building mass, with
once every 10 years is recommended; for use in con- height Z, 0 < Z < H, kg/m,
junction with Equations (13) and (14) the lower
value might be considered appropriate for apart- g = acceleration due to gravity = 9.81 m/s2,
ment buildings, the higher value for office
buildings. The application of Equations (13) and u(Z) = displacement at height Z, 0 < Z < H, m,
(14) tends to give conservative results insofar as P(Z) = distribution of equivalent static wind
they assume that the wind always comes from the pressure with height Z, 0 < Z < H, kPa.
most sensitive direction; this factor has also been
considered in setting the above limitation. A de- Other symbols are as defined earlier.
signer who has more detailed information available
can make suitable allowances. From Equations (15), (16) and (17)
61. Owing to the relative sensitivity of expres-
sions (13) and (14) to the natural frequency of
vibration, and in (14) to the corresponding building (20)
stiffness, these should be determined using fairly
rigorous methods, and approximate formulas
From Equations (18), (19) and (20)
should be used with caution. For example, the
adoption of a natural frequency of 10/N, where N
is the number of storeys, may not be consistent with
the assumption that the displacement under wind
loading is as large as H/500. (21)

Commentary B
Substituting Equation (21) into Equation (18), the Therefore, aW/g = 7.1%.
deflection at height H becomes
Step 3: Calculate q
q = CV̄2
(22) = 650 10−6 27.42
One possible expression for P(Z) assumes a = 0.488 kN/m2

power law variation with a maximum at the top of
qCeCgCp Step 4: Calculate aD/g
(23)
where Cp = 0.8 –(– 0.5) = 1.3 and is the appro-

priate exponent from Equations (2), (3) and (4).
Substituting Equations (22) and (23) into (14)
64. In this example, clearly the across-wind

accelerations overshadow the along-wind accelera-
(24) tions. Table B-2 gives the results of calculations for
two other sample buildings. One is rectangular and
is examined in cases 1 and 2 for wind along both
63. Sample Calculation of aW and aD. A de- axes. Results are given for three different terrains
tailed calculation for aW and aD using Equations (13) and for three different wind pressures, correspond-
and (24) will be made for the sample problem ing to Montreal, Toronto and Vancouver. A tall
worked earlier to illustrate the calculation of a gust building with a waterfront location may be exposed
effect factor: to all three terrain conditions for different wind di-
rections.
assume that nW = nD = 0.2 Hz
Pressure Differences Across Interior Walls
and Partitions
65. Considerable pressure differences can result

across interior walls and partitions in high-rise
buildings and in low-rise buildings in exposed loca-
Step 1: Calculate ar tions, if windows are broken during a storm. In
certain locations almost the full pressure difference
between the windward and leeward sides of the
building could be applied across interior walls or
partitions. For example, a large window on the
windward side might be broken by flying debris
and the full positive pressure exerted on the walls
of a small room located at the broken window. Simi-
lar conditions could prevail in an apartment
Step 2: Calculate aW building with operable windows or doors. This
pressure difference could be aggravated by stack ef-
fects in a tall building in the winter. On the other
hand, experience does not indicate many failures of
interior walls due to this cause, and thus interior
walls and partitions are not required to be designed
for the maximum possible pressure difference. A de-
sign pressure difference of the order of 0.5 kPa may
be appropriate.

Commentary B
Table B-2
Wind Induced Building Motions: Examples of Calculated Peak Accelerations (a) in Along-Wind (D) and Across-Wind (W)
Directions at Top of Building (H)
q = 0.31 kPa (1/10 basis) q = 0.39 kPa (1/10 basis) q = 0.45 kPa (1/10 basis)
Zone Exposure
Factor, Ce V (H), V (H), V (H),
aD, % g aW, % g aD, % g aW, % g aD, % g aW, % g
m/s m/s m/s
Case 1: 120 x 50 x 30 m building(1)
open 2.01 30.9 1.61 2.99(2) 34.7 2.30 4.38(2) 37.3 2.87(2) 5.55(2)
suburban 1.54 27.1 1.24 1.93 30.4 1.78 2.82(2) 32.6 2.22 3.52
city 1.09 22.8 0.93 1.08 25.5 1.33 1.58 27.4 1.67 2.01
open 2.01 30.9 1.80 1.84 34.7 2.53(2) 2.69(2) 37.3 3.12(2) 3.41(2)
suburban 1.54 27.1 1.40 1.18 30.4 1.98 1.73 32.6 2.45 2.19
city 1.09 22.8 1.07 0.66 25.5 1.51 0.97 27.4 1.88 1.23
open 2.43 34.1 1.75 3.23(2) 38.2 2.45 4.73(2) 41.1 3.01(2) 5.99(2)
suburban 2.17 32.2 1.59 2.68(2) 36.1 2.24 3.92(2) 38.8 2.75(2) 4.96(2)
city 1.79 29.2 1.43 1.94 32.8 2.01 2.83(2) 35.2 2.48 3.59(2)
Notes to Table B-2:

(1) Full dimensions and properties of Cases 1 to 3 are given in the following Table.
(2) Value might exceed acceptable limits and detailed wind tunnel tests might be warranted.
Weight in D Direction in W Direction

Case Height H, m Density, kN/
m3 Dimen., m Frequency Damping Dimen., m Frequency Damping
1 120 1.5 50 0.250 0.015 30 0.200 0.010
2 120 1.5 30 0.200 0.010 50 0.250 0.015
3 240 1.9 50 0.125 0.010 50 0.125 0.010
Tornadoes (i) anchorage of home floors into the foundation

or ground (floor takes off with the occupants
66. Although the probability of any specific on it), and
building being hit by a tornado is very small (less (ii) anchorage of roofs down through concrete
than 10-5 per year(43)) in total, tornadoes account for block walls (roof takes off and unsupported
the greatest incidence of death and serious injury of block wall collapses onto occupants).
building occupants due to structural failure and a
considerable economic loss, of the order of $5 mil- 67. The first detail, anchorage of home floors, is
lion per year in Canada. With some exceptions, such essentally covered by NBC Article 9.23.6.1. for nor-
as nuclear power plants, it is generally not economi- mal housing with permanent foundations. Also for
cal to design buildings for tornadoes beyond what is mobile homes, CAN/CSA Z240.10.1-94 contains an-
currently required by NBC Subsection 4.1.8. because chorage recommendations for tornado protection.
of the low risk of loss to individual owners (insur- The second detail, roof anchorage in block walls, is
ance is cheaper). It is, however, important to provide essentially covered in CAN/CSA S304.1-94 by limit
key construction details for the safety of building states requirements for wind uplift and, for the em-
occupants. Investigations of tornado-damaged areas pirical method of masonry design, by Clause 16.1.4
in Eastern Canada(44)(45) have shown that the build- of the standard; however, it is not adequately cov-
ings in which well over 90% of the occupants were ered by the empirical method in CAN3 S304-M84
killed or seriously injured by tornadoes did not sat- (allowable stress design). Deficiency of this
isfy two key details of building construction: construction detail is especially serious for open
assembly occupancies because there is nothing in-
side, such as stored goods, to protect the occupants

Commentary B
from wall collapse. For such buildings in tornado- (11) A.G. Davenport, Note on the Distribution of
prone areas it is recommended that the block walls the Largest Value of a Random Function with
contain vertical reinforcing linking the roof to the Application to Gust Loading. Proc., Inst. Civ.
foundation. Eng., London, Vol. 28, June 1964, pp. 187-196.
(12) B.J. Vickery and R.I. Basu, Simplified
68. For tornado protection, key details such as Approaches to the Evaluation of the Across-
those indicated above should be designed on the Wind Response of Chimneys. Journal of Wind
basis of a factored uplift wind suction of 2 kPa on Eng. and Indust. Aerodynamics, Vol. 14, De-
the roof and a factored lateral wind pressure of cember 1983, pp. 153-166.
1 kPa on the windward wall and suction of 2 kPa (13) M. Jensen and N. Franck, Model Scale Tests in
on the leeward wall. Turbulent Wind, Part II. Danish Technical
Press, Copenhagen, 1965.
69. Guidance for determining if a given locality (14) D. Surry, R.B. Kitchen and A.G. Davenport,
is prone to tornadoes may be obtained from Energy Design Effectiveness of Wind Tunnel Studies
and Industrial Relations Section, Canadian Climate for Buildings of Intermediate Height. Can. J.
Centre, Atmospheric Environment Service, Environ- Civ. Eng., Vol. 4, No. 1, 1977, pp. 96-116.
ment Canada, 4905 Dufferin Street, Downsview, (15) T. Stathopoulos and X. Zhu, Wind Pressures
Ontario M3H 5T4. on Buildings with Appurtenances. Journal of
References Wind Eng. and Indust. Aerodynamics. Vol. 31,
1988, pp. 265-281.
(1) Canadian Commission on Building and Fire (16) T. Stathopoulos and X. Zhu, Wind Pressures
Codes, National Building Code of Canada on Buildings with Mullions. Journal of Struc-
1995. National Research Council of Canada, tural Eng., ASCE, Vol. 116, No. 8, 1990, pp.
Ottawa, NRCC 38726 2272-2291.
(2) E. Simiu and R.H. Scanlan, Wind Effects on (17) T. Stathopoulos, Wind Loads on Eaves of Low
Structures: An Introduction to Wind Engi- Buildings. Journal of Structural Division,
neering. John Wiley & Sons, New York, 1986. ASCE, Vol. 107, No. ST10, October 1981, pp.
(3) J.E. Cermak, Application of Fluid Mechanics 1921-1934.
to Wind Engineering. Freeman Scholar Lec- (18) T. Stathopoulos and H.D. Luchian, Wind-
ture, Journal of Fluid Engineering, ASME, Induced Forces on Eaves of Low Buildings.
Vol. 97, No. 1, March 1975. Journal of Wind Eng. and Indust. Aerody-
(4) D. Surry and N. Isyumov, Model Studies of namics, Vol. 52, 1994, pp. 249-261.
Wind Effects – A Perspective on the Problems (19) D. Surry and E.M.F. Stopar, Wind Loading of
of Experimental Technique and Instrumenta- Large Low Buildings. Can. J. Civ. Eng., Vol.
tion. Int. Congress on Instrumentation in 16, 1989, pp. 526-542.
Aerospace Simulation Facilities, 1975 Record, (20) T. Stathopoulos and A. Baskaran, Wind Pres-
pp. 79-90. sures on Flat Roofs with Parapets. Journal of
(5) A.G. Davenport, Gust Loading Factors. Jour- Structural Division, ASCE, Vol. 113, No. 11,
nal of Structural Division, Proc., Am. Soc. Civ. Nov. 1987, pp. 2166-2180.
Eng., Vol. 93, June 1967, pp. 12-34. (21) T. Stathopoulos, Wind Pressures on Flat Roof
(6) D.R. Lemelin, D. Surry and A.G. Davenport, Edges and Corners. Proc. of Seventh Interna-
Simple Approximations for Wind Speed-Up tional Conference on Wind Engineering,
Over Hills. 7th International Conference on Aachen, West Germany, July 6-10, 1987.
Wind Engineering, Aachen, West Germany, (22) T. Stathopoulos and H.D. Luchian, Wind Pres-
July 6-10, 1987. sures on Building Configurations with
(7) P.S. Jackson and J.C.R. Hunt, Turbulent Wind Stepped Roofs. Can. J. Civ. Eng., Vol. 17, No.
Flow Over a Low Hill. Quart. Journal R. Met. 4, 1990, pp. 569-577.
Soc., Vol. 101, 1975, pp. 929-955. (23) T. Stathopoulos and H.D. Luchian, Wind
(8) J.L. Walmsley, P.A. Taylor and T. Keith, A Loads on Flat Roofs with Discontinuities.
Simple Model of Neutrally Stratified CSCE Annual Conf., Vancouver, May 1991.
Boundary-Layer Flow Over Complex Terrain (24) D. Meecham, D. Surry and A.G. Davenport,
With Surface Roughness Modulations. The Magnitude and Distribution of Wind-
Boundary-Layer Meteorology, Vol. 36, 1986, Induced Pressures on Hip and Gable Roofs,
pp. 157-186. 8th Coll. on Ind. Aerodynamics, Aachen, Ger-
(9) P. Sachs, Wind Forces in Engineering. Second many, September 1989.
Edition, Pergamon Press, Toronto, 1978. (25) J.D. Holmes, Wind Loading on Multi-span
(10) L. Christensen and S. Frandsen, A Field Study Building. 1st National Structural Eng. Conf.,
of Cross Wind Excitation of Steel Chimneys: Melbourne, Australia, August 1987.
Safety of Structures under Dynamic Loading.
Norwegian Institute of Technology, Trond-
heim, June 1977, pp. 689-697.

Commentary B
(26) T. Stathopoulos and P. Saathoff, Wind Pres- (36) P.N. Georgiou and B.J. Vickery, Wind Loads
sures on Roofs of Various Geometries. Journal on Building Frames. Proc. Fifth International
of Wind Eng. and Indust. Aerodynamics, Vol. Conference on Wind Engineering, Colorado
38, 1991, pp. 273-284. State University, July 1979, Pergamon Press.
(27) T. Stathopoulos and A.R. Mohammadian, (37) D.E. Walshe, Measurements of Wind Force on
Wind Loads on Low Buildings with Mono- a Model of a Power Station Boiler House at
sloped Roofs. Journal of Wind Eng. and Various Stages of Erection. NPL Aero Report
Indust. Aerodynamics, Vol. 23, 1986, pp. 81-97. 1165, National Physical Laboratory, Tedding-
(28) D. Surry and T. Stathopoulos, The Wind ton, England, September 1965.
Loading of Low Buildings with Mono-sloped (38) Wind Loading and Wind-Induced Structural
Roofs. Final Report BLWT-SS38, University of Response, Wind Effects Committee, American
Western Ontario, London, Ont., 1988. Society of Civil Engineers. ASCE, New York,
(29) P. Saathoff and T. Stathopoulos, Wind Loads 1987.
on Buildings with Sawtooth Roofs. Journal of (39) N. Isyumov, The Aeroelastic Modelling of Tall
Structural Eng., ASCE, Vol. 118, No. 2, 1992, Buildings. International Workshop on Wind
pp. 429-446, Paper No. 675. Tunnel Modeling Criteria and Techniques in
(30) Y. Lee, H. Tanaka and C.Y. Shaw, Distribution Civil Engineering Applications, Gaithersburg,
of Wind and Temperature Induced Pressure Maryland, April 1982. Cambridge University
Differences Across the Walls of a Twenty Press, 1982.
Story Compartmentalized Building. Journal of (40) P.W. Chen and L.E. Robertson, Human Per-
Wind Eng. and Indust. Aerodynamics, Vol. 10, ception Thresholds of Horizontal Motion.
1982, pp. 287-301. Journal of Structural Division, Proc., Am. Soc.
(31) T. Stathopoulos, D. Surry, and A.G. Daven- Civ. Eng., Vol. 98, August 1972, pp. 1681-1695.
port, Internal Pressure Characteristics of (41) F.K. Chang, Human Response to Motions in
Low-Rise Buildings Due to Wind Action. Tall Buildings. Journal of Structural Division,
Proc. Fifth International Conference on Wind Proc., Am. Soc. Civ. Eng., Vol. 99, June 1973,
Engineering, Colorado State University, July pp. 1259-1272.
1979, Pergamon Press. (42) R.J. Hansen, J.W. Reed and E.H. Van Marcke,
(32) D. Surry, T. Stathopoulos and A.G. Daven- Human Response to Wind-Induced Motion of
port, The Wind Loading of Low Rise Buildings. Journal of Structural Division,
Buildings. Proc. Can. Struct. Eng. Conference, Proc., Am. Soc. Civ. Eng., Vol. 99, July 1973,
Toronto, 1978. pp. 1587-1605.
(33) Normen fur die Belastungsannehmen, die In- (43) M.J. Newark, A Design Basis Tornado. Can. J.
betriebnahme und die Uberwachung der Civ. Eng., Vol. 18, 1991, pp. 521-524.
Bauten. (Standards for Load Assumptions, (44) D.E. Allen, Tornado Damage at Blue Sea Lake
Acceptance and Inspection of Structures). and Nicabong. Building Research Note 222,
Schweizerischer Ingenieur und Architekten Institute for Research in Construction,
Verein (Swiss Association of Engineers and National Research Council of Canada,
Architects), No. 160, Zurich, Switzerland, Ottawa, 1984.
1956. (45) D.E. Allen, Tornado Damage in the Barrie/
(34) R.J. Kind and R.L. Wardlaw, Model Studies of Orangeville Area, Ontario, May 1985.
the Wind Resistance of Two Loose-Laid Roof- Building Research Note 240, Institue for
Insulation Systems. Laboratory Technical Research in Construction, National Research
Report, LTR-LA-234, National Aeronautical Council of Canada, Ottawa, 1986.
Establishment, National Research Council of
Canada, Ottawa, May 1979.
(35) R.J. Kind and R.L. Wardlaw, Design of
Rooftops Against Gravel Blow-Off. National
Aeronautical Establishment, National
Research Council of Canada, Ottawa, Septem-
ber 1976. NRCC 15544.

Commentary B
Cp : External pressure
h : b : l = 1 : 1 : 10
coefficients
φ A B C D
b l
0° +0.8 –1.2 –1.4 –1.5
DOORS CLOSED
C h
A B
Cpi : Internal pressure
D 2h coefficients
OPENINGS φ = 0°
Uniformly distributed –0.5
Predominating on side “A” +0.7
Predominating on side “B” –1.1
Predominating on side “C” –1.3
Figure B-16
Closed passage between large walls
END WALLS Fn = C · q · C g · C e · h · L
f
l / h = 10 → ∝ l C – FORCE COEFF. FOR
h f
l / h = 10 WALLS ABOVE GROUND
h
h l/h=1 l/h 10 → ∝ 10 1
(End Walls)
l l
φ = 0°
Fn a Fn a Fn a 2.0 1.3 1.15
a = .5l
φ = 40°
1.6
a = .3l
φ = 0° φ φ φ = 50°
1.8
END WALLS l a = .4l
l / h = 10 → ∝ Cf – FORCE COEFF. FOR

h h l / h = 10 WALLS ON THE GROUND
h l/h=1
l/h 10 → ∝ 10 1
(End Walls)
l l
Fn ρ = 0°
Fn Fn
1.2 1.2 1.1
a a a = .5l
a
ρ = 40°
1.5
a = .3l
= 0° ρ = 50°
1.5
a = .4l
Figure B-17
Free standing plates, walls and billboards

Commentary B
TOTAL FORCE F = C · q · Cg · Ce · A, where A = d · h

f
h/d = 25
C : FORCE COEFFICIENT FOR d qCe > 0.167
f
h/d = 7
Slenderness h/d = 25 7 1
h/d = 1
Cross section and roughness Cf Cf Cf
Moderately smooth,
(metal, timber, concrete) 0.7 0.6 0.5
–
l/d = 2
h/d = 25
p Rough surface
h/d = 1
e 0.9 0.8 0.7

(rounded ribs h = 2%d)
h/d = 7
d Very rough surface

p (sharp ribs h = 8%d) 1.2 1.0 0.8
h α° i
+
l
Smooth and rough
surface sharp edges 1.4 1.2 1.0
Cp : EXTERNAL PRESS. COEFF. FOR d qCe > 0.167 and moderately smooth surface
h/d l/d α= 0° 15° 30° 45° 60° 75° 90° 105° 120° 135° 150° 165° 180°
25 50 Cp +1.0 +0.8 +0.1 –0.9 –1.9 –2.5 –2.6 –1.9 –0.9 –0.7 –0.6 –0.6 –0.6
7 14 Cp +1.0 +0.8 +0.1 –0.8 –1.7 –1.6 –2.2 –1.7 –0.8 –0.6 –0.5 –0.5 –0.5
1 2 Cp +1.0 +0.8 +0.1 –0.7 –1.2 –1.6 –1.7 –1.2 –0.7 –0.5 –0.4 –0.4 –0.4
pi = Cpi · q · Cg · Ce
Stack fully operating Cpi = +0.1; Stack throttled Cpi = –0.8
∆p = pi – pe
pe = Cp· q · Cg · Ce
Figure B-18
Cylinders, chimneys and tanks
TOTAL FORCE F = Cf · q · Cg · Ce · A; A = π d
2
4
for d qCe > 0.8 and moderately smooth surface
– C f : FORCE COEFFICIENT
pe C f = 0.2
d pi
α° + + p = pi – pe pi for closed tanks = working pressure
Fn pe = C p · q · C g · C e
Cp : EXTERNAL PRESS. COEFF. FOR d qCe > 0.8 and moderately smooth surface
α= 0° 15° 30° 45° 60° 75° 90° 105° 120° 135° 150° 165° 180°
Cp +1.0 +0.9 +0.5 –0.1 –0.7 –1.1 –1.2 –1.0 –0.6 –0.2 +0.1 +0.3 +0.4
Figure B-19
Spheres

Commentary B
Cp : EXTERNAL PRESSURE COEFFICIENTS

RAD. r = 5/6 b h:b:l = 1:12:12
1/6 b
φ A B C D E F G H J K
F G H J 0° +0.7 –0.2 –0.3 –0.3 –0.1 –0.5 –0.8 –0.8 –0.4 –0.1
E K
h A r = 5/6 b B
30° +0.6 –0.3 +0.2 –0.4 –0.1 –0.4 –0.7 –0.9 –0.7 –0.4
b
Y = 0.1b x = 0.1 b φ A B C D L M N O P Q
C
L m
90° –0.3 –0.3 +0.9 –0.3 –0.8 –0.7 –0.5 –0.3 –0.1 –0.1
φ M
N 30° Section “m” Cp * = –1.8 with Cp * min. = –2.5
0° A l
O B
P Cpi : INTERNAL PRESSURE COEFFICIENTS
Q
D OPENINGS φ = 0° φ = 30° φ = 90°
Shaded Area to Scale Uniformly distributed ±0.2 ±0.2 ±0.2
Window Y open on side “A” +0.4 +0.7 –1.0
All doors open on side “C” –0.1 +0.6 +0.8
Only door X open on side “C” –1.5 +0.7 +0.4
Figure B-20
Hangar, curved roof with moderately smooth surface
h : d : r = 1 : 1 : 1.5
Total force on roof F = (pi – pe) A
pi = working pressure in kN/m2

0.1 d
F pe = Cp · q · Cg · Ce
P π 2
i r A= d
h 4
d
Cp external pressure
coefficient = –1.0
Figure B-21
Roof load on smooth closed tank
C f = FORCE COEFFICIENTS
l/d > 100
Total force F = C f · q · C g · C e · A d q Ce
<0.167 >0.167
Smooth wires, rods, pipes 1.2 0.5
A=d·l Mod. smooth wires and rods 1.2 0.7
Fine wire cables 1.2 0.9
Thick wire cables 1.3 1.1
Figure B-22
Poles, rods and wires

Commentary B
l = Length of member
A = h · l = Area
For wind normal to axis of member: Normal force Fn = k · Cn · q · C g · Ce · A
Tangential force Ft = k · Ct · q · C g · Ce · A
Cn and Ct : Force coefficients for an infinitely long member
+Ft 1/2 h +Ft +Ft +Ft +Ft +Ft

h h
+Fn +Fn +Fn +Fn +Fn
0° 0° h 0°
h 0° 0° h 0° h
α 0.1h
0.1h 0.43h
+Fn h 1.1h
0.45h
αw C n Ct Cn Ct Cn Ct Cn Ct Cn Ct Cn Ct
0° +1.9 +0.95 +1.8 +1.8 +1.75 +0.1 +1.6 0 +2.0 0 +2.05 0
45° +1.8 +0.8 +2.1 +1.8 +0.85 +0.85 +1.5 –0.1 +1.2 +0.9 +1.85 +0.6
90° +2.0 +1.7 –1.9 –1.0 –0.1 +1.75 –0.95 +0.7 –1.6 +2.15 0 +0.6
135° –1.8 –0.1 –2.0 +0.3 –0.75 +0.75 –0.5 +1.05 –1.1 +2.4 –1.6 +0.4
180° –2.0 +0.1 –1.4 –1.4 –1.75 –0.1 –1.5 0 –1.7 ±2.1 –1.8 0
2/3h +Ft +Ft +Ft +Ft +Ft +Ft
h h h h
h
h
+Fn +Fn +Fn +Fn +Fn +Fn
0° 0° 0° 0° 0° 0°
α
0.48h h 0.1h 0.5h h
1.6h
αw C n Ct Cn Ct Cn Ct Cn Ct Cn Ct Cn Ct
0° +1.4 0 +2.05 0 +1.6 0 +2.0 0 +2.1 0 +2.0 0
45° +1.2 +1.6 +1.95 +0.6 +1.5 +1.5 +1.8 +0.1 +1.4 +0.7 +1.55 +1.55
90° 0 +2.2 ±0.5 +0.9 0 +1.9 0 +0.1 0 +0.75 0 +2.0
For slenderness, Ft
k : Reduction factor for members of finite
hα is to be used: Fres slenderness (in general use full length
not panel length)
0° Fn
α
l/hα 5 10 20 35 50 100
l
hα k 0.60 0.65 0.75 0.85 0.90 0.95 1.0
Figure B-23
Structural members, single and assembled sections

Commentary B
A s = Section area
A = ht · L
A s / A = Solidity ratio
For wind normal to surface A : Normal force Fn = k · C n · q · Cg · Ce · As

ht Fn
Cn : Force coeff. for an infinitely k : Reduction factor for trusses of

long truss, 0 ≤ A s / A ≤ 1 finite length and slenderness
0.3 to A s /A
As/A 0 0.1 0.15 0.2 0.8 0.95 1.0 L/h t 0.25 0.5 0.9 0.95 1.0
Cn 2.0 1.9 1.8 1.7 1.6 1.8 2.0 5 0.96 0.91 0.87 0.77 0.60
20 0.98 0.97 0.94 0.89 0.75
50 0.99 0.98 0.97 0.95 0.90
1.0 1.0 1.0 1.0 1.0
Figure B-24
Plane trusses made from sharp-edged sections
k x SHIELDING FACTOR
PLANE OF PLANE OF A s /A
MEMBER I MEMBER II x / h 0.1 0.2 0.3 0.4 0.5 0.6 0.8 1.0
0.5 0.93 0.75 0.56 0.38 0.19 0 0 0
h
or q qx 1 0.99 0.81 0.65 0.48 0.32 0.15 0.15 0.15
ht 2 1.00 0.87 0.73 0.59 0.44 0.30 0.30 0.30
x
4 1.00 0.90 0.78 0.65 0.52 0.40 0.40 0.40
q x = kx · q 6 1.00 0.93 0.83 0.72 0.61 0.50 0.50 0.50
Figure B-25
Shielding factors

Commentary B
L B = Length of bridge
CASE I WITHOUT VEHICLES
k, Cn , As, kx from Figs. B-23 & B-24
x Windward girder FI = kCn · q · Cg · Ce · As
FI Leeward girder FII = kCn · kxq · Cg · Ce · As

q d F II
ht Fh Deck horiz. load Fh = 1.0 · q · Cg · Ce · LB

Deck vert. load Fvert. = 0.6 · q · Cg · Ce · b · LB
F vert.
0.4b
Lv = Length of vehicle; A1 = h v · Lv
b 1
A2 = h v · Lv
2
Windward girder FI = kCn · q · Cg · Ce · As
CASE II WITH VEHICLES
Leeward girder FII = kCn · kxq · Cg · Ce · As
hv
1 Deck horiz. load Fh = 1.2 · q · Cg · Ce · d · LB
hv Fv Deck vert. load Fvert. = 0.8 · q · Cg · Ce · b · LB
1
q Fv F II Traffic load Fv = Cnq · Cg · Ce · A1
FI hv 2 1
2
Fv = Cn · 2/3 q · Cg · Ce · A2
Fh d 2
0.4 b
Height and force coefficients hv Cn
F vert.
Railway vehicle 3.8 m 1.5
b Highway vehicle 3.0 m 1.2
Pedestrian 1.7 m 1.0
Figure B-26
Truss and plate girder bridges
A = d · L or h · L
As/A ≤ 0.3 L = true length of member
β = angle formed by wind direction
and the normal to member axis
kx = a function of As/A and x/b
TOTAL LOAD IN WIND DIRECTION F = ΣFm
x b
0° Fm = FORCE ON MEMBER
90° d Fm = k · C β · q · Cg · Ce · A cos β
x b
45° β
Fm (Shielded member Fm =
x b h k·C β · kxq · Cg · Ce · A cos β)
l
0°
Fn β
Coeff. C β : For sharp-edged members C β = kβ · Cn and kβ · Ct

Coeff. C β, kβ, k, kx : SEE FIG. B-23 FOR Cn and Ct values
ROUND MEMBERS, SMOOTH ROUND MEMBERS,

& ROUGH SURFACES MODERATELY SMOOTH
SHARP-EDGED MEMBERS
β d qCe < 0.167 SURFACES, d qCe < 0.167
kβ k kx C β k kx C β k kx
0° 1.00 1.20 0.60
15° 0.98 See See 1.16 See See 0.58 0.9
Fig. Fig. Fig. Fig. for 0.95
30° 0.93 1.04 0.53 constant
B-23 B-25 B-23 B-25 l/d = 25
45° 0.88 0.85 0.42
60° 0.80 0.60 0.28
Figure B-27
Three-dimensional trusses

Commentary B
Case (a) Case (c)
Cs+ 0.75 Cs+

Cs+ Cs– Cs–

0.75 Cs–
Case (b) Case (d)
0.25 Cs+
0.25 Cs+ 0.25 Cs–
Figure B-28
Full and partial loads due to combined and torsional loads (see Sentence 4.1.8.3.(1))

Commentary C
Structural Integrity
1. The strength and stability of building struc- Safety Measures

tural systems is addressed by Sentence 4.1.1.3.(1)
and by the specific requirements in Part 4 of the 6. Measures to prevent the occurrence of wide-
National Building Code 1995 and in the CSA mate- spread collapse resulting from such accidental
rial design standards referenced in Section 4.3. of events include:
the Code. This commentary provides guidance on (a) Control of accidental events. Such measures
additional considerations for structural integrity as include protective devices (curbs, guards)
addressed in Sentence 4.1.1.3.(1) and Appendix A. against vehicle impact, inspection of key ele-
ments or ground conditions for deterioration
2. Structural integrity is defined as the ability of during use, and blow-out panels to reduce ex-
the structure to absorb local failure without wide- plosion pressures.
spread collapse. For example, a cellular or frame (b) Local resistance. This consists of designing key
arrangement of components well tied together in members to resist accidental events.(2) Some
three dimensions has good structural integrity. major structural members, for example, are so
strong that most accidental events are unlikely
3. Building structures designed in accordance to cause serious structural damage. Ductility of
with the CSA design standards will usually possess the key members and of their connections to
an adequate degree of structural integrity, generally the structure can also provide substantial addi-
through detailing requirements for connections tional resistance to accidents not normally
between components. Situations where structural in- considered in design.
tegrity may require special attention include (c) Design of tie forces. Structural integrity can of-
medium/high rise building systems made of ten be achieved indirectly by providing certain
components of different materials, whose intercon- minimum vertical, horizontal and peripheral
nection is not covered by existing CSA design ties in buildings.(3)(4)(5)
standards, buildings outside the scope of existing (d) Alternate paths of support. Here it is assumed
CSA design standards, and buildings exposed to that the key member has failed, and the dam-
severe accidental loads such as vehicle impact or ex- aged building is checked to ensure that it can
plosion. The following provides guidance for such support the dead load plus a portion of the
situations. live load and wind load.
(e) Control of widespread collapse. This measure
4. A significant number of failures, many of consists of dividing the structure into areas
them progressive, occur during construction. The separated by planes of weakness which pre-
construction sequence should, therefore, be carefully vent a collapse in one area from propagating
planned and monitored to ensure that partially com- into adjacent areas. This method is described
pleted structural systems have sufficient strength, in Commentary M: Structural Integrity of Fire-
ductility and lateral stability to resist progressive walls.
collapse if a construction accident causes significant
damage to a structural element or if local failure of a 7. Any building system should be considered as
permanent or temporary structural element occurs. a whole, and effectively tied together in such a way
as not to be sensitive to local accidental failure.
Identification of Hazard(1)
8. Additional information for specific building
5. The hazard is the risk of widespread collapse
structural systems is contained in References (3) to
with serious consequences arising from local failure (9). Reference (6) includes additional references for
caused by accidental events not addressed by the concrete building systems.
loads specified in Part 4. Key components which
can be severely damaged by an accident with a sig-
nificant probability of occurrence (approximately
10-4 per year or more) should therefore be identified,
and measures taken to ensure adequate structural
safety.

Commentary C
References
(1) H. Griffiths, A. Pugsley and O. Saunders, Re-
port of the Enquiry into the Collapse of Flats
at Ronan Point, Canning Town. Her Majesty’s
Stationary Office, London, 1968.
(2) Dansk Standard DS 410 - English Translation.

Loads for the Design of Structures: Chapter
17 - Accidental Action. Dansk Ingeniorforen-
ing, Copenhagen, 1983.
(3) J.E. Breen, Developing Structural Integrity in
Bearing Wall Buildings. Journal of the Pre-
stressed Concrete Institute. Vol. 25, No. 1,
January-February 1980, pp. 42-73.
(4) M. Fintel and G. Annamalai, Philosophy of
Structural Integrity of Multistorey Load-
bearing Concrete Masonry Structures.
Concrete International, Vol. 1, No. 5, May
1979.
(5) I.J. Speyer, Considerations for the Design of
Precast Concrete Bearing Wall Buildings to
Withstand Abnormal Loads. Journal of the
Prestressed Concrete Institute, Vol. 21, No. 2,
March-April 1976.
(6) Canadian Portland Cement Association.
Structural Integrity. Concrete Design Hand-
book, Ottawa, 1985, pp. 1-25 to 1-30.
(7) British Standards Institute. BS 5628: Code of
Practice for the Structural Use of Masonry:
Part 1 – Unreinforced Masonry: Section 5 –
Design: Accidental Damage, London, 1978.
(8) B.R. Ellingwood and E.V. Leyendecker, Ap-
proaches for Design Against Progressive
Collapse. Journal of Structural Division, Proc.,
Am. Soc. Civ. Eng., March 1978, pp. 413-423.
(9) D.A. Taylor, Progressive Collapse. Can. J. Civ.
Eng., Vol. 2, No. 4, December 1975.

Commentary D
Effects of Deformations in Building
Components
Structural Effects gradient in temperature through the thickness of

materials or assemblies. Again this may complicate
1. When building materials expand and contract the assessment of deformations or stresses, but a ra-
due to temperature changes, considerable forces tional judgment must be made in design if building
may be produced in restrained structural elements, elements are to perform in a satisfactory manner.
i.e., those elements that are not free to expand and
contract with the changes in temperature. Often 3. If these forces are not properly considered,
these forces are compounded by those produced by the stresses resulting from such forces can lead to
shrinkage, creep and moisture content changes and serious failures (usually cracking) in materials and
are therefore difficult to analyse or predict. In many structural members. Failures occur when clearances
situations, however, the structural designer must are insufficient, when fasteners do not allow move-
consider the probable structural effects of the forces ment or deformations, or, in the case of restrained
produced by temperature changes along with all elements, when the elements are not strong enough
other forces; indeed the designer is required to do to withstand the stresses induced. An elementary
so by Sentence 4.1.2.1.(1) of the National Building review of thermal and moisture deformations in
Code of Canada 1995. buildings is given in References (1) and (2).
Table D-1 indicates the order of magnitude of move-
2. In addition to expansion and contraction, ment to which various materials are liable. Actual
temperature changes may produce differential de- values can vary significantly from those in the Table.
formation or warping of materials as a result of a
Table D-1
Typical Deformation Properties for Some Common Building Materials
Moisture Movement, mm/m Modulus of Creep

Material Thermal Movement,
Elasticity, Coefficient,(1)
mm/m per 100 C Permanent Reversible MPa 103
Plainconcrete(2)
normal weight 1.0 0.5 ± 0.1 30 3
Glass 0.9 0 0 70 0
Masonry(3)
clay 0.7 –0.2 ± 0.1 20 1
(expansion)
calcium silicate 1.0 0.2 ± 0.1 15 2
concrete
normal weight 1.0 0.4 ± 0.2 15 2
lightweight (autoclaved) 1.0 0.4 ± 0.2 10 2
aerated (autoclaved cellular) 1.0 0.7 ±0.2 6 2
Metal
aluminum 2.4 0 0 70 0
copper 1.7 0 0 110 0
lead 3.0 0 0 14 0
steel 1.2 0 0 200(4) 0

Commentary D
Table D-1 (Continued)
Moisture Movement, mm/m Modulus of Creep

Thermal Movement,
Material Elasticity, Coefficient,(1)
mm/m per 100 C Permanent Reversible MPa 103
Natural stone
limestone 0.4 — ± 0.1 60 0
marble 0.5 — ± 0.1 35 0
sandstone 1.2 — ± 0.3 20 0
Wood (spruce-pine-fir)
across grain
radial 4.0 30(5) 1
± mc(5) 1 (6)
tangential 6.0 50(5) 1

±2 mc(5) 0.5 (6)
parallel to grain 0.4 1(5) 1

± mc/30(5) 10 1
Notes to Table D-1:

(1) Deformation under sustained loading = short term deformation based on modulus of elasticity 2 (1 + ).
(2) For reinforced concrete see CAN3-A23.3-94, “Design of Concrete Structures for Buildings.”
(3) For further information see CAN/CSA-S304.1-94, “Masonry Design for Buildings (Limit States Design).”
(4) For cold-formed steel see CAN/CSA-S136-94, “Cold Formed Steel Structural Members.”
(5) 1
Initial drying from green condition to equilibrium is assumed to be 12%; mc = per cent change in moisture content from 12%.(4)
(6) Such application is usually avoided.
Design Temperature Ranges Table D-3

Temperature Decrease below Ambient Temperature due to
4. In a country like Canada, with its many cli- Radiation Loss into a Dark Clear Sky
matic regions, the extremes of air temperature that
have to be considered in the design of exteriors of Surface
Temperature Loss, C
buildings vary greatly. One way of approaching this
problem is to use temperature maps like those given Dark roofing 10
in the Ontario Highway Bridge Code(3) giving Steel and other metal 5 – 10
maximum summer and minimum winter air tem- Concrete and masonry 5
peratures. Such a detailed approach may not be
necessary for buildings. Instead, the 2.5% July and
January air temperatures for the design of cooling 6. The values in Table D-2 will depend on the
and heating systems in the Table “Design Data for colour, slope, orientation and insulation backing of
Selected Locations in Canada” in Appendix C of the the surface.
NBC 1995 are suggested. This will be illustrated by
the three examples below. Examples: For a horizontal dark-coloured metal
surface in three typical climate regions (coastal, cen-
5. Because of solar heat gain in summer and tral and interior), the range of temperatures for
radiation heat loss in winter, the range of design purposes might be as follows:
temperatures that building elements undergo is Coastal (Victoria):
greater than the ambient air temperature. Tables D-2 (24(i) + 25(ii)) − (− 5(iii) −10(iv)) = 64C
and D-3 show typical annual ranges of temperature Central (Ottawa):
differences between such elements and ambient air (30(i) + 25(ii)) − (−25(iii) − 10(iv)) = 90C
temperatures due to these effects.(2) Interior (Regina):
(31(i) + 25(ii)) − (−34(iii) − 10(iv)) = 100C
Table D-2 7. Except for the very temperate parts of

Temperature Increase in Excess of Ambient Canada referred to as Coastal, as a simple rule, one
Air Temperature due to Solar Radiation may assume a range of exterior surface tempera-
tures of about 100C for a horizontal relatively dark
Surface material. Because of thermal insulation, thermal
Temperature Gain, C
Dark roofing 20 – 40 (i) July 2.5% temperature.
Steel and other metal 15 – 25 (ii) Dark metal temperature gain.
Concrete and masonry 10 – 15 (iii) January 2.5% temperature.
(iv) Dark metal temperature loss.

Commentary D
inertia and other factors, however, the range of ex- (2) Estimation of Thermal and Moisture Move-
treme temperatures in structural components of a ments and Stresses. Parts 1-3, British Building
certain thickness will often be somewhat smaller Research Establishment Digests No. 227-229.
than those in the preceding examples. Building Research Station, Garston, Watford,
Great Britain, August 1979.
8. Temperature variations can be particularly (3) Ontario Highway Bridge Design Code 1984.
significant in multi-storey apartment and office Ontario Ministry of Transportation and Com-
buildings with exterior columns partially, and in munications, Toronto, 1984.
some cases fully, exposed to the weather. Exposed (4) W.G. Plewes, Failure of Brick Facing on High-
columns, when subjected to seasonal temperature Rise Buildings. CBD 185, Division of Building
variations, change their length relative to interior Research, National Research Council Canada,
columns, which remain unchanged in a controlled Ottawa, April 1977.
environment. Although in low buildings this causes (5) Principles of Modern Building, Vol. I. Building
insignificant structural problems, in tall buildings Research Station of DSIR. Her Majesty’s Sta-
temperature stresses become significant and must be tionery Office, London, 1959 (in particular see
investigated thoroughly. Chapter 2 on Dimensional Stability).
(6) D.G. Stephenson, Extreme Temperatures at the
9. Dimensional changes occur not only as the Outer Surface of Buildings. CBD 47, Division
result of temperature changes, but also from shrink- of Building Research, National Research
age, moisture content changes, chemical processes Council Canada, Ottawa, November 1963.
and creep deformation in the component materials (7) F.R. Khan and M. Fintel, Effects of Column
of a building. If the building or component is not Exposure in Tall Structures. Paper in three
free to contract or expand, tensile or compressive parts. (a) Temperature Variations and Their
stresses result. These stresses can be relieved or Effects, (b) Analysis of Length Changes in
reduced to tolerable limits by contraction and ex- Exposed Columns, and (c) Design Considera-
pansion joints. Such joints are particularly important tions and Field Observations of Buildings.
to allow contraction to take place along certain pre- Journal of American Concrete Inst., Vol. 63,
selected lines rather than to produce cracks along No. 8, August 1966 and Vol. 65, No. 2, Febru-
accidental lines of least resistance. ary 1968.
(8) P. Weidlinger, Temperature Stresses in Tall
Effects on Cladding Reinforced Concrete Buildings. Civil Engi-
neering, New York, Vol. 34, No. 8, August
10. In the design of all buildings, but particu- 1964.
larly very long and very high buildings, the effects (9) K. Jones, Restraint of Structures Attached to
of movements of the structural members on the Mass Concrete. Journal of Structural Division,
cladding elements should be considered. Shortening Am. Soc. Civ. Eng., Vol. 87, No. ST8, Decem-
and lengthening of columns due to temperature and ber 1961.
shrinkage effects and creep can crack, buckle or oth- (10) W.T. Marshal, Shrinkage and Temperature
erwise overstress cladding materials and their Stresses in Reinforced Concrete. Civil Engi-
fastenings. Deflections and linear movements of neering, London, Vol. 56, No. 665, December
beams and spandrels and building sidesway can 1961.
have similar effects. Failure to consider these differ- (11) P. Fisher, Differential Temperature Movements
ential movements has caused many cases of in Rigid Frame. Journal of American Concrete
cladding damage. For example, spalling, cracking Inst., Vol. 59, No. 6, June 1962.
and bulging have occurred to brick and stone veneer (12) D.W. Allen, The Calculation of Temperature
on a number of tall concrete buildings,(4) necessitat- Stresses. Concrete & Constructional Engineer-
ing extensive repairs. The phenomenon is not, ing, Vol. LVII, No. 9, September 1962.
however, limited to concrete frames, nor are the ef- (13) G.L. England and A.D. Ross, Reinforced Con-
fects limited to stone and brick cladding. References crete under Thermal Gradients. Magazine of
(5) to (23) discuss these effects in greater detail. Concrete Research, Vol. 14, No. 40, March
1962.
Bibliography of the Effects of (14) J.H. Slack and M.J. Walker, Movement Joints
Deformations on Structures in Concrete. Concrete Society Limited,
Grosvenor Gardens, London, Technical Paper,
(1) M.C. Baker, Thermal and Moisture Deforma- 1967.
tions in Building Materials. CBD 56, Division (15) Deflections of Reinforced Concrete Flexural
of Building Research, National Research Members. Report of ACI Committee 435, ACI
Council Canada, Ottawa, August 1964. Manual of Concrete Practice 1970, Part 2.

Commentary D
(16) H. Mayer and H. Rusch, Building Damage

Caused by Deflection of Reinforced Concrete
Building Components. Deutsher Ausschuss
fur Stahlbeton, Heft 193, Berlin 1967, National
Research Council Technical Translation
TT1412.
(17) W.G. Plewes, Cladding Problems Due to

Frame Movements. CBD 125, Division of
Building Research, National Research Council
Canada, Ottawa, May 1970.
(18) R.E. Copeland, Flexible Anchorage of Ma-
sonry Walls. Concrete Products, Vol. 71, No. 7,
1968, p. 54.
(19) M. Fintel and F.R. Khan, Effects of Column
Creep and Shrinkage in Tall Structures – Pre-
diction of Inelastic Column Shortening.
Journal of American Concrete Inst., December
1969, Proc. V66, No. 12, p. 957.
(20) D. Foster, Some Observations on the Design
of Brickwork Cladding to Multi-storey R/C
Framed Structures. BDA Tech. Note, Vol. 1,
No. 4, September 1971, The Brick Develop-
ment Association, 3-5 Bedford Row, London
WC1 4BU.
(21) CIRIA, Movement and Cracking in Long Ma-
sonry Walls. Practice Note, Special Publication
44, Construction Industry Research and Infor-
mation Association, London, U.K., 1986.
(22) S.J. Alexander and R.M. Lawson, Design for
Movement in Buildings. CIRIA Technical
Note 107, Construction Industry Research and
Information Association, London, U.K., 1981.
(23) A.T. Hansen, Effects of Wood Shrinkage in
Buildings. Canadian Building Digest 244,
Institute for Research in Construction,
National Research Council Canada, Ottawa,
February 1987.

Commentary E
Load Combinations
Introduction mining load combinations can be determined by the

following principle.(1) Each load can be separated
1. Article 4.1.3.2. and Subsection 4.1.4. of the into two components, one a sustained or frequently
National Building Code of Canada 1995 are in- occurring component (e.g., dead load, storage), the
tended to provide an acceptable and relatively other a transient component which acts rarely, for a
uniform degree of safety in the design of structural short time only (e.g., impact, wind, earthquake, rare
members under different load combinations. The accumulation of people or equipment). Since tran-
rules pertain to the ultimate limit states (allowable sient loading components are unlikely to occur
stresses, required structural resistance) and not, in simultaneously, the critical load combination for any
general, to serviceability considerations such as de- structural effect can be estimated by combining the
flection. The rules take into consideration the load having the largest transient component with
probability of simultaneous occurrence of the design the sustained or frequent components of all other
loads stipulated in NBC Subsections 4.1.5. to 4.1.10.. loads. This principle can be applied in place of the
They do not take into account the change in mate- load combination factor both for limit states design,
rial strengths with changes in duration of loading. where the loads are factored, and for working stress
design, where the loads are not factored.
Load Combinations
Counteracting Loads
2. As dead load is nearly constant throughout
the life of a structure, a combination of dead load 5. Counteracting loads cause overturning, uplift,
with any other load constitutes a combination in sliding of structures as a whole and stress reversal
which the basic safety or load factors apply. When in structural members. For these cases, the dead
dead load is combined with two or more other load acts to resist failure and deviations which de-
loads, the full design values of each of the load ef- crease rather than increase the dead load are critical.
fects are less likely to occur simultaneously than is For limit states design, an underload factor of 0.85
the basic combination above. Sentences 4.1.3.2.(5) is, therefore, applied in accordance with NBC
and 4.1.4.2.(1) of the NBC take this into considera- Sentence 4.1.3.2.(2) as follows:
tion by allowing reductions in the total effect due to
combinations of the dead load with two or more
other loads. (1)
3. Because of the very short duration of some
design loads, the probability of their simultaneous where R is the factored resistance of the anchorage
occurrence is extremely small. Thus, according to or member component. For the case of earthquake
NBC Sentences 4.1.3.2.(3) and (8), earthquake load is load combinations, Sentence 4.1.3.2.(8) applies, i.e.,
not considered simultaneously with wind load. Fur-
thermore, a reduced load factor is given for live
load in combination with earthquake load to ac- (2)
count for the lower value of live load expected to be
acting when the earthquake event occurs. (See also
paragraph 13 of Commentary F). For industrial where an underload factor of 1.0 is used to account
buildings with multiple crane operations, the hori- for the greater expected dead load plus gravity live
zontal crane load should be determined on the basis load that would be acting when the earthquake
that the largest crane only is operating at maximum event occurs.
capacity in the most critical location.
6. Working stress design does not take into ac-
4. For building structures subjected to unusual count the need for an underload factor for dead
load combinations, for example those involving the load and, therefore, special rules are required to
use of heavy equipment or the storage of liquids, provide sufficient safety.(2)(3) NBC Article 4.1.4.4. re-
the simple rules of NBC Article 4.1.3.2. or Subsection quires a safety factor of not less than 2.0 on loads
4.1.4. may not apply. In such cases the rule for deter- tending to cause overturning or sliding. In cases of

Commentary E
uplift or stress reversal, special rules may be re-

quired which essentially reflect Equation (1).
7. With regard to overturning, the designer

should consider that: (1) unless the foundation ma-
terial has a high strength, the point of overturning
is not at the toe of the building, and (2) for flexible

structures the dead load acts through the centre of
gravity of the deflected structure.
8. A special situation arises in the case of over-

turning or sliding during an earthquake. Since the
lateral earthquake force depends on the combined
weight of the structure and its contents, two cases
should be considered in applying Equation (1): one
with the structure empty, where dead load only acts
to resist overturning, and one with the structure
full, where dead load plus the contents act to resist
overturning. In the latter case, D in Equation (1)
should also include the contents.
Full and Partial Loading

9. Full and partial loading considerations are re-
quired in accordance with NBC Article 4.1.6.3. for
floor live load and Sentence 4.1.7.2.(2) for snow
load. For limit states design, such pattern loading
requirements should be considered in conjunction
with the dead load multiplied either by 1.25 on all
spans or 0.85 on all spans, whichever produces the
most unfavourable effect.
References
(1) C.J. Turkstra, Theory of Structural Safety. SM
Study No. 2, Solid Mechanics Division, Uni-
versity of Waterloo, Waterloo, Ontario, 1970.
(2) D.E. Allen, Safety Factors for Stress Reversal.
Publications, International Association for
Bridge and Structural Engineering, Vol. 29/II,
1969.
(3) Report of the Committee of Inquiry into Col-
lapse of Cooling Towers at Ferrybridge,
Monday, 1 November 1965. Central Electricity
Generating Board, London.

Commentary F
Limit States Design
1. Subsection 4.1.3. of the National Building in some cases, a stability analysis for overturning.
Code of Canada 1995 allows the use of limit states The appropriate theory will either be indicated in
design as an alternative to existing procedures for the structural material standard or chosen by the
design calculations of building structures. This designer.
Commentary describes what is meant by limit states
design and the reasons for introducing it into the 6. Furthermore, existing design methods em-
National Building Code and associated CSA phasize only one limit state, usually associated with
Structural Standards. In addition, a background ex- a limiting stress or member strength. Due to
planation is given of the safety and serviceability changes to lighter composite-acting construction
criteria contained in NBC Subsection 4.1.3. The cri- with less in the way of stiffening and damping from
teria were produced by a CSA/NBC Joint Liaison curtain walls and partitions, serviceability require-
Committee on Limit States Design, which repre- ments such as deflection and vibration are
sented all structural and foundation standards used becoming more critical in structural design, and
by the National Building Code of Canada. deserve the same consideration as strength require-
ments. In contrast to existing design methods, limit
Limit States states design applies to all kinds of failure such as
collapse, overturning and vibration, and to all mate-
2. All building structures have the same basic rials and types of construction.
functional requirements, namely that they should be
safe from collapse during construction and that they 7. In summary, limit states design provides a
should be safe from collapse and be serviceable dur- unified rational basis for design calculations of
ing the useful life of the building. The onset of building structures of all materials. This is the main
various types of collapse and unserviceability are reason that it is being adopted for international
called limit states. Those concerning safety are standardization.(1)
called ultimate limit states and include the exceed-
ing of load-carrying capacity, fracture, overturning, Safety and Serviceability Criteria
sliding and large deformation. Those concerning
serviceability are called serviceability limit states, 8. As already stated, the aim of limit states de-
and include excessive deflection, permanent defor- sign calculations is to prevent failure, that is, the
mation, cracking and vibration. attainment of a limit state. Unpredictable factors
such as loads and workmanship enter into the cal-
3. The primary aim of limit states design is to culations, however, and so the further qualification
prevent the attainment of limit states, that is to pre- is added “that the probability of failure be suffi-
vent various kinds of failure. This should be clearly ciently small.” The more serious the consequences
understood by the designer when reading and inter- of failure, the smaller should be the probability. Sat-
preting structural standards, since any detailed isfactory failure probabilities are achieved through
requirement is aimed at preventing the attainment the use of reliable materials, through competent
of a particular limit state. structural engineering, manufacture and erection,
and by the use of safety and serviceability criteria in
4. Limit states design is not new; it is basically a the design calculations. The safety and serviceability
clarification of well-known principles. There is, criteria should provide adequate human safety and
however, a change of emphasis. serviceability on the one hand, and economy on the
other hand, i.e., optimum or smaller failure proba-
5. Existing design methods – allowable stress bilities.(2) This is achieved in limit states design
design, plastic design, ultimate strength design – through the statistical definition of specified loads
put the main emphasis on a particular structural and material properties and the use of partial fac-
theory such as elastic or plastic theory. No particu- tors.
lar theory, however, applies universally to all limit
states and all types of construction. Elastic theory is 9. The general form of safety and serviceability
generally applicable for serviceability limit states criteria contained in NBC Subsection 4.1.3. can be
and fatigue, plastic theory for ultimate limit states expressed as follows:

Commentary F
A special situation arises in cases of overturning,

(1) uplift and stress reversal, where the load effects
tending to cause failure are counteracted by the
dead load effect. For such cases positive anchorage
for load combinations except earthquake, and is required if the factored load effects tending to
cause failure are greater than the stabilizing effect of
(2) the dead load multiplied by a dead load factor of

0.85 in Equation (1) and 1.0 in Equation (2).
for earthquake load combinations, where D, L, W, E 12. For the serviceability limit states, instead of
and T are the specified loads (dead, live, wind, a factored resistance, R represents a criterion such
earthquake and temperature) defined in NBC as an allowable deflection, acceleration or crack
Sentence 4.1.2.1.(1), width. Equation (1) therefore results in serviceability
requirements of the same type as in the past. It is
= load factor applied to one of the speci- important, however, to understand which limit state
fied loads which takes into account a particular criterion is attempting to prevent.
variability of the load and load patterns,
bias in the relationship between the 13. A new load combination for earthquake
nominal load and the expected value of (Equation (2)) different from that for wind (Equation
the load for the event considered and, to (1)) was introduced because an earthquake causing
some extent, inaccuracy in the structural structural damage is a rare event for any location in
analysis, Canada and because an earthquake, a ground shak-
ing event, is an entirely different kind of load than
= load combination factor applied to loads wind, an applied pressure. The design approach for
other than dead load to take into account earthquake, based on ductile behaviour (capacity
reduced probability of simultaneous oc- design principles) and deformation control, as well
currence of loads from different sources, as strength, is, therefore, different than for wind.
Furthermore, the prediction of earthquake forces
= importance factor applied to loads other and earthquake response by the procedures of NBC
than dead load which takes into account Subsection 4.1.9. is much more uncertain than the
the consequences of collapse as they re- prediction for wind forces in NBC Subsection 4.1.8.
late to the use and occupancy of the Equation (2) represents a load combination
building, i.e., the danger to human recommended by ISO,(1) which is appropriate for ac-
safety and the economic loss, cidental events such as earthquake, where loads
other than earthquake are represented by their ex-
R = calculated resistance of a member, pected values during the design rare event. For
connection or structure based on the such rare events, the stability of the structure is
specified material properties, maintained for life safety, although damage is likely
to occur. The load T is not included in Equation (2)
= resistance factor applied to the resistance because this category of forces would be dissipated
or specified material property which by inelastic action. The expected value for live loads
takes into account variability of material other than storage and assembly occupancies is con-
properties and dimensions, workman- siderably less than their nominal values; hence, a
ship, type of failure (e.g., brittle versus load factor of 0.5 is used. The load factors for live
ductile) and uncertainty in the predic- load in Sentence 4.1.3.2.(8), however, are higher than
tion of resistance. those used to determine the expected total weight
of the building in Sentence 4.1.9.1.(2) because struc-
10. When the specified loads and resistances are tural components are subjected to localized
multiplied by the appropriate partial factors, the concentrations of live load.
products are called the factored loads and factored
resistances. The load factors, load combination fac- Definition of Specified Loads
tor and importance factor in the right side of and Resistances
Equation (1) are given in NBC Subsection 4.1.3., the
same for all building structures. The resistance and 14. For limit states design, specified loads and
resistance factor in the left side of Equation (1) are specified material properties used to calculate resis-
contained in the appropriate structural standard, tance are defined on the basis of probability of
different for different materials, type of structural occurrence. Values so defined are called characteris-
element and type of behaviour. tic values. Material properties are controlled by
statistical sampling and the characteristic value cor-
11. In the case of the ultimate limit states, Equa- responds to a limiting probability of unfavourable
tion (1) states that the factored resistance must be test values. Climatic loads are based on measure-
greater than or equal to the effect of factored loads. ments taken at weather stations, and the

Commentary F
characteristic value corresponds to a return period. 18. The partial factors contained in NBC Sub-
Characteristic values for material properties and section 4.1.3. were determined on the following
loads used in the National Building Code are given basis:
in Table F-1. Where statistical information is lacking,
e.g., for live loads, the specified values correspond Load factors were first chosen on the basis of
to the existing nominal values. variability of the loads and load patterns only, ex-
cluding approximations in structural analysis. Based

on the probabilistic assumptions given in Reference
Table F-1 (3) for a 30-year life, the following load factors were
Characteristic Values for Loads and Material Properties in obtained, corresponding to a probability level three
the National Building Code standard deviations above the mean: 1.2 for dead
load and 1.4 for live load and wind load.
Materials Probability Level To take into account approximations in structural
Concrete (cylinder test) 10% analysis, these load factors were increased to 1.25
Wood (tests on small clear specimens) 5% for dead load and 1.5 for live load and wind load.
Steel (yield in tension) Not defined (1–2%) The load factor for imposed strains (temperature,
shrinkage, differential settlement) was taken as 1.25,
Masonry (for prism tests) 10% the same as for dead load. As a general rule, how-
Load Return Period ever, inaccuracy in structural analysis should be
Dead Not defined taken care of by means of conservative assumptions
in the analysis. Inaccuracies in member resistance
Floor Not defined
are taken into account in the factored resistance.
Snow 30 years
Wind The load combination factor, , was determined
ultimate limit states 30 years on the basis of a probabilistic study for the combi-
serviceability limit states 10 years nation of dead, live and wind loads for office or
Earthquake residential buildings.(3) These studies indicate that
ultimate limit states approx 500 years(1) = 0.7 gives a safety consistent with that for the basic
combination of dead load plus live load. The same
Notes to Table F-1: studies also indicate that for overturning, uplift and
(1) See Commentary J.
stress reversal, safety consistent with that for dead
load plus live load is obtained when D = 0.85 for
15. For new materials or new control methods, load combinations without earthquake.
material resistance should be defined on the basis of
the 5% probability level and material stiffness on For buildings of normal human occupancy the
the basis of the 50% probability level; where statisti- importance factor was taken equal to 1.0 for practi-
cal sampling is used, a 75% confidence level is cal purposes, since this classification corresponds to
recommended. most buildings. For post-disaster buildings, impor-
tance factors greater than 1.0 are applied to wind
16. Since the characteristic values in Table F-1 load in NBC Subsection 4.1.8. (by means of in-
refer to standard tests or measurements (for exam- creased return period) and earthquake load in NBC
ple the standard cylinder test for concrete, hourly Subsection 4.1.9. The reason for retaining this
wind speed at an airport), the probability levels can- approach is that the importance factor should be ap-
not be directly applied to what happens in the plied only to those loads which can cause a disaster.
structure without further considerations. For buildings of low human occupancy, such as
farm buildings or storage sheds, an importance fac-
Partial Factors tor less than 1.0 applied to loads other than dead
load is justified on the basis of reduced risk to hu-
17. To provide sufficiently small failure proba- mans. The factor 0.8, along with the reduction in
bilities for the limit state under consideration, limit dead load factor from 1.5 to 1.25, closely corre-
states design makes use of partial factors in contrast sponds to the 25% increase in allowable stress
to the total safety factor used by existing design contained for some years in the National Farm
methods. The use of partial factors gives more con- Building Code of Canada.
sistent safety for different load combinations as well
as for different combinations of materials, with a For the serviceability limit states, in accordance
consequent economy of materials. It also provides a with NBC Article 4.1.3.3., the partial factors are gen-
better basis for development of new types of erally taken as 1.0; an exception is the reduction for
construction or for unusual situations, since all par- load combinations. Two reasons for continuing this
tial factors, including resistance factors for the traditional approach are: the consequences of ser-
well-known basic structural materials, will either be viceability failure are considerably less serious than
known or can be established on a rational basis. for collapse, and therefore the partial factors are

Commentary F
closer to 1.0; since the criteria for serviceability fail-

ures cannot be accurately defined, only simple
empirical rules which absorb safety factors are justi-
fied.
19. The live load factor 1.5 applies to live loads

for most buildings. For live loads whose uncertainty

deviates considerably from those specified in NBC
Article 4.1.2.1., the load factor should be adjusted
accordingly. For example, the uncertainty in weight
or pressure of fluids in storage tanks is similar to
uncertainty in dead load, and a load factor of 1.25
may be appropriate in some cases. Design of such
structures is often governed by a serviceability limit
state of water tightness or by differential settlement.
20. The dead load factor of 1.25 does not take

into account any increase in dead load as a result of
alterations (e.g., to the roof or floor) during the life
of the building.
21. During construction all permanent and tem-

porary structural members should have sufficient
factored resistance to carry the effect of factored
construction loads. Adequate load factors not less
than 1.25 (greater if the construction load is more
uncertain than the dead load), with the importance
factor equal to 1.0, are recommended.
Resistance Factors
22. The resistance factors will be determined by
each structural materials standard as it develops
limit states design. Guidance on the determination
of resistance factors is given in Reference (4).
References
(1) General Principles on Reliability for Struc-
tures. International Standard ISO 2394.
Geneva, 1986, 18 pp.
(2) M.K. Ravindra and N.C. Lind. Theory of
Structural Code Optimization. Journal of
Structural Division, Proc. Am. Soc. Civ. Eng.,
Vol. 99, ST7, July 1973, p. 1541.
(3) D.E. Allen. Limit States Design – A Proba-
bilistic Study, Can. J. Civ. Eng., Vol. 2, No. 1,
March 1975, p. 36.
(4) Guidelines for the Development of Limit
States Design. CSA Technical Publication
S408-1981. Canadian Standards Association,
Rexdale, Ontario.

Commentary G
Tributary Area
1. Because live loads are generally given as uni- lowed when determining the forces that members
formly distributed loads over a floor area, and carry. In determining live load reduction, however,
because dead loads can usually be considered as the following simplifications are recommended.
uniform loads, either over an area or along the
length of a flexural member, design engineers have Decks and Slabs
for years used the concept of tributary area to deter-
mine the loads that beams, girders and columns 7. No live load reduction factors should be ap-
carry. Once the concept is applied to any floor, it is plied to wooden or sheet metal decks, precast units
easily extended for multi-storey columns to any or one-way slabs because of the uncertainty of the
number of floors. degree of lateral distribution of loads.
2. Earlier design standards recognized that the 8. The tributary area for a flat slab or the slab
probability that all the floors of a multi-storey portions of two-way slabs with beams is the area
building would be loaded to the full live load si- bounded by column lines or by a combination of
multaneously was very remote. Therefore, to design column lines and lines of supporting members such
the columns for the full live load of a number of as beams and girders, whichever is the lesser, as
floors was unduly restrictive, and reductions in the shown in Figures G-1, G-2 and G-3.
live load were devised as a function of the number
of floors supported by the columns. Beams and Girders
3. In the 1960 edition of the National Building
9. The tributary area for a member supporting a
Code, with the recognition that the average live
portion of a floor is the area enclosing the member
load was a function of the area supported, the ratio-
and bounded by the lines of zero shear in the mem-
nalization was carried one step further and a
bers supported. For buildings with fairly regular
reduction of 15% was allowed for beams, girders
bays the lines of zero shear in the members sup-
and trusses supporting areas greater than 20 m2.
ported can be assumed to be half-way between lines
4. In subsequent editions, provisions have been of support. Figures G-2 and G-3 illustrate the tribu-
included for live load reduction based on tributary tary area of beams supporting two-way slabs.
areas with two different expressions, one for office Figures G-4 and G-5 illustrate the tributary areas for
and apartment buildings and the other for storage joists, beams and girders supporting a one-way slab.
or similar areas.
Negative Moments in Continuous
5. Therefore, for determining the total dead load Members
to be supported by a given member, and to deter-
mine what live load reduction factor should be 10. Tributary area for negative moment over a
applied, a clear definition of tributary area, about support may be taken as the sum of the tributary
which some confusion existed, is needed. areas of the beams on either side of the support. For
cantilever, drop-in beam systems, the tributary area
6. In the case of a member which supports the for calculation of the live load reduction factor, as it
load directly, such as a slab, the tributary area is de- affects the negative moment at the support, is taken
fined as the area supported by the member bounded as the sum of the areas tributary to the cantilevered
by the lines of support. In the case of a member section and one half the length of the drop-in
which does not support the load directly but sup- section.
ports other members, the tributary area is defined
as the area bounded by the lines of support of the Columns
member and the lines of zero shear in the members
supported, assuming a uniformly distributed load is 11. For a column the tributary area per floor is
acting on the structure. These definitions, which for the area of floor supported, bounded by the lines of
continuous construction require a structural analysis zero shear. For buildings with fairly regular bays
to determine locations of zero shear, should be fol- these can be assumed to be half-way between the

Commentary G
column lines, as shown by the dotted area in Fig-

ures G-1 to G-5. In structures with beams, joists or
girders the tributary area per floor is half the sum of
the tributary areas of each of the floor members
framing into it.
12. In multi-storey buildings the tributary area

for a column supporting one use and occupancy is
the sum of the tributary areas per floor for that col-
umn on all levels above the storey in question.
13. For a column supporting more than one use

and occupancy, NBC Article 4.1.6.9. requires that the
Figure G-3
tributary area for each use and occupancy be con-
sidered separately for determining reduction in live Tributary areas for a two-way slab with joists, beams and
load and that the area supporting snow load, which girders
has no reduction, not be included.
Flat and two-way slabs
slab span
Joist, beam or girder
Column
l4
l4
2
l3
2
l3
l1 l2 Figure G-4
2 2
l1 l2 Tributary areas for a one-way slab with girders
Figure G-1
joist girder
Tributary areas for flat slabs without beams and girders
deck span beam

joist span
l1 l2
4 3
45°
l1 l2
Figure G-2 Figure G-5

Tributary areas for a one-way deck or slab with joists,
Tributary areas for a two-way slab with beams
beams and girders

Commentary H
Snow Loads
1. Snow loads on roofs vary according to geo- snow load with a 1-in-30 annual probability of
graphical location (climate), site exposure, shape exceedence, Ss, based on measured depths and den-
and type of roof, and also from one winter to an- sities and a load, Sr, due to the associated rain
other. To account for these varying conditions, the which may fall into the snow cover (not including
specified snow load, S, on a roof or other surface is any rainfall that exceeds the weight of the snow
expressed in Subsection 4.1.7. of the National cover).(2) (See Paragraph 5.)
Building Code 1995 as the sum of two components,
one being the product of a series of factors The snow loads for a given town or city are for
the exact latitude and longitude defined in the
Canada Gazetteer(3) for that town or city. In cities
having a large change in elevation, snow loads may
vary within the city. Maps from which Ss and Sr can
be obtained for locations not listed in Appendix C
where are available from the Atmospheric Environment
Service, Environment Canada, 4905 Dufferin Street,
Ss = ground snow load in kPa with a 1-in-30
Downsview, Ontario, M3H 5T4.
probability of exceedence per year,
Variations with Climate
Sr = the associated rain load in kPa. However,
the rain load at any location on a roof 3. The wide climatic variations across the coun-
need not be taken greater than the load try produce large variations in snow conditions. The
due to snow (i.e., Sr ≤ Ss(CbCwCsCa)). heaviest snow loads occur in the mountainous re-
gions of British Columbia and Alberta; they last the
Cb = the basic roof snow load factor,
entire winter and vary considerably with elevation.
Cw = the wind exposure factor, In certain limited coastal locations of British Colum-
bia, little drifting of snow occurs. The Prairie
Cs = the roof slope factor, and provinces, Yukon and the Northwest Territories
have very cold winters, with small annual snowfalls
Ca = the accumulation factor. but frequent strong winds, which cause consider-
able drifting of snow on roofs and on the ground.
The factors are discussed separately in this Com- The region that includes Ontario, Quebec, and inte-
mentary and a series of figures are provided to rior regions of the Atlantic provinces is marked by
illustrate their application to various shapes of moderate winds and snowfalls, and sufficiently low
roofs. The factors are based on measurements temperatures in most places to allow snow accumu-
obtained during surveys of snow on roofs, on ana- lation all winter. In this region, moderate uniform
lytical studies of the loads on large flat roofs, and and high drift loads occur. Also, cold northwesterly
on judgment. Since surveys of sufficient length winds often cause locally heavy snowfalls to the lee
(about 10 years or more)(1) cover only a limited se- of bodies of water such as the Great Lakes and the
lection of the most common and simplest roof Saint Lawrence River, which lead to increased snow
shapes, the factors are of limited accuracy and may loads.
be subject to change as more data become available.
Local Variations – Mountainous Areas
Snow Loads on the Ground
4. In mountainous areas ground snow loads in-
2. In Canada, ground snow loads are used as a crease with elevation. Observations by the Institute
basis for the determination of roof snow loads. for Research in Construction of the National
Therefore, they form part of the basic climatic infor- Research Council on a number of mountains in
mation needed for building design and are given in British Columbia indicate significant increases in
Appendix C of the NBC 1995 in the Table “Design ground snow load with increases in elevation, de-
Data for Selected Locations in Canada.’’ Each pending on the local topography and climate.(4)
ground snow load is composed of a load due to the Individual mountains or groups of mountains may

Commentary H
cause significant changes in local or micro climate radiation has little effect in reducing loads in cold
within short distances. Hence, snow loads listed in weather. Similarly, during cold weather, heat loss
Appendix C apply only at a particular elevation at from roofs is not very effective in melting the snow,
the individual location as defined by the name and particularly on well insulated and well ventilated
latitude/longitude coordinates given by the roofs. These two factors cannot, therefore, be relied
Gazetteer of Canada.(3) For locations not listed, the upon to reduce the snow load significantly during
Atmospheric Environment Service should be con- the colder periods. During thaws and toward the
sulted for specific recommendations. end of winter, however, when air temperatures ap-
proach the freezing point, solar radiation and heat
Unit Weight of Snow on the Ground loss do result in melting.
5. Falling snowflakes usually consist of very Roof Snow Load Factors
large complex ice crystals. Because of their large
surface area to weight, they fall to the ground rela- 8. The factors Cb, Cw, Cs and Ca were not ob-
tively slowly. On arrival, this snow accumulates in a tained by rigorous statistical analyses due to the
loose and fluffy layer with a unit weight of about lack of data, but they have been found to give ac-
0.5 to 1.0 kN/m3. Immediately, however, the snow ceptable and conservative designs.
crystals start to change: the thin, lacy needle-like
projections begin to sublime and the crystals be- 9. Basic roof snow load factor, Cb. The basic
come smaller irregularly shaped grains. Settlement roof load has been set at 80% of the ground load
of the snow results and the unit weight, , increases (i.e., Cb = 0.8). This figure is based on the results of
after a short time to about 2.0 kN/m3 or greater, a countrywide survey of snow loads on roofs car-
even at below freezing temperatures. The unit ried out by the Institute for Research in
weight of the snowpack continues to increase with Construction and a number of volunteers.
age, ranging from 2.0 to 5.0 kN/m3. As explained in
Appendix C, average values for the seasonal snow 10. Wind exposure factor, Cw. Observations in
pack have been derived for different regions across many areas of Canada have shown that where a
the country for use in the ground snow load calcu- roof or a part of it is fully exposed to wind, some of
lations.(2) The snow surveys from which the unit the snow is blown off or prevented from accumulat-
weight is derived are made four times per month ing and the average snow load is reduced.
(at most). While the survey measurements reflect to
some extent the portion of rainfall that is trapped in 11. Therefore, for roofs fully exposed to the
the snowpack over a period of time, only a small wind, the wind exposure factor, Cw, may be taken
proportion of measurements would have been made as equal to 0.75 rather than 1.0 (or 0.5 rather than
directly after a rainfall. Therefore, the measurements 1.0 for exposed sites north of the treeline). This sub-
probably do not adequately represent the short term stitution applies under the following conditions:
density increase due to the wetting of snow by rain, (a) the building is in an open location containing
and for this reason, the rain load, Sr, is included in only scattered buildings, trees or other such
the calculation of roof snow loads.(2) obstructions, so that the roof is exposed to the
winds on all sides and is not shielded by
Snow Loads on Roofs obstructions higher than the roof within a dis-
tance from the building equal to 10 times the
Unit Weight of Snow on Roofs height of the obstruction above the roof level;
6. In calculation of loads due to snow on roofs a (b) the area of roof under consideration does not
measurement or good estimate of the unit weight is have any significant obstructions such as para-
necessary. The unit weight of snow on roofs, , ob- pet walls within a distance of at least 10 times
tained from measurements at a number of stations the difference between the height of the ob-
across Canada varied from about 1.0 to 4.5 kN/m3. struction and CbCwSs/ metres;
An average value for use in design in lieu of better (c) the loading case under consideration does not
local data is = 3.0 kN/m3. (5) In some locations the involve accumulation of snow due to drifting
unit weight of snow may be considerably greater from adjacent surfaces such as, for example,
than 3.0 kN/m3. Such locations include regions the other side of a gable roof.
where the maximum roof load is reached only after
contributions from many snowstorms, coastal A value for Cw of 1.0 must be applied to other
regions, and regions where winter rains are consid- than the loadings marked Case 1 in Figures H-1 to
erable and where a unit weight as high as H-3.
4.0 kN/m3 may be appropriate.
12. The value CbCwSs/ is the height of
Solar Radiation and Heat Loss uniformly distributed snow on a roof without ob-
structions, including parapets. Any obstructions
7. Some factors which modify snow loads occur lower than this height do not generate additional
only under special conditions. For example, solar snow loading.

Commentary H
13. In practice it is sometimes difficult to make gutters, heated drips or some other means to
a clear distinction between roofs that will be fully prevent the growth of dangerous icicles due to melt-
exposed to the winds and those that will not. The water from the snow retained on roofs may also be
designer should, in consultation with the owner, required. Snow and ice falling from the roof of a
weigh the probability of the roof becoming shel- building may be deflected against the building,
tered by an addition to the building or by adjacent causing damage.
higher buildings or trees. Such changes could cause

either drift loads or higher average loads. In consid- 19. Accumulation factor, Ca. The accumulation
ering drift loads, which are the more serious, a factor, Ca, is described in Figures H-1 to H-6 for a
minimum distance of at least 5 m should be main- number of different roof shapes. For cases where
tained from another existing or future building or Figures H-1 to H-6 do not apply, accumulation fac-
from the property line to justify disregarding drift tors should be determined by the designer based on
loads. This corresponds to the distance used in NBC applicable field observations, special analyses usu-
Subclause 4.1.7.1.(7)(c)(iii) for multi-level roofs. With ally accounting for local climate effects(11)(12)(13) or on
regard to higher average loads, it is important to model tests.(6) In an effort to provide guidance, the
use a wind exposure factor, Cw, equal to 1.0 for any Institute for Research in Construction has published
roof area whose exposure may decrease. two collections of interesting non-uniform snow
loads as case histories.(14)(15)
14. The designer should also be aware that the
snow loads on the roof of an existing building on 20. Effect of wind on snow accumulation on
the same or adjacent property may also be affected roofs. When the wind encounters obstructions, re-
by the location of a new higher building or other gions of accelerated and retarded flow result. The
obstruction. regions of retarded flow are said to be regions of
“aerodynamic shade.”(16) Because a minimum veloc-
15. The installation of solar collectors on roofs ity is required to transport the snow, it settles out
may result in reduced exposure similar to that where the flow velocity is too low and forms drifts
around obstructions unless the clear gap under whose shapes are indicated by Ca. In general, the
them is sufficiently large to allow scouring and re- longer the wind duration, the deeper are the drifts
moval by the wind rather than deposition.(6)(7) on roofs, especially if it is also snowing, and the
greater the wind speed, the less uniform are the
16. Regional variation in Cw. For the exposure snowdrifts. For large roofs the wind is less effective
factor to have any application there must be wind. in removing snow due to the greater quantities in-
Therefore, designers should use Cw = 1.0 in the few volved.(17) Increased values of Ca are specified,
areas of Canada, such as winter-calm mountain val- therefore, in NBC Clause 4.1.7.1.(7)(b) to account for
leys, where winter winds are not strong or frequent this effect for large roofs.
enough to produce significant reductions in roof
loads. 21. Roofs situated below adjacent roofs are par-
ticularly susceptible to heavy drift loads, because
17. Roof slope factor, Cs. Snow loads on a slop- the upper roofs can provide a large volume of snow
ing surface act on the horizontal projection of the to form drifts.(5)(17)(18)(19)(20) Canopies, balconies and
surface. Under most conditions, less snow accumu- porches are similarly susceptible. The drifts that ac-
lates on steep roofs than on flat and moderately cumulate on these roofs depend mainly on the
sloped roofs, because of sliding, creep, better difference in elevation and on the size of the upper
drainage and saltation.(8)(9)(10) The coefficient, Cs, as roof.(17) Where the lower level roof area is large,
defined in NBC Sentence 4.1.7.1.(4) accounts for wind blowing for a considerable time at an angle
these effects by reducing the snow load linearly towards the raised portion of the roof may form an
from full snow load at 30 slope to zero at 70. A elongated “spike” or quartering drift extending lee-
lesser value of Cs is permitted in Sentence 4.1.7.1.(5) ward of the change in elevation.(21) The provisions
for unobstructed smooth, slippery roofs such as in paragraph 35 cover the typical types of snow
glass or metal. In this case, the load may be reduced load that arise on lower level roofs. For unusual ge-
linearly from full load at 15 to zero at 60. In order ometries, especially where the roof areas involved
for the designer to use the full reductions as de- are large, model studies can be useful in identifying
scribed in either of these relationships, the snow unusual drift formations.
should be able to slide completely off the roof sur-
face under consideration. 22. Projections such as penthouses or parapet
walls on flat roofs may collect triangular snow drifts
18. Situations in which public safety may be that reach the tops of the projections, but the magni-
compromised by snow and ice falling from roofs tude of the loads is usually less than on roofs
should be avoided. If snow fences or barriers are re- situated below adjacent roofs.
quired to keep snow and ice on roofs, they should
be designed to transmit the substantial forces in- 23. Wind flow over gable, or arch roofs, is accel-
volved into the building structure.(8)(9) Heat-traced erated by being deflected upwards on the windward

Commentary H
sides. On the leeward sides, velocities drop and the authority having jurisdiction is convinced that
snow entrained in the wind and scoured from the drifting will not occur, drift effects need not be con-
other side is deposited. Heavy unbalanced loads of- sidered. However, the influence of creep and sliding
ten occur as a result of the transfer of snow from one snow causing unbalanced loads should be consid-
side to the other.(22)(23) This unbalance is especially ered on gable roofs of slope >15, arches with rise to
important for domes and also for buildings such as span, h/b greater than 0.1, and other roofs of signif-
arenas, which have long spans and in which a col- icant slope.
lapse might be catastrophic. (1) Lightweight curved
structures, such as cold-formed metal arch build- 26. In deep snow areas of Canada, in particular
ings, are particularly sensitive to unbalanced snow the mountains and valleys of British Columbia, the
loads, as the self-weight of the structure is relatively following should be kept in mind:(5)(24)
small. These structures can generally be analyzed as (a) Snow cornices can become very large and can-
arches. However, the flexibility of such arches sug- tilever beyond the edges of flat or sloping
gests that a second-order analysis is likely to be roofs a distance equal to the snow depth on
required to predict the structural behaviour.(18)(22) the roof. This can occur in sheltered or com-
The structures can also be analyzed as shells when pletely calm areas and on the leeward side of
special consideration is given to shear transfer and roofs where there is wind. Cornices have been
to the axial capacity of longitudinal stiffeners. Load known to overload walls and columns, result-
tests may be needed to assess the behaviour and ing in failures. In addition, cornices are a
load carrying capacity of the structural elements, es- hazard if they break off. They can destroy bal-
pecially when transverse corrugations are present. conies, stairs, porches, attachments of wires,
etc., to the building and they could be very
24. When the wind flows over peaked or dangerous to people below.
smooth domes, unbalanced snow load will also oc- (b) When deep snow is deposited on slippery slop-
cur. Data on snow load distributions on domes are ing roofs, it has been known to shear off vents,
not available and wind tunnel or water flume tests chimneys, aerials, wiring, stacks, skylights and
are therefore recommended to assist in the selection ventilators when it slides. It is a menace below
of appropriate design loads. In the absence of such when it falls. In addition, it may creep off the
tests, the following approximate distributions may roof, rotating slowly at the eaves, and may
be considered; in plan view: even break windows if it hits the side of the
(a) a uniformly distributed load (adjusted for building. Protrusions through the roof should
slope) over the whole dome, and be located at the ridge, or be especially pro-
(b) the Case II loading (Figures H-2(a) and H-2(b)), tected against shearing forces due to the snow.
applied over a 90 sector, tapering linearly to (c) Where a roof is L-shaped or has dormers, the
zero at 22.5 beyond the sector boundaries and snow on each slope will slide in the direction
with no snow on the remaining 225 sector.) of the ribs or corrugations and jam in the val-
leys. If one slope is longer or steeper, the snow
Local experience should also be considered. on this slope will predominate and may force
Snow accumulations due to sliding and drifting oc- the whole mass of snow to slide across the op-
cur regularly at the bases of domes where they meet posing corrugations on the other slope,
the ground. These should not be neglected. resulting in a tearing or flattening of these cor-
rugations. If the corrugations hold and the
25. In completely calm areas, snow covers roofs snow does not slide, the restraining load on
and ground in uniform layers. For these locations, the lower opposing slope may be very high.
the design load can be considered as a uniformly
distributed load equal to some suitable fraction of 27. Redistribution of load due to melting. Re-
the ground snow load if sliding is not a factor. Truly distribution of loads may occur as a result of snow
uniform loads, however, are rare and have been ob- or ice melting and flowing or sliding to other areas
served only in certain mountain valleys of British where it refreezes, or falling to a lower roof where it
Columbia and occasionally in other parts of the accumulates as slush or ice. On sloped roofs, melt-
country, on roofs that are well sheltered on all sides water from warm, perhaps poorly insulated, parts
by high trees. Generally, the winds which usually may refreeze on colder areas or on the eaves and
accompany or follow snowfalls transport new snow cause high ice loads and also ice damming, water
from exposed to protected areas. Hence, the proba- back-up under shingles and danger from falling ici-
bility that high uniform loads will occur on exposed cles. These can be alleviated by taking steps to
roofs is reduced and the probability that drifts will decrease heat loss from the warm surfaces.
form is increased. Drifting does not occur in certain
local areas on the B.C. coast where heavy snow- 28. Since drainage under the snow cover on flat
storms invariably consist of wet snow. In these or nearly flat roofs is not generally as good as on
specific locations, the drift requirements of Sentence those with slopes, meltwater, slush and ice may be
4.1.7.1.(7) may be overly conservative. Where the retained longer. Also snow accumulations near pro-

Commentary H
jections can melt as a result of heat loss through the to or less than 15, the load distribution is deter-
roof or solar radiation or exhausted warm air. The mined by Case I, but is also subject to the general
resulting meltwater may migrate to the lower areas requirements of NBC Article 4.1.7.2. for “full and
causing heavy loads. The centres of bays are partic- partial loading’’ which now apply to the Case I
ularly vulnerable if the drains are located at the loading only. On slopes over 15, Case II, which ac-
columns (high points). This redistribution of load counts for unbalanced loading, and Case I both
may cause further deflection and lead to an instabil- apply. Case II loading is intended to account for
ity similar to that produced by rain ponding (see snow blown from the windward over to the lee-
Commentary I). ward side as well as snow removed by sliding from
one side. Flat and shed (single sloped) roofs are
Detailed Explanations of Figures H-1 to H-6 subject to Case I and “full and partial’’ loading only.
29. In the following, Figures H-1, H-2(a) and H- 31. Arch roofs (Figures H-2(a) and H-2(b)). r2
2(b) apply to the basic shapes: the simple flat and Uniform and unbalanced load distributions are par-
shed roofs, the simple gable roofs and the simple ticularly important for the design of curved roofs.
arch and curved roofs. More complex shapes can of- (19)(22)(23) In addition, the requirements for “full and
ten be considered as combinations of these. When partial loading’’ apply. Case II loading may also be
the roofs shown in Figures H-1, H-2(a) and H-2(b) used for the design of domes (see Paragraph 24).
are adjacent to higher roofs or have projections or
are combined to form valleys, reference should also
be made to Figures H-3 to H-6. For all these simple
and complex roofs, the basic snow load coefficient,
Cb, is 0.8 in all loading cases.
30. Gable, flat and shed roofs (Figure H-1). On

gable roofs both uniformly distributed and unbal-
anced loads should be considered for all slopes less
than 70 (or 60 for unobstructed slippery roofs), as
shown in Figure H-1. Where both slopes are equal
Roof profile
α
Factors
Load Roof
case slope
Cw Cs Ca
Distribution of snow load, S
α
Case I 0.75 (3)

(4)
I 0° ≤ α ≤ 90° or f (α) (1) 1.0
1.0
II 15° ≤ α ≤ 20° 1.0 f (α) (1) 0.25 + α/20

Case II(2) 20° ≤ α ≤ 90° 1.0 f (α)(1) 1.25
Figure H-1
Snow distributions and snow loading factors for gable, flat and shed roofs
Notes to Figure H-1:

(1) Varies as a function of slope as defined in NBC Sentences 4.1.7.1.(4) and (5)

(2) Case II loading does not apply to gable roofs with slopes of 15 or less or to single-sloped (shed) roofs or to flat roofs.
(3) Cw = 0.75 may be reduced to 0.5 for exposed areas north of the treeline as defined in NBC Sentence 4.1.7.1.(3).
(4) For large flat roofs Ca, as defined in NBC Clause 4.1.7.1.(7)(b), may exceed 1.0.

Commentary H
Wind
Roof profile
x
(x, hx)
α = 30° α
h
αe
hx
hw ≥ 0
b
Range Factors
α > 30° α < 30° α > 30° Load
of
case Cw Cs Ca
Distribution of snow load, S application
f( α)
0.75 (3)
Case I I All or f(α) (5) 1.0
1.0
f( α)
Max.
h γ hx 3 kPa
S ≤ 3 + Sr II (1) (6) > 0.05 1.0 f(α) (5) but not more than A
Case II b S s Cb Ss Cb
f( α) For 0 < x < x 30 (4)

Max. 2 x
S = 2 Ss + Sr(2) Ca = A
Case III h Cb x 30
III (1) > 0.05 1.0 f(α) (5)
h b
For x 30 ≤ x
2
where A = b , A ≤ 1.0 Ca = A
0.12 Cb
Figure H-2(a) r2
Snow distributions and snow loading factors for simple arch or curved roofs
Notes to Figure H-2(a):

(1) Use Case II or Case III, whichever produces the lower total load per unit length of building perpendicular to the span. r2
(2) Max. S = 2 Ss+ Sr occurs at = 30o or at the edge of the roof if e < 30o.
(4) x30 = value of x where = 30o or value of x at edge of roof if e < 30o.
(5) Varies as a function of slope as defined in NBC Sentence 4.1.7.1.(4).
(6) The full and partial loading provisions of NBC Sentence 4.1.7.2.(2) also apply to all curved roofs. r2

Commentary H
Wind
Roof profile
x
(x, h x)
α = 15° α
h
αe
hx
hw ≥ 0
b
α > 15° α < 15° α > 30°
Range Factors
Distribution of snow load, S
Load
of
case Cw Cs Ca
application
f( α)
0.75 (3)
Case I I All or f(α) (5) 1.0
1.0
f( α)
Max. h γ hx 3 kPa
S ≤ 3 + Sr II (1) (6) > 0.05 1.0 f(α) (5) but not more than A
Case II b Ss C b Ss C b
f( α) For 0 < x < x 30 (4)

Max. 2 x
S ≤ 2 Ss + Sr(2) Ca = A
Case III h Cb x 30
III (1) > 0.05 1.0 f(α) (5)
b
h For x 30 ≤ x
2
where A = b , A ≤ 1.0 Ca = A
0.12 Cb
Figure H-2(b) r2
Snow distributions and snow loading factors for simple arch or curved roofs with unobstructed slippery surfaces
Notes to Figure H-2(b):

(1) Use Case II or Case III, whichever produces the lower total load per unit length of building perpendicular to the span. r2
(2) Max. S = 2 Ss + Sr occurs at α = 15o or at the edge of the roof if e < 15o. r2
(4) x30 = value of x where = 30o or value of x at edge of roof if e < 30o.
(6) The full and partial loading provisions of NBC Sentence 4.1.7.2.(2) also apply to all curved roofs. r2

Commentary H
x1 x2
α1 α2
α1 α2 Factors
Roof
Load
Distribution of snow load, S slope
case Cw Cs Ca
α
b1 b2
2 2 0.75 (2)
I α ≤ 90° or f (α) (1) 1.0
Case I 1.0
S = Ss + Sr For 0 < x ≤ b/4:

Ca = 1/C b
Case II II 10° < α ≤ 90° 1.0 1.0
For b/4 < x ≤ b/2:
S = 0.5 S s + S r
b1 b2 Ca = 0.5/C b
4 4
S = 1.5 S s + S r For 0 < x ≤ b/8:
Ca = 1.5/C b
III 10° < α ≤ 90° 1.0 1.0
Case III For b/8 < x ≤ b/2:
S = 0.5 S s + S r Ca = 0.5/C b
b1 b2
8 8
Figure H-3
Snow distributions and snow loading factors for valley areas of roofs

32. Snow accumulations caused by wind and by higher, it is an obstruction; if higher than a “non-
snow sliding off the surface regularly occur on ei- obstructing” parapet and wide enough to provide a
ther or both sides of arches where they meet the significant source of snow on its upper surface, it is
ground and should not be neglected. (22) an “upper level” roof.
33. Valleys in curved or sloped roofs (Figure 35. The lower roof (Figures H-4(a) and H-4(b)).
H-3). In the design of roofs with valleys, uniform The load distribution on roofs adjacent to higher
loads and loads accounting for drifting, sliding or ones is taken to be a triangular shape as illustrated
creep, and the movement of meltwater are impor- in Figure H-4(a). Thus Ca varies with distance x
tant. A reduction due to slope is allowed for Case I from the step in roof elevation, being Ca(0) at x = 0
loading, because as the snow creeps down the slope and decreasing linearly to a value Ca(xd) at the tail
and wrinkles and layers at the bottom of the valley, of the drift defined by x = xd. The magnitude of the
the loads on the upper slopes are reduced. Since drift on the lower roof depends primarily on the
Cases II and III describe the worst loads due to amount of snow that can drift off the upper roof
drifting and slope effects, the Cs factor is taken as and be trapped in the step. Therefore the loading in
equal to 1.0. the step tends to increase significantly as the size of
the upper level roof increases.(17) For roofs with rela-
34. Multi-level roofs, obstructions and para- tively small step heights and, in particular, for large
pets. Multi-level roofs, obstructions and parapets roofs, the step can fill to the top. In this case Ca(0)
are all “bluff objects” creating turbulent wakes should be taken as:
downwind, where snow accumulates in drifts. Such
roofs and obstructions can be considered as geomet- (1)
rical variations of a rectangular object situated on or
adjacent to a lower flat roof. If the object is narrow
and lower than the design depth of uniformly dis- where h is the difference in elevation between the
tributed snow on the roof, it is, for the purposes of lower roof surface and the top of the parapet on the
this Commentary, a “non-obstructing” parapet; if upper roof, and is the specific weight of snow.

Commentary H
(Note that in the region of the drift, Cw is 1.0, since

the lower roof is sheltered by the roof step.) How- (6)
ever, in many cases there may be insufficient snow
to fill in the step. Therefore an upper limit on Ca(0)
equivalent to for the case where the drift is limited by other fac-
tors accounted for by Equation (3). The value of Ca
in Equations (3), (5) and (6) is that applicable to the

(2) lower roof, i.e., Ca would normally be taken as 1.0
except where the plan dimensions of the lower roof
are so large as to produce a value of l* for the lower
may be used, where the factor F is to be taken as
roof exceeding 75 m (in which case a higher value
the greater of
of Ca would apply as provided for in NBC Clause
4.1.7.1.(7)(b)). Drifts deposited as a result of a
change in elevation occur not only when the upper
roof is part of the same building but also when it is
(3) on an adjacent building not more than 5 m away, as
shown in Figure H-4(a). Where the upper roof is
very large, the limiting gap, a, of 5 m should be
l* is a characteristic length of a roof defined by confirmed by model tests. Where the drift obtained
from Figure H-4(a) is longer than the lower roof, the
(4) drift should be truncated at the edge of the lower
roof.
where w and l are the shorter and longer dimen- 36. Canopies or small roofs adjacent to tall
sions of the roof plan form. buildings. Where the lower roof is small relative to
the difference in elevation between the two roofs,
In Equation (3), l* for the upper roof is used, Ca such as for an entrance canopy at the base of a
is the accumulation factor applicable to the lower high-rise building, the loads will be less than those
roof, and hp is the height of the upper roof’s parapet described in Paragraph 35 because the snow from
as illustrated in Figure H-4(a). Figure H-4(b) shows the upper roof is dispersed over a wide area and
graphically the variation of the factor F as a func- drifting over the small lower roof is insufficient to
tion of l*/Ss. Background information on the build up a large accumulation. While research has
derivation of Equation (3) is provided in Reference not been sufficient to fully evaluate the reduced
(17). For buildings in sheltered locations (Exposure loadings, the following approach is suggested. For
B or C as defined in Commentary B, Wind Loads, small area lower roofs with a plan area less than
Paragraphs 13 and 14), an upper limit of 5.0 can be 25 m2, situated more than 20 m below the upper
placed on the value of F as shown in Figure H-4(b). level roof, Ca may be taken as 1.0. Where the height
The horizontal length xd of the drift extending out difference, h, is less than 10 m, Ca should be as de-
from the step should be based on the top surface of scribed in Paragraph 35. For cases between h = 10 m
the drift having a 1:5 slope. This implies that and h = 20 m, the form of drift described in Para-
graph 35 should be used but with Ca(0) reduced in
linear fashion as h varies from 10 to 20 m, i.e.:
(5)
for the case where the step fills to the top, or (7)

Commentary H
x
hp
Value of xd :
h
α Lesser of
xd = 5 (h - Cb Ca Ss/γ)
and
xd = 5(Ss/γ) (F - Cb Ca)
h
α Value of h' :
Cb Cw Ca Ss
h' = h –
γ
a
Distribution of snow load, S Value of Ca(0) :
γh
Lesser of Ca(0) =
C b Ss
and
xd Ca(0) = F/Cb
x = 10 h'
Factors
x Cw Cs C a(2)
0 1.0 f (α) (1) Ca(0)
(Ca(0) – Ca (xd)) x
0 < x ≤ xd 1.0 f (α)(1) Ca(0) -
xd
xd < x ≤ 10 h' 1.0 f (α)(1) 1.0(4)
0.75(3)
> 10 h' or f (α)(1) 1.0(4)
1.0
Figure H-4(a)
Snow distributions and snow loading factors for lower levels of adjacent roofs
Notes to Figure H-4(a):

(1) Varies as a function of slope as defined in NBC Sentences 4.1.7.1.(4) and (5).
(2) If a > 5 m or h ≤ 0.8Ss/ drifting need not be considered. Where the upper roof is very large, the limiting gap, a, of 5 m should be confirmed by
model tests.
(4) For large flat roofs Ca, as defined in NBC Clause 4.1.7.1.(7)(b) may exceed 1.0.

Commentary H
10 pending on the relative sizes, slopes and positions

9 of the two roofs. If all the sliding snow cannot be
8 retained on the lower roof because it is too small,
exposed appropriate reductions may be made. A profile of
7
the snow depth on the roof should be drawn to con-
6 sheltered
firm that the loading is reasonable.
F 5 equation 3
4 upper limit
39. Areas adjacent to obstructions (Figure H-6).
lower limit Consideration should also be given to triangular
3
drift loads adjacent to significant vertical obstruc-
2 γhp
0 1 2 3 tions, such as elevator, air conditioning and fan
Ss
1 housings, small penthouses and wide chimneys.
0 The peak load adjacent to the obstruction in Figure
10 100 1000 H-6 is assumed equal to 0.67 h + Sr, where h is the
γl*/Ss obstruction height in metres and the unit weight
Figure H-4(b) in kN/m3. It decreases to the design roof load at a
distance of 2 h from the obstruction. The peak load
Variation of the factor, F, for lower roof Ca = 1.0 need not be larger than 2 Ss + Sr (Ca ≤ 2/Cb ) nor is
it necessary to consider the drift load if the width, b,
of the obstruction in Figure H-6 is less than 3.0 Ss/ .
37. If the building is at a sheltered site, the
wind exposure factor Cw is 1.0 for all areas of the
lower roof. For exposed sites Cw on the lower roof
should still be taken as 1.0 within a region sheltered Roof profile
by the step extending outwards from the step a dis- upper roof
tance equal to 10 h’, where h’ is the difference in
elevation between the top of the upper roof’s para-
h
pet and the snow surface on the exposed portion of α
the lower roof. This definition implies that
(8) Distribution of snow load (1, 2, 3), S
in which the applicable value of Cw is either 0.75 or

0.5, as provided for in NBC Sentence 4.1.7.1.(3), and sliding snow load
Ca is the appropriate value for the exposed portion
of the roof, i.e., normally 1.0, but may be higher for
roofs of large dimensions as specified by NBC drift load
Clause 4.1.7.1.(7)(b).
38. Multi-level roofs with the upper roof

sloped (Figure H-5). A lower roof should be de- Figure H-5
signed for the loads provided in Figure H-4(a) plus Snow distributions on lower roof with sloping upper roof
an additional load produced by the snow that may
slide from an upper roof. The following guide is
recommended. Because of the low probability that Notes to Figure H-5:
(1) Lower roof is designed for drift load (see Figures H-3 and H-4(a))
both upper and lower roofs will have the full load
and sliding snow load (see Paragraph 36).(3)
over their entire areas simultaneously when sliding
(2) Upper roof is designed in accordance with NBC Subsection 4.1.7.
occurs, the lower roof should be assumed to carry
(see Figures H-1, H-2(a) and H-2(b)).(3)
its full load according to Figure H-4(a) plus 50% of
(3) In the calculation of Ca and Ca(0) in Notes (1) and (2), “l*” and “w”
the total weight of the Case 1 snow load from the
are taken as the plan dimensions of the upper or lower roof.
portion of the upper roof that slopes toward the
lower roof. The distribution should be made de-

Commentary H
Roof projection Value of xd :
xd = 2h but 3 m ≤ xd ≤ 9 m
h Value of h' :
b Cb Cw C Ss
a
h' = h –
γ
Distribution of snow load, S Value of Ca(0):

x
γh
Ca(0) = 0.67
Cb Ss
2 2
xd when Ca(0) < 1.0 use 1.0, when Ca(0) > use
Cb Cb
x = 10 h'
Factors
b,
x
metres Cw Cs C a(4)
0.75(3)
3 Ss (2)
≤ ALL or f (α)(1) 1.0
γ 1.0
3 Ss
> 0 1.0 f (α)(1) Ca(0)
γ
3 Ss (Ca(0) – 1) x
> 0 < x ≤ xd 1.0 f (α)(1) Ca(0) –
γ xd
3 Ss
> xd < x ≤ 10 h' 1.0 f (α)(1) 1.0
γ
0.75(3)
> 10 h' or f (α)(1) 1.0
1.0
Figure H-6
Snow distribution and snow loading factors for areas adjacent to roof obstructions

(1) Varies as a function of slope as defined in NBC Sentences 4.1.7.1.(4) and (5).
(2) If b is less than 3 Ss/ , m the effect of the obstruction on the snow loading can be ignored.
(3) Cw = 0.75 may be reduced to 0.5 for exposed areas north of the treeline, as defined in NBC Sentence 4.1.7.1.(3).
(4) For large flat roofs Ca, as defined in NBC Clause 4.1.7.1.(7)(b), may exceed 1.0.

Commentary H
Unusual Roofs is immobilized and snow removal crews can-

not be obtained,
40. In some cases, particularly roofs of unusual (b) snow cannot be effectively removed from the
shapes, exceptionally large roofs and roofs over centre of large roofs, and
which the air flow is significantly affected by other (c) unbalanced loading can occur as a result of
buildings or topographic features, the prediction of certain patterns of snow removal.
snow loads is difficult. In such cases, the designer

should calculate and plot the snow depths to scale 45. In special cases, roofs have been designed
applying a unit weight of 3.0 kN/m3 to judge with reduced design loads because of the incorpora-
whether the distributions look reasonable. In some tion of melting systems which periodically clear
circumstances wind tunnel or water flume tests them of snow. Decisions to use such systems should
might be used to assist in the evaluation. be considered carefully, because adequate energy for
melting may not be available when required. Fur-
Parking Decks
ther, as the years pass, the importance of keeping
41. Roofs used as parking decks should be de- the system functioning (perhaps at great cost) may
signed for the loads noted in NBC Tables 4.1.6.3. be forgotten.
and 4.1.6.10. or the roof snow loads, whichever are
greater. Where snow removal may occur, considera- Ice Loading on Structures
tion should be given to the loads due to snow
removal equipment and to the weight of piled snow. 46. Loads due to ice accretion on exposed sur-
faces of superstructure members, railings, lattice
Full and Partial Loading towers and signs are described in References (26)
and (27). The Atmospheric Environment Service,
42. All roof areas, including those to be de- Downsview, Ontario has a model, based on climate
signed for increased or decreased loads according to data at weather stations, to compute ice loading on
Figures H-1 to H-6, must be designed for the full vertical and horizontal surfaces and cables.
specified load given in NBC Article 4.1.7.1. over the
entire area. However, only the flat, shed, low slope Minimum Roof Load
gable (<15) and curved roofs (h/b < 0.1) of Figures
H-1 and H-2 need be designed for Case I loading 47. Articles 4.1.6.3. and 4.1.6.10. of the National
distributed on one portion of the area and half of Building Code provide for a minimum uniform roof
this on the remainder of the area, the location and load of 1 kPa and a minimum concentrated load of
size of such partial areas being chosen to give the 1.3 kN. These loads are “use and occupancy loads”
greatest effects in the members and joints con- intended to provide for maintenance loadings,
cerned. These requirements do not imply workmen and so forth. These loads are not reduced
checkerboard loading because the probability that as a function of area or as a function of the roof
checkerboard loading will occur in such a way as to slope.
give the worst conditions for supporting members is
generally too remote to be considered in design.(25) History of Snow Loads in the
On most of the roofs in Figures H-1 and H-2 a num- National Building Code
ber of separate cases of full and partial loading will
be required to ensure proper design of all elements. 48. In the 1953 National Building Code of
Canada, design snow loads were equal to the
43. The reason for these requirements is that
ground snow load, with reductions allowed for
snow seldom accumulates according to the simple
configurations in Figures H-1 and H-2. Conse- sloped roofs only. Such values were very approxi-
quently, full and partial loading must be considered mate and resulted in overdesign in some roofs
while allowing underdesign in others, particularly
for the design of structural members which are sen-
sitive to changes in load distribution (e.g., truss in areas subject to high drift loads. Information on
diagonals and cantilevers) and which would not which to base a more refined assessment of the
loads was not available, however, until a country-
otherwise be designed for unbalanced loads.
wide survey of snow loads on roofs was undertaken
Snow Removal by the Institute for Research in Construction with
the help of many volunteer observers. This survey
44. Although it is fairly common practice in provided evidence on the relationship between
some areas to remove snow from roofs after heavy ground and roof loads and enabled the committees
snowfalls, the National Building Code does not al- responsible for the 1960 edition of the National
low a reduction of the design load to account for Building Code of Canada to make changes. The roof
this because: load was set at 80% of the ground load and the
(a) snow removal cannot be relied upon. Experi- ground load was based on a return period of 30
ence in several countries has shown that years and adjusted to allow for the increase in the
during and after extreme snow storms, traffic load caused by rainwater absorbed by the snow.

Commentary H
49. With the introduction of the 1965 Code and on large flat upper or lower roofs. Additional infor-
its Commentary, further changes made by the Revi- mation is also provided in this Commentary for
sion Committee on Structural Loads and Procedures snow loads on lower roofs.
led to a more rational approach to design loads. All
roof loads were directly related to the snow load on References
the ground and, consequently, the roof snow loads
were removed from the table of Design Data for (1) D.A. Taylor, A Survey of Snow Loads on the
Selected Locations in Canada. The basic load re- Roofs of Arena-Type Buildings in Canada.
mained at 80% of the ground load, except that a Can. J. Civ. Eng., Vol. 6, No. 1, 1979, pp. 85-96.
snow load of 60% of the ground load was allowed (2) M.J. Newark, L.E. Welsh, R.J. Morris and W.V.
for roofs exposed to the wind. This reduction was Dynes, Revised Ground Snow Loads for the
made because at the same time allowance was 1990 National Building Code of Canada. Can.
made for a variety of influences causing accumula- J. Civ. Eng., Vol. 16, No. 3, June 1989, pp. 267-
tions of snow on roofs. This was done by means of 278.
“snow load coefficients” or shape factors, which (3) Canadian Permanent Committee on Geo-
were shown in the form of simple formulas and dia- graphic Narmes, Gazetteer of Canada (12
grams similar to Figures H-1 to H-6. In addition, the Volumes by Province and Territory). Natural
slope reduction formula was changed from the step Resources Canada, Available from Mail Order
function used in 1960 to a linear function. Services, Canadian Government Publishing
Centre, Dept. of Supply and Services, Ottawa,
50. In the 1970 Code and Commentary, minor Ontario K1A 0S9.
changes were made to the provisions for gable and (4) B.R. Claus, S.O. Russell and P.A. Schaerer,
arch roofs and more severe “full and partial loading Variation of Ground Snow Loads with Eleva-
provisions”: “full and zero loading” rather than tion in Southern British Columbia. Can. J. Civ.
“full and half.” Eng., Vol. 11, No. 3, September 1984, pp. 480-
493.
51. In the 1975 Code and Commentary, few
(5) D.A. Taylor, Snow on Two-level Flat Roofs –
changes were made, except that the requirement for Measured vs 1990 NBC Loads, Can. J. Civ.
full and partial loading was considered too severe Eng., Vol. 19, No. 1, 1992, pp. 59-67.
at “full and zero” and was changed back to “full
(6) Ontario Ministry of Municipal Affairs, Hous-
and half” loading. ing, Research and Development Section.
52. In the 1977 and 1980 Commentaries, the pro- Prevention of Excess Snow Accumulation due
visions for loads on arch roofs were changed and a to Roof Mounted Solar Collectors. (Report
number of rationalizations made as an aid to better prepared under contract by MHTR Ltd.,
understanding of snow loads on roofs. Guelph, Ontario) Toronto, December 1981,
76 pp.
53. The 1985 Code and Commentary provisions (7) D. Nixon, Solar Collectors – Briefing Docu-
were rewritten to simplify the presentation and ment S-2. Public Works Canada, Design/
to clarify the intent of the minimum roof loading of Construction Branch, Sir Charles Tupper
1.0 kPa. Further, the minimum roof loading Building, Ottawa, March 1981, 25 pp.
was made independent of slope, the unit weight (8) D.A.Taylor, Snow Loads on Sloping Roofs.
of roof snow was increased 1.9% to give = Two Pilot Studies in the Ottawa Area. Can. J.
2.4 kN/m3, “full and partial loading’’ was restricted Civ. Eng., Vol. 12, No. 2, 1985, pp. 334-343.
to Case I loadings on buildings in Figures H-1 and (9) D.A. Taylor, Sliding Snow on Sloping Roofs.
H-2, and the unbalanced loading on arches was CBD 228, Division of Building Research,
simplified. National Research Council Canada, Ottawa,
1983, 4 pp.
54. In the 1990 Code and Commentary a new (10) R.L. Sack, Snow Loads on Sloped Roofs.
slope reduction formula was given for unobstructed ASCE Journal of Structural Eng., Vol. 114,
slippery sloping roofs, the unit weight of roof snow No. 3, March 1988, pp. 501-517.
was increased to = 3.0 kN/m3, the need for unbal- (11) N. Isyumov, Roof Snow Loads – Their
anced snow loads on domes was emphasized, the Variability and Dependence on Climatic Con-
minimum Cw was reduced to 0.5, rather than 0.75, ditions. Symposium on the Structural Use of
for exposed roofs north of the tree line, and design Wood in Adverse Environments, 15-18 May
roof snow loads were separated into snow and rain 1978, Vancouver, Van Nostrand Reinhold, 510
components consistent with the new ground snow pp.
loads given in Chapter 1 of the 1990 Supplement. (12) N. Isyumov and M. Mikitiuk, Climatology of
Snowfall and Related Meteorological Vari-
55. In the 1995 Code and Commentary, new for- ables with Application to Roof Snow Load
mulae are given for the accumulation factor, Ca, for Specifications. Can. J. Civ. Eng., Vol. 4, No. 2,
the calculation of uniformly distributed snow loads 1977, pp. 240-256.

Commentary H
(13) N. Isyumov and A.G. Davenport, A Proba-

bilistic Approach to the Prediction of Snow
Loads. Can. J. Civ. Eng., Vol. 1, No. 1, 1974,
pp. 28-49.
(14) W.R. Schriever, Y. Faucher, and D.A. Lutes,
Snow Accumulation in Canada: Case Histo-
ries: I. Division of Building Research,

National Research Council Canada, Ottawa,
January 1967. NRCC 9287.
(15) D.A. Lutes, and W.R. Schriever, Snow Accu-
mulation in Canada: Case Histories: II. DBR
Technical Paper 339, Division of Building
Research, National Research Council Canada,
Ottawa, March 1971. NRCC 11915.
(16) J.T. Templin and W.R. Schriever, Loads due to
Drifted Snow. Journal of Structural Division,
Proc., Am. Soc. Civ. Eng., Vol. 108, No. ST8,
August 1982, pp. 1916-1925.
(17) P.A. Irwin, S. L. Gamble, and D. A. Taylor, Ef-
fects of Roof Size, Heat Transfer and Climate
on Snow Loads: Studies for the 1995 NBC.
Can. J. Civ. Eng., Vol. 22, No. 4, 1995.
(18) M.J. O’Rourke and E. Wood, Improved Rela-
tionship for Drift Loads on Buildings. Can. J.
Civ. Eng., Vol. 13, No. 6, 1986, pp. 647-652.
(19) D.A. Taylor, Roof Snow Loads in Canada.
Can. J. Civ. Eng., Vol. 7, No. 1, 1980, pp. 1-18.
(20) N. Isyumov and M. Mikitiuk, Wind Tunnel
Model Tests of Snow Drifting on a Two-level
Flat Roof, J. of Wind Engineering and Indus-
trial Aerodynamics. Elsevier Science
Publishers, N.Y. Vol. 36, 1990.
(21) D.J.L. Kennedy, N. Isyumov and M. Mikitiuk,
The Effectiveness of Code Provisions for
Snow Accumulations on Stepped Roofs. 2nd
International Conf. on Snow Eng’g. Engineer-
ing Federation, Santa Barbara CA, June 1992.
(22) D.A. Taylor, Snow Loads for the Design of
Cylindrical Curved Roofs in Canada 1953-
1980. Can. J. Civ. Eng., Vol. 8, No. 1, 1981,
pp. 63-76.
(23) T.H.R. Kennedy, D.J.L. Kennedy, J.G. MacGre-
gor and D.A. Taylor, Snow Loads in the 1985
National Building Code of Canada: Curved
Roofs. Can. J. Civ. Eng., Vol. 12, No. 3, 1985,
pp. 427-438.
(24) D.A. Taylor, Snow Loads on Roofs in British
Columbia – Results of a Survey, Proc. 1991
CSCE Annual Conference. Vol. 111, 1991, pp.
205 - 213.
(25) R.L. Booth, and D.A. Taylor, Discussion, De-
sign of Light Industrial Buildings. Can. J. Civ.
Eng., Vol. 7, No. 4, 1980, pp. 660-661.
(26) Ontario Ministry of Transportation and Com-
munications, Highway Engineering Division,
1983 Ontario Highway Bridge Design Code,
Toronto, p. 46.
(27) Canadian Standards Association CAN/CSA-
S237-M86. Antennas, Towers and
Antenna-Supporting Structures. Rexdale,
Ontario, 1986, 73 pp.

72
Commentary I
Rain Loads
1. In accordance with National Building Code where

Sentence 4.1.7.3.(1), any roof which can accumulate
water must be designed for the load that results E = modulus of elasticity,
from a 24 h rainfall on the horizontal projected area
of the roof. This requirement applies whether or not I = moment of inertia of the beam or deck-
the surface is provided with drainage, such as rain- ing,
water leaders. The distribution of rain load should
be determined by the designer, who should take L = span,
into account the shape of the roof, including cam-
ber, with or without creep deflection due to dead S = spacing of the beam or decking,
load, and also deflection due to rain.
= mass density of water, kg/m3.
2. Notwithstanding the above requirement, it is
considered good practice in the location of roof 5. In the case of a two-way system of roof joists
drains to take into account not only the roof slope on girders, the critical stiffness can be approximated
but also deflection of the roof due to creep, snow by
and rain. Drains should be provided with suitable
devices to prevent clogging by leaves or, when ap-
propriate, suitable overflows should be provided
through parapet walls. (2)
3. In some areas of Canada the primary

drainage system for a roof could block due to where EIjcrit and EIgcrit are given by Equation (1) for
freeze-thaw conditions. Roofs in these areas should joists and girders, respectively.
be designed accordingly.
6. Even if the roof system is stiffer than the crit-
Ponding Instability ical values determined by Equations (1) and (2),
calculated moments and deflections may be ampli-
4. If a flat roof is too flexible, rainwater will not fied due to ponding effect. A practical criterion is to
accumulate evenly over the roof but will flow to require roof stiffness to be at least twice the critical
form ponds in a few local areas. This may lead to stiffness. In the case of a one-way system on rigid
an instability similar to buckling, which can result supports, in terms of existing deflection require-
in failure of the roof due to local overloading. In the ments, this can be expressed as follows:
case of one-way roof beams or decking simply sup-
ported on rigid supports, ponding instability will
occur when the beam or decking stiffness is less
than EIcrit given by (3)
where w is the design load in kilopascals specified

(1)
1
for deflection calculation and ( /L)allowable is the al-
lowable deflection to span ratio (see Table A-1,
Commentary A, “Serviceability Criteria for Deflec-
tions and Vibrations”). If for a one-way system the
design load w is less than the critical value given in
Table I-1, the effects of ponding should be consid-
ered. This applies particularly to large flat roofs in
areas of heavy rainfall. Further information is given
in References (1) to (7).

Commentary I
Table I-1
Critical Values of w for Ponding, kPa
(one-way system – Equation (3))
w, kPa
Deflection/Span
L=5 m L=10 m L=20 m L=30 m
Requirement
1:180 0.43 0.86 1.71 2.57
1:240 0.32 0.64 1.28 1.93
References
(1) D.A. Sawyer, Ponding of Rainwater on Flexi-
ble Roof Systems. Journal of Structural
Division, Proc., Am. Soc. Civ. Eng., , Vol. 93,
ST1, February 1967, p. 127.
(2) R.W. Haussler, Roof Deflection Caused by
Rainwater Pools. Civil Engineering, Vol. 32,
October 1962, p. 58.
(3) F.J. Marino, Ponding of Two-Way Roof Sys-
tems. Engineering Journal Am. Inst. of Steel
Construction, Vol. 3, No. 3, July 1966, p. 93.
(4) Commentary on the Specification for the De-
sign, Fabrication and Erection of Structural
Steel for Buildings. Am. Inst. of Steel
Construction, New York, February 1969.
(5) A.E. Salama and M.L. Moody, Analysis of
Beams and Plates for Ponding Loads. Journal
of Structural Division, Proc., Am. Soc. Civ.
Eng., Vol. 93, ST1, February 1967, p. 109.
(6) J. Chinn, A.H. Mansouri and S.F. Adams,
Ponding of Liquids on Flat Roofs. Journal of
Structural Division, Proc., Am. Soc. Civ. Eng.,
Vol. 95, ST5, May 1969, p. 797.
(7) D.A. Sawyer, Roof-Structural Roof-Drainage
Interactions. Journal of Structural Division,
Proc., Am. Soc. Civ. Eng., Vol. 94, ST1, Jan-
uary 1969, p. 175.

Commentary J
Effects of Earthquakes
Objectives of Earthquake-Resistant 4. It is beyond the scope of the NBC to cover

Design the entire range of problems involved in the
earthquake-resistant design of all structures. Un-
1. The earthquake-resistant design requirements usual structures, highly irregular buildings and
of the National Building Code of Canada 1995 pro- special-purpose industrial structures such as nuclear
vide an acceptable level of public safety, which is reactors, power plants and stacks should be treated
achieved by designing to prevent major failure and as special problems with special design criteria in
loss of life. Structures designed in conformance with each instance, including possibly a dynamic analy-
these provisions should be able to resist moderate sis. Likewise, tanks, piping and structures whose
earthquakes without significant damage and major failure constitutes an unusually high involuntary
earthquakes without collapse. For the purpose of hazard to health or life as a result of emission of
this section, collapse is defined as the state at which toxic or explosive substances must be designed to
exit of the occupants from the building becomes im- standards determined by an evaluation of the prob-
possible because of failure of the primary structure. abilities of failure and their consequences. Some of
these structures have earthquake-resistant design
2. Structures could be designed to resist major provisions identified in other standards, e.g.,
earthquakes without damage; this, however, would CANDU nuclear power plants,(3) liquefied natural
be uneconomical and unwarranted because of the gas facilities(4) and fixed offshore production struc-
relatively small likelihood of such events in Canada. tures.(5) The advice of an experienced engineer
Instead, the objective of the NBC provisions is to re- should be sought to arrive at suitable design criteria.
duce the probability of fatalities to an appropriately
small value and to accept some structural damage Seismic Regionalization
in major earthquakes. Damage caused by landslides
such as have occured in Anchorage, Alaska,(1) or 5. Earthquake-resistant designs should be con-
damage due to earth consolidation or liquefaction sidered for structures built in all regions of Canada,
as in Niigata, Japan,(2) will not be prevented by con- although the severity of expected earthquake effects
forming to the seismic requirements of the NBC. The varies considerably across the country. Detailed in-
NBC regulations provide buildings with resistance formation on earthquakes that have occurred in
to earthquake ground motions but not to slides, sub- Canada is contained in the publications of the Geo-
sidence, or active faulting in the immediate vicinity logical Survey of Canada, Natural Resources
of the structure; such cases require special study. Canada.(6-10) From these studies, the present seismic
zoning maps for Canada have been developed.(11-15)
3. To design an earthquake-resistant structure,
one needs to know the characteristics and probabil- 6. The seismic zoning maps (Figures J-1 and J-2)
ity of occurrence of the “design” seismic ground are based on a statistical analysis of the earthquakes
motion, the characteristics of the structure and the that have been experienced in Canada and adjacent
foundation, the allowable stresses in the materials of regions. These maps differ from the previous (1970)
construction, including the foundation soils, and the seismic zoning map employed in the 1980 and ear-
amount of damage that is tolerable. The design lier editions of the NBC in that
must provide not only sufficient structural strength (a) the data were analyzed using a method
to resist the ground motion, but also the proper proposed by Cornell(16) and a seismic risk com-
stiffness to limit the lateral deflection or drift. Dam- puter program developed by McGuire,(17)
age to nonstructural elements may be minimized by instead of the extreme-value method. The
proper limitation of distortions and by attention to present method provides for inclusion of geo-
the details of their connection to the primary struc- logical and tectonic information in support of
ture. The minimum requirements given in the NBC the seismic data,
incorporate the above considerations. This commen- (b) new strong seismic ground motion attenuation
tary elaborates on the quantitative and qualitative relations(12) were employed,
bases for the NBC requirements, and in some cases, (c) both peak horizontal acceleration and peak
recommends procedures. horizontal velocity have been mapped, and

Commentary J
(d) the probability of exceedance of the seismic seismic risk and the adoption of the second ground
ground motion parameters has been changed motion parameter, peak horizontal velocity.(15)
to 10% in 50 years (which is mathematically
equivalent to a probability of exceedance of 7. The range of the ratio, a, of peak horizontal
0.0021 per annum) from the previous value of ground acceleration to the acceleration due to grav-
0.01 per annum. ity (taken nominally as 10 m/s2) and of the ratio, v,
of peak horizontal ground velocity to a velocity of

The probability of exceedance of 10% in 50 years 1 m/s and the zonal ratios associated with each
for the ground motion parameters is considered to zone are given in Table J-1. Table J-2 includes repre-
be more appropriate than the previously employed sentative values of acceleration and velocity ratios
0.01 per annum for the objectives of earthquake- for several levels of probability of exceedance for a
resistant design described in Paragraph 1. However, number of Canadian cities. Similar calculations for
on average across the country, the calculation of any site in Canada may be obtained at cost by writ-
minimum lateral seismic forces has not changed rel- ing to Geophysics Division, Geological Survey of
ative to the NBC 1980, although at any geographical Canada, 1 Observatory Crescent, Ottawa, Ontario
location the forces will vary in some detail from pre- K1A 0Y3, or the Pacific Geoscience Centre, P.O. Box
vious values because of the improved estimates of 6000, Sidney, B.C. V8L 4B2.
Table J-1
Definition of Seismic Zones
Range of Peak Horizontal Ground Acceleration, g, or Peak

Acceleration-Related or Horizontal Ground Velocity, m/s, for 10% probability of
Velocity-Related Seismic Zonal Acceleration Ratio, a
exceedance in 50 years Zonal Velocity Ratio, v
Zone, Za, Zv
Equal to Less than
0 0.00 0.04 0.00
1 0.04 0.08 0.05
2 0.08 0.11 0.10
3 0.11 0.16 0.15
4 0.16 0.23 0.20
5 0.23 0.32 0.30
6 0.32 or greater 0.40
8. The zonal acceleration and velocity ratios for arrive at suitable design seismic ground mo-
Zone 6 (a = v = 0.40) should be considered nominal tion.
values. These ratios may be applicable for deriva-
tion of design loads for structures, but it should be 9. There is, and always will be, an inherent un-
recognized that certainty in the appropriate seismic ground motions
(a) site specific values of the acceleration and ve- to be used in earthquake-resistant design. Because
locity ratios in Zone 6 can exceed 0.40 by of the nature of earthquakes, their size and location
substantial margins, and and the ground motion effects that they will pro-
(b) depending on the type of structure, it may be duce cannot be accurately defined. The ground
necessary to characterize more explicitly the motions computed to produce Figures J-1 and J-2
nature of expected ground motion in the dom- are the current best estimates for building code pur-
inant frequency range of structural response. poses. Estimates of seismic ground motions at lower
In the latter case, the advice of an experienced levels of probability of exceedance than those given
engineering seismologist should be sought to in Table J-2 would require additional investigations.

Commentary J
Table J-2
Peak Horizontal Ground Acceleration (PHA, g) and Peak Horizontal Ground Velocity (PHV, m/s)
for Selected Localities and Probabilities of Exceedance
Probability of Annual Exceedance

Locality(1)
0.01 0.005 0.0021(2)

PHA PHV PHA PHV PHA PHV

Inuvik (68.30 N, 133.48 W) 0.033 0.052 0.043 0.066 0.060 0.083

Prince Rupert (54.30 N, 130.43 W) 0.074 0.13 0.096 0.20 0.13 0.27

Victoria (48.65 N, 123.43 W) 0.12 0.088 0.18 0.15 0.34 0.29

Vancouver (49.18 N, 123.17 W) 0.089 0.077 0.13 0.12 0.23 0.21

Calgary (51.10 N, 114.02 W) 0.011 0.026 0.014 0.032 0.019 0.040

Toronto (43.67 N, 79.63 W) 0.029 0.014 0.039 0.023 0.056 0.038

Ottawa (45.32 N, 75.67 W) 0.084 0.031 0.12 0.054 0.20 0.098

Montreal (45.47 N, 73.75 W) 0.078 0.031 0.11 0.053 0.18 0.097

Quebec City (46.80 N, 71.38 W) 0.075 0.035 0.11 0.066 0.19 0.14

Fredericton (45.87 N, 66.53 W) 0.046 0.020 0.066 0.036 0.096 0.066

Halifax (44.88 N, 63.52 W) 0.027 0.016 0.038 0.030 0.056 0.056

St. John’s (47.61 N, 52.75 W) 0.022 0.013 0.033 0.026 0.054 0.052
Notes to Table J-2:

(1) Geographical coordinates used for the computation are indicated.
(2) Equivalent to the probability of 10% in 50 years employed for Figures J-1 and J-2. P(50 yrs) = 1 - (1 - p(per annum))50.
0.08
0.16 0.16 g Za
0.08 0.32 0
0
0.04
0.3 1
2 0.08
0.16 2
0.32 0.11
0.32 0.1
.04
6 3
0.08 0.16
4
0.23
5
0.32
0.08
0.16
0.04 6
0.32
0.04
0. 04
08
0.
0.16 0.16
0.32
0.16
0.08
8
0.0 0.08
0.32
0.04
04 0.
16
0.16
0.
km
0 200 400 600
Figure J-1
Contours of peak horizontal ground accelerations, in units of g, having a probability of exceedance of 10% in 50 years.

Commentary J
04
0.
0.08
m/s v Zv
08
0
0.
0 0
0.04
0.05 1
0.16 0.08
16
0.10 2
0.
0.08 0.11
0.15 3
0.32
0.16
0.04 0.20 4
0.16
0.16
0.23
0.30 5
0.08
0.32
0.32
0.40 6
4
0.0
0.32
0.16
0.04 0.0
8 0.08 0.32
0.16
4
0.0 km
0 200 400 600
Figure J-2
Contours of peak horizontal ground velocities, in m/s, having a probability of exceedance of 10% in 50 years.
Direction of Earthquake Motions component. The ratio of vertical to horizontal mo-

[NBC 4.1.9.1.(3)] tion varies widely depending on site conditions; an
average value for this ratio is 2/3 to 3/4.
10. In general, ground motion in an earthquake
is multi-directional. This complex motion is im- 12. Under abnormally high vertical accelera-
parted to the supports of a structure; the structure tions, columns at the upper floors, especially at the
then responds according to its stiffness and inertial roof level, could be adversely affected. However,
properties. At any instant during the earthquake, there is usually sufficient reserve strength in vertical
the state of stress in the structure is a function of load-carrying members that vertical accelerations
the inertial forces and the deformations of the struc- can be safely neglected. In certain special
ture. In the most general case, seismic analysis structures,(18) these accelerations may have led to in-
would involve the simultaneous translation along stability or unusual reductions in the factors of
the two horizontal axes, rocking plus vertical and safety. Cantilevered structures or cantilevered
torsional motions. For normal buildings, however, building components are also sensitive to vertical
independent design about each of the horizontal accelerations. When this becomes a governing de-
axes together with the associated torsional forces is sign consideration, dynamic analysis should be
considered to provide adequate resistance against employed.
earthquake motions applied in any direction. This
simplification forms the basis for the earthquake re- Structural Response to Ground
quirements in the NBC. Particular attention should
be paid, however, to the effect of the combined
Motion [NBC 4.1.9.1.(4)]
stresses at the external and re-entrant corners,
13. The elastic response of a single-degree-of-
which are especially vulnerable to the effect of con-
current translational and torsional motions.(1) freedom system to ground motion depends on the
fundamental period and the damping characteristics
Vertical Accelerations of the system, and on the frequency content and
amplitude of the ground motion. The base shear
11. The multi-directional ground motion during which can be used as a measure of this response is
an earthquake may contain a substantial vertical expressed as the product of the mass of the system

Commentary J
and the spectral acceleration as given by the re- Equivalent Lateral Seismic Force
sponse spectrum. Representing Elastic Response, Ve
[NBC 4.1.9.1.(5)]
The acceleration response spectrum reflects the
dependence on the natural period, T, the damping 18. The lateral force
of the system and also the characteristics of the
ground motion. Details of its derivation can be

found in Reference (19). (2)
14. This concept of calculating base shear can be

is a good approximation of the base shear of an
applied to multi-storey buildings having many
elastic structure having nominal damping and sub-
modes of vibration. For usual buildings of low or
jected to a seismic ground motion characterized by
moderate height, the principal earthquake response
peak ground velocity, for which S represents an ide-
is due to the fundamental mode of vibration. For
alized elastic response spectrum.(25) This is a useful
taller structures, some allowance for contributions of
reference in terms of the response of a structure
the higher modes is made in the base shear calcula-
having period T. Therefore, in the 1995 NBC, is
tions in most building codes,(20) including the NBC.
taken as the “equivalent lateral force representing
the elastic response.” This elastic response parame-
15. With the onset of inelastic material be- ter is then modified by R to account for various
haviour, the base shear induced by earthquakes is forms of energy dissipation and by the calibration
reduced as compared to that of elastic behaviour. factor U, to arrive at the base shear force, V.
Such a reduction in base shear is implicit in the pro-
visions of most building codes, including the NBC. Zonal Velocity Ratio, v
The lower base shears, however, are justified only if
a structure possesses ductility, i.e., the capacity to 19. The zonal velocity ratio, v, is derived from
deform beyond the yield point without major struc- the probabilistic study of ground motions in the 1 s
tural failure.(21)(22)(23) period range, normalized to a spectral velocity of
1.0 m/s. Similarly, a zonal acceleration ratio, a, was
Minimum Design Earthquake Forces obtained from probabilistic evaluation of the ground
[NBC 4.1.9.1.(4)] motions in the 0.2 s period range, normalized to
1.0 g. At any location, the two ground motion pa-
16. The NBC specifies that a structure should be rameters, a and v, affect the behaviour of structures
designed for a minimum base shear, V, given by in the short period and intermediate period range,
respectively. In the formulation for the NBC seismic
provisions, only the zonal velocity ratio, v, is speci-
fied explicitly, whereas the zonal acceleration ratio,
(1) a, is used implicitly via the seismic response factor,
S.(15) Values for v and the corresponding velocity-
related seismic zone, , and accceleration-related
where is the equivalent lateral seismic force rep- seismic zone, , are given in Appendix C of the
resenting elastic response, R is the force modification NBC.
factor, and U = 0.6 is a calibration factor. This cali-
bration factor is applied to maintain the design base 20. The seismic hazard maps for peak ground
shears at the same level of protection for buildings acceleration and for peak ground velocity having a
with good to excellent capability of resisting seismic probability of exceedance of 10% in 50 years are
loads consistent with the R factors used.(24) Each of shown as Figures J-1 and J-2. The correspondence
these factors will be discussed in greater detail. between zone boundaries, zone numbers and zonal
acceleration and velocity ratios is presented in
17. The base shear, V, corresponds to that at the Table J-1.
ultimate limit state, where the structure is assumed
to be at the point of collapse. For the 1990 NBC, the Seismic Response Factor, S
seismic load factor in Subsection 4.1.3.. was sepa- [NBC 4.1.9.1.(6)]
rated from that of wind and assigned a value of 1.0.
in contrast to 1.5 for the 1985 NBC. This follows 21. The seismic response factor, S, is plotted in
from the argument that the seismic ground motions Figure J-3. The factor is a function of the fundamen-
specified in the NBC should be considered to be ex- tal period, T, of the structure, and the relative
treme or accidental actions, rather than live loads or values of the velocity-related and acceleration-
variable actions, as for example, live or wind loads. related seismic zones, and , pertaining to a
(25) Although the load factors have changed, the fac- particular geographical location. This factor repre-
tored base shear has remained approximately sents the idealized elastic response of 5% damped
constant compared to the 1985 NBC. multi-degree-of-freedom systems, for unit values of

Commentary J
the zonal velocity ratio, v, and weight, W. Research formulae are used. The new formulae for steel and
assessing the validity of S, on the basis of dynamic concrete moment resisting frames are generally
response studies using a number of representative more accurate than T = 0.1N and can be an alterna-
earthquake records, is reported in Reference (26). tive to this value.
24. For pure shear wall structures, the natural

period is generally closer to the true building period

when rather than D, also expressed in metres, is
used in Equation (4). However, when the length of
Za > Z v
the lateral force resisting system is not well defined,
Seismic response factor, S
4.2
then the Code requires that D shall be used instead
Za = Z v
of .
3.0
S = 1.5/ T 25. The period, T, for a structure may be deter-

Za < Z v
2.1 mined by more refined methods of calculation. One
such method is to use the following expression,
which represents the Rayleigh approximation for
determining the natural period of the fundamental
mode:
0 0.25 0.50 0.75 1.00 1.25
Period, T, s
Figure J-3
Seismic response factor S for 1995 NBC. (5)
where (i = 1 ... n) are the elastic deflections in

22. For the case where > , the maximum storeys i, due to the forces applied horizontally
value of S in the short period range, T < 0.25, has at storeys i. is as defined in Sentence 4.1.9.1.(2),
been limited by the multiplier 1.4 relative to the g is the acceleration due to gravity and the units
case where = , whereas for < the mini- used must be consistent.
mum value is limited by the multiplier 0.71. This is
for code purposes, to prevent extreme variations in 26. When the fundamental period is calculated
lateral force due to differing velocity and accelera- by the Rayleigh method or any other analytical
tion zones. The evolution of the seismic response method, the NBC requires that the resulting base
factor, S, over the last few decades is reviewed in shear be not less than 0.8 times the base shear ob-
Reference (23). tained by using T by the Code formulae. The reason
Fundamental Period, T [NBC 4.1.9.1. (7)] for limiting the calculated base shear is that the
Code is primarily calibrated based on the period
23. The following empirical formulas are to be calculation of Equations (3) and (4). Large devia-
used for the determination of the fundamental pe- tions of T resulting in significant reductions in base
riod, T, for buildings: shear are undesirable.
for moment resistant space frames only,
Force Modification Factor, R
[NBC 4.1.9.1.(8)]
(3)
27. The force modification factor, R, assigned to
different types of structural systems reflects design
for all others and construction experience, as well as the evalua-
tion of the performance of structures in major and
moderate earthquakes. It endeavours to account for
(4) the energy-absorption capacity of the structural sys-
tem by damping and inelastic action through
several load reversals. Types of construction that
where and must be expressed in metres. have performed well in earthquakes are assigned
Other symbols are defined in Sentence 4.1.9.1.(2) of higher values of R. The values of R in Table
the NBC. Period calculations based on the formulae 4.1.9.1.B. of the NBC recognize the following:
given in NBC Sentence 4.1.9.1.(7) give values that (1) The capability of a structure to absorb energy,
are, for the most part, in reasonable agreement with within acceptable deformations and without
measured values.(27)(28) However, variations in T in failure, is a very desirable characteristic of any
the order of 50% have been observed when these earthquake-resistant design.

Commentary J
(2) The existence of alternate load paths or redun- Importance Factor, I [NBC 4.1.9.1.(10)]
dancy of a structural system is a desirable
characteristic. It increases the locations where 31. Normally, structures are assigned an I factor
energy can be dissipated and reduces the risk of 1.0. For school buildings, which are usually well
of collapse when individual members should distributed throughout a community and whose
fail or become severely damaged. grounds and buildings may be needed for post-
(3) A building designed with a value of R greater disaster services, the importance factor is assigned a
than 1.0 is presumed to be capable of undergo- value of I = 1.3, which gives an increased degree of
ing inelastic cyclic deformations. Members and life safety for schools.
connections in such systems must therefore be
detailed to accommodate these deformations 32. Some structures are designed for essential
in a ductile manner. public services, and these post-disaster structures
(4) Buildings are normally endowed with a multi- should be operative immediately after an earth-
plicity of nonstructural elements and resisting quake. Examples of such structures are buildings
elements not considered in the analysis. Fur- that house or control electrical generating and dis-
thermore, buildings generally have higher tribution systems, fire and police stations, hospitals,
damping values during large amplitude vibra- radio stations and towers, telephone exchanges, wa-
tions than do mere skeleton structures. ter and sewage pumping stations, fuel supplies and
civil defence buildings. Such structures are assigned
28. The R values for structural systems need to an importance factor of I = 1.5. However, the mere
reflect the degree of continuity and ductility pro- application of I = 1.5 will not necessarily ensure op-
vided. A building with a value of R equal to 1.0 erational readiness of a facility after an earthquake.
corresponds to a structural system exhibiting little The requirements for a functional survival would
or no ductility. Values of R higher than 1.0 reflect entail a detailed study of the anticipated behaviour
the fact that structures can be designed and detailed of equipment and structural components. Building
to accommodate the corresponding inelastic cyclic contents, such as equipment and services, which are
deformations. The definitions, together with the required to remain functional immediately after an
minimum design and detailing requirements for earthquake, should be capable of accommodating
each of the specific cases in Table 4.1.9.1.B., are the deflections specified in Sentence 4.1.9.2.(2).
given in the corresponding material standards. These more stringent drift requirements, coupled
with the factor I = 1.5 for post-disaster structures
29. In choosing the structural system for a and for parts and portions of such structures, are
building, large dissimilarities in the stiffness and intended to provide a higher probability of main-
ductility characteristics of framing systems in the or- taining these functional requirements immediately
thogonal directions should be avoided. For example, after a major earthquake.
a moment-resistant ductile frame in one direction
and reinforced masonry walls in the other would be 33. The factor I = 1.5 is not intended to cover
unsuitable, whereas reinforced concrete ductile flex- the design considerations associated with special
ural walls and reinforced concrete walls with purpose structures whose failure could endanger
nominal ductility in orthogonal directions would be the lives of a large number of people or affect the
acceptable. The reason for this recommendation is environment well beyond the confines of the
that seismic displacements induced in flexible fram- building. These special purpose structures would
ing systems would probably cause failure in the include, for example, facilities for the manufacture
relatively brittle and weak directions of elements or storage of toxic materials and may require more
that resist the load in the orthogonal direction. sophisticated analysis.
30. Internal structures such as multi-storey racks Foundation Factor, F [NBC 4.1.9.1.(11)]
that are free-standing on the ground but are sur-
rounded by, but not otherwise connected to, the 34. The soil conditions at a site have been
building structure should be analyzed as separate shown to exert a major influence on the type and
structures. Separation by Sentence 4.1.9.2.(6) must amount of structural damage resulting from an
also be satisfied. Appropriate R values need to be earthquake.(29)-(32) As the motions propagate from
chosen in accordance with the structural system em- bedrock to the surface, the soil may amplify the mo-
ployed, and adequate resistance to the lateral forces tions in selected frequency ranges around the
needs to be provided throughout the height of the natural frequencies of the surficial layer. In addition,
structure. When connections are made between the a structure founded on the surficial layer, with some
racks and the enclosures, the combined system has natural frequencies close to those of the layer, may
to resist the lateral seismic forces in proportion to undergo even more intense shaking due to the de-
the relative stiffnesses of the components. velopment of a state of quasi-resonance between
structure and foundation soil. Direct calculation of
these effects by suitable mathematical models such

Commentary J
as lumped masses, wave propagation and by two- Lateral Force Distribution

dimensional finite element models using realistic [NBC 4.1.9.1.(13)]
soil properties is possible with the assumption that
the earthquake motions are shear waves propagated 40. For translational vibrations, the Code
vertically from bedrock. This simplified model of formula assumes that the building response is pri-
wave transmission ignores the source mechanism of marily due to its fundamental mode and the mode
the earthquake, the geology of the travel path and shape is assumed to be linear. As the inertial forces,
the effect of surface waves. Because of the uncer- , induced at any level, x, are proportional to the
tainties and complexities in the realistic estimation weight, , at that level, the distribution of seismic
of site effects on the seismic response of structures forces is approximated by
and ground, only a rough allowance for site effects
can be made at this time.
35. The seismic provisions of the NBC incorpo-

rate site effects by categorizing the wide variety of
possible soil conditions into four types and assign- (6)
ing values to a foundation factor, F, as per Sentence
4.1.9.1.(11), depending on soil type and depth. The
factor, F, reflects experience with these soil condi- For a lumped mass system the use of this equa-
tions in the field, and in an approximate way tion is reasonable for structures with fundamental
integrates the effect of possible soil amplification periods less than about 0.7 s. For buildings having
and soil-structure resonance into the estimation of longer periods, higher forces are induced at the top
the seismic design forces for buildings having no portion of the structure due to increasing contribu-
unusual structural characteristics. tions to top storey amplitudes by all the contributing
modes.(21)(37) This redistribution of forces is ac-
36. Sites underlain by deposits of very soft to counted for by applying part of the base shear as a
soft fine-grained soils with depths greater than 15 m concentrated force, , to the top of the structure, in
are assigned a foundation factor F = 2.0. This which case V in Equation (6) is replaced by V – .
provision is based on the observation of large am- The top force, , increases with the period of the
plifications of incoming earthquake motions in the structure, starting at T = 0.7 s and reaching a maxi-
clay deposits of Mexico City during the September mum of 25% of the base shear. This is considered a
19, 1985 earthquake.(29)(32)(33) more rational approach to account for the effects of
higher modes in tall structures than the criterion of
slenderness employed in the 1985 NBC.
37. When a site is underlain by a number of dif-
ferent soil layers, an F value appropriate to some 41. , the shear force for the top storey, and
average conditions may be used. However, when , the shear force transmitted to the supporting
substantial layers of soft clay are present, F = 1.3 structure just below the level i = x, are given by
should be used if the total thickness of clay is less
than 15 m and F = 2.0 when the thickness is 15 m or
greater.
(7)
38. The seismic design procedures outlined in
the NBC are based on the assumption that the struc-
tures are founded on a rigid base moving with the
ground surface motion. Real foundations possess
(8)
both flexibility and damping capacity, which alter
structural response. The flexibility of the foundation
increases the fundamental period of a structure, and
the damping capacity arises from dissipation of en- Dynamic Analysis
ergy by radiation away from the structure and by [NBC 4.1.9.1.(13)(b)]
hysteretic damping in the foundation, thus increas-
ing the effective damping of the structure. These 42. Sentence 4.1.9.1.(13) states that the distribu-
effects are described as soil-structure interaction and tion of forces in the building may also be
are not considered explicitly in the NBC. For most determined by a dynamic analysis. This would
buildings covered by the NBC, neglecting soil- apply especially to buildings with significant irregu-
structure interaction results in conservative design. larities either in plan or elevation, and buildings
with setbacks or major discontinuities in stiffness or
39. Liquefaction potential of foundation soils mass. Performing a dyamic analysis will lead to a
may be assessed as outlined in References (34), (35) better representation of modal contribution in tall
and (36). buildings.

Commentary J
43. An independent determination of base shear when the periods are well separated. For build-
using a dynamic analysis is not intended in this ings having closely spaced periods, a more
procedure. The dynamic analysis in this Commen- refined rule such as the complete quadratic
tary is based on the linear spectral modal technique. combination (CQC) rule should be used.(38)
If the results of such calculations yield a lower base Enough modes should be included that at least
shear, they have to be calibrated to the static base 90% of the participating mass of the structure
shear as required in NBC Subclause 4.1.9.1.(7)(c). is included in the calculation of response for
The reason is that Code values represent the mini- each principal horizontal direction.
mum values commensurate with an acceptable level (e) To obtain the design responses of interest, mul-
of public safety. tiply the dynamic responses of interest by V/
Vdyn, where V is as given in Sentence 4.1.9.1.(4)
44. The dynamic analysis should be carried out (f) When a two dimensional mathematical model
according to the following procedure: is used, torsional effects need to be considered
(a) Multiply the normalized 5% damping spec- statically as specified by Clause 4.1.9.1.(28)(a)
trum given in Figure J-4 by v to obtain the and combined with the design parameters of
design spectrum. interest described in Clause (e) to arrive at the
(b) Determine the natural periods, the associated final design values.
mode shapes and the modal participation fac- (g) When a three dimensional mathematical
tors. model is used, the accidental torsional effects
(c) Obtain the modal base shears and other modal need to be considered statically as specified by
responses of interest. Clause 4.1.9.1.(28)(b) and combined with the
(d) Obtain the dynamic base shear, Vdyn, and other design parameters of interest described in
dynamic responses of interest using a suitable Clause (e) to arrive at the final design values.
mode combination rule. The square-root-sum-
square (SRSS) combination rule may be used
10
8
6
4
80 0
10 8
10
60
6
40
2
4
20
2
sp
m
ec
10
d,
Spectral velocity, S v, m/s
1
8
S
1
tra
0. 6
6
8
la
0.
0.8
l.,
4
cc
isp
0.
0.6
4
el
ld
.,
0.
tra
2
0.4
Sa
ec
1
0. 8
,
0. 8
0. 06
sp
0.
1
g
Za > Z v
6
0.
4
0.
0.
04
0.2
2
0.
0.
02
Za < Z v
0. 0.1
Za = Zv
0. 008
0.
0. 08
0.1
01
0.
06
00
0.08
6
0.
04
00
0.
4
0.06
0.
02
00
0.
2
0.04
0. 008 1
0.
0. 0.0
00
1
6
00
4
00
0.02
0.
2
00
0.
1
00
0.01
0.
0.02 0.04 0.06 0.08 0.2 0.4 0.6 0.8 2 4 6 8 20 40 60 80

0.01 0.1 1 10 100
Period, T, s
Figure J-4
Normalized design distribution spectrum for peak horizontal ground velocity v = 1 m/s

Commentary J
45. The normalized design distribution spec- the structural design of post-disaster buildings and
trum shown in Figure J-4 was derived from a schools.
response spectrum for 5% damping.
51. An exemption is made for some non-
46. For cases where more detailed knowledge of structural components in non-post-disaster
the building motion or forces near the ultimate limit buildings located in zones where Za and Zv = 1.0 or
state is required, a non-linear step-by-step approach less and F is equal to or less than 1.3. This is to
is indicated, with consideration of a number of ap- reflect the low risk to life safety posed by seismic ef-
propriately selected ground motions. These motions fects on the non-structural components such as
might be chosen from both historical records and suspended ceilings and mechanical/electrical equip-
artificially generated ones which incorporate the de- ment in buildings located in low seismic zones.
sired characteristics pertaining to the site and Items in categories 1 to 6 of Table 4.1.9.1.D. are not
seismological considerations. A comprehensive trea- exempt in these provisions.
tise on aspects of earthquake engineering and
determination of seismic response is presented in Architectural Components [NBC
Reference (39). 4.1.9.1.(17), Table 4.1.9.1.D.]
Parts of Buildings [NBC 4.1.9.1.(15), 52. Reduced forces are permitted for connection
Tables 4.1.9.1.D., 4.1.9.1.E.] design provided that the connection responds in a
ductile manner. This provision is intended to enable
47. The present provisions for non-structural designers to develop connections which respond in
building components have been rationalized to re- a controlled ductile fashion and to avoid brittle fail-
flect current understanding of force levels and ures. To avoid non-ductile response, the yield
practice. Significant departures from previous edi- mechanism should be confined to the body of the
tions of the NBC have been made to reflect current connection. This requires that the attachment of the
thinking. Generally, force levels are lower or similar connection to the component and the structure
to those in previous codes. should be substantially stronger than the body of
the connection.
48. During an earthquake, components attached
to a building respond to support excitations which 53. Examples of damage to chimneys in recent
are the ground motions modified, amplified and fil- earthquakes have led to an increase in the force fac-
tered by the supporting building. As a result, the tor specified.
excitations are very different from the incoming
ground motions. To allow for this effect and others, 54. Suspended ceilings have been included as a
the design forces for components of buildings are special category. It should be emphasized that the
generally higher than the forces specified for the de- lateral forces specified for suspended ceilings
sign of the building structure. There are four main should take into account the weight of all compo-
reasons for the higher force levels specified for non- nents attached to the ceiling system. A minimum
structural building components: loading of 0.19 kPa (4 psf) is recommended. All me-
(a) support motion of the equipment is higher chanical/electrical components of the ceiling system
than ground motion; should be attached directly to the building structure.
(b) additional amplification can take place within
the component; 55. Force factors for the attachment of veneers
(c) components usually have limited ductility; to the building structure are given as a new cate-
(d) highly concentrated forces can lead to connec- gory in Table 4.1.9.1.D.
tion failures.
56. Factors for access floors are not included in
49. The seismic response of non-structural Table 4.1.9.1.D. There is limited experience available
components can also be greatly affected by the de- on the seismic response of these systems. However,
formations of the building to which they are considerable interest is currently being focused on
attached. The design of components and their at- the response of access floors in earthquakes. Until
tachments requires consideration of the deformation further information becomes available, a value of Sp
of the structural and non-structural elements, as = 2.0 is suggested.
well as force levels. Connections and attachments
must be ductile enough to accommodate the im- Mechanical/Eectrical Components [NBC
posed building displacements or be designed (for 4.1.9.1.(19), Table 4.1.9.1.E.]
example, slotted connections) to allow certain dis-
placements to occur without affecting the connector. 57. Some judgement has been employed in the
definition of rigidly mounted equipment as having a
50. The importance factor, I, is included in the fundamental period of 0.06 seconds or less. This pe-
force level calculation for portions of the structure riod range precludes amplification due to resonance
to establish compatibility of design risk levels with between the building and the component and allows

Commentary J
for the fact that the energy content of the ground tre of gravity of the masses. If the centre of mass
motion is generally very low at such short periods. and the centre of rigidity do not coincide because of
asymmetrical arrangement of structural elements or
58. A special case can be made for flexible pip- uneven mass distributions, torsional moments will
ing, ducting, electrical conduit and cable trays arise. The centres of rigidity are defined as the set of
which are manufactured from ductile materials and points located at floor levels such that lateral forces
which are attached to the building with ductile con- Fx and Ft acting through them will produce no floor
nections. Experience has shown that these systems rotations about a vertical axis. Their locations can be
are capable of extensive inelastic strains and defor- determined by a procedure given in Reference (47).
mations prior to failure. For such systems, where The designer should endeavour to make the struc-
ductile characteristics can be demonstrated by anal- tural system as symmetrical as possible and should
ysis or experience, an Ar factor of 1.0 is justified. consider the effect of torsion on the behaviour of
the structural elements.
59. The force factor, Ar for flexible or flexibly
mounted equipment located above ground has been
64. A realistic approach to aseismic torsional de-
reduced to take into account the limited capacity of
sign should consider the effect of the dynamic
isolator systems to transmit seismic forces to the
magnification of the torsional moments,(47–50) the ef-
component.
fect of simultaneous action of the two horizontal
60. Racking and shelving systems which are components of the ground disturbance, and acci-
connected to the building should be considered as dental torsion. Accidental torsion moments are
“fixtures” in Table 4.1.9.1.E. The components, as- intended to account for the possible additional tor-
semblies, connections and attachments of such sion arising from variations in the estimates of the
systems should be capable of withstanding the spec- relative rigidities, uncertain estimates of dead and
ified forces. live loads at the floor levels, addition of wall panels
and partitions after completion of the building, vari-
61. Some suggested design considerations and ation of the stiffness with time, and inelastic or
details are presented in References (40) to (46). The plastic action. The effects of possible torsional mo-
failures of interior partitions, finishes and hung ceil- tion of the ground should also be considered. For
ings also pose hazards to occupants. most practical situations, however, these concepts
and effects can only be accounted for by the use of
Overturning Moments adjustment factors.
[NBC 4.1.9.1.(23–27)]
65. The torsional provisions of the NBC deal
62. The lateral forces induced in a structure by
with the complex nature of torsion by increasing or
earthquakes give rise to moments which are the
decreasing the computed eccentricity by 50%,
product of the induced lateral forces times the dis-
whichever produces the worst effect in a member.
tance to the storey level under consideration, where
To use Clause 4.1.9.1.(28)(a) the algebraic sign of ex
they have to be resisted by axial forces and mo-
must be retained. The part played by accidental tor-
ments in the vertical load-carrying members. While
sion and torsional ground motion is recognized by
the base shear contributions of modes higher than
specifying an additional torsion due to an eccentric-
the fundamental can be significant, the correspond-
ity of 0.10 times the plan dimension perpendicular
ing modal overturning moments for the higher
to the direction of the loading.
modes are small. As the equivalent static lateral
base shear in the NBC also includes the contribu-
66. In lieu of determining eccentricity, ex, using
tions from higher modes for moderately tall and tall
Clause 4.1.9.1.(28)(a), the following procedure may
structures, a reduction in the overturning moments
be used.(47) The structure shall be analysed subjected
computed from these lateral forces appears justified.
to three load cases. Load case 1 consists of applying
This is achieved by means of the multiplier J as
the floor forces Fx at the mass centres (CM). Load
given in NBC Sentence 4.1.9.1.(23). If, however, a
case 2 consists of applying Fx at CM, but restraints
structure did respond exclusively in its fundamental
are imposed to prevent floor rotations about a verti-
mode, the overturning moment at the base would
cal axis. Load case 3 consists of applying torques
be the sum of the moments corresponding to the
equal to 0.1 FxDnx in the same direction at each floor.
forces, , about the base without any J-factor re-
Denoting the design parameters of interest (forces
ductions. A more refined method of accounting for
or deformation) that correspond to these load cases
the maximum overturning moments is through dy-
as r1, r2, and r3, respectively, implementation of
namic analysis.
Clauses 4.1.9.1.(13)(a) and 4.1.9.1.(28)(a) is achieved
Torsional Moments by using the combinations 1.5r1 − 0.5r2 + r3 or 0.5r1
[NBC 4.1.9.1.(28)] + 0.5r2 + r3, whichever gives the more severe effect.
63. The inertial forces induced in the structure 67. For structural elements to resist torsional
by earthquake ground motions act through the cen- moments most effectively, they should be located

Commentary J
near the periphery of the building, i.e., some dis- should be carried out to demonstrate that the offsets
tance from the centre of rigidity. Wall elements that are fully compatible with the setback conditions.
are intended for resisting torsional forces should be
oriented so that their in-plane forces are associated Deflection and Separation of
with as large a moment arm as possible about the Buildings [NBC 4.1.9.2.(1)-(9)]
centre of rigidity. In buildings with complete di-
aphragms, such as complete reinforced concrete 72. Deflection refers to the lateral deflection at
floor slabs, all elements interconnected by such any point in the structure relative to the ground. In-
members can be counted on to resist torsional forces. cremental deflection or interstorey deflection refers
to the lateral deflection of a storey relative to the
68. In core-type buildings, where all stiffening one just below it. The calculations of deflections are
elements are located in a central core away from the intended to be based on accepted practice and
periphery, accidental torsion and torsional ground 1
should include such items as P – effects, founda-
motion are particularly significant. In odd- and tion rotations, and the effects of cracked concrete
irregularly-shaped buildings (e.g., L-shaped), and in sections, when these have an important effect on the
buildings with the core located at one side or cor- structure.
ner, large torsional oscillations are induced by
horizontal ground motion. These are examples of 73. The requirement to calculate deflection as R
torsion situations that should be avoided in times the elastic deflection resulting from the design
building layouts. Torsional effects should also be earthquake forces as specified in the NBC includes
evaluated for parts of structures relative to the allowance for some plastic deformation to occur in
whole. For example, the torsional effects of project- the structural system.
ing wings on buildings should be considered in
relation to the motion of the building as a whole. 74. Deflection limitations have been established
in consideration of the acceptable damage to the
Setbacks [NBC 4.1.9.1.(29)] non-structural components. The specified deflections
appear reasonable approximate limits for the control
69. A setback is a sudden change in plan di- of damage to structures and their contents.(52) Signif-
mension or a sudden change in stiffness along the icantly lower deflection limits are required for
height of a building. For moderate or small changes, post-disaster buildings to ensure a higher probabil-
the equivalent static procedure in the NBC usually ity of functioning after an earthquake.
gives acceptable lateral forces. The effects of major
changes in stiffness or geometry are best investi- 75. The separation of two adjacent structures is
gated by dynamic methods (see Paragraphs 42–44). required to prevent collision of buildings in an
The fundamental period of a setback building may earthquake. Collision of buildings was observed to
be less than the fundamental period of a regular be destructive, particularly when adjacent buildings
building of the same height and number of storeys. have different heights or different storey heights.
Calibration to the minimum lateral seismic force These requirements also apply to the expansion
based on a fundamental period, T, as determined by joints within buildings. Such joints could be con-
Clause 4.1.9.1.(7)(a) or (b) may underestimate the nected for seismic forces, or special details could be
base shear.(51) developed to ensure local failure without damage to
the principal structural elements.
Structural Requirements for Setbacks
76. Sentence 4.1.9.2.(6) is inserted to ensure con-
70. In addition to the determination of the seis- sistency with Clause 4.1.9.1.(9)(b).
mic loads on the building, the “notch” effect must
be considered in design, particularly where the P-Delta Effects
tower framing does not extend downward through
the base. A shear wall type of design for both a 77. When a flexible building is subjected to lat-
tower and base could produce severe stresses at and eral seismic forces, the gravity loads acting through
about the 90 notches. the lateral displacements lead to additional mo-
ments throughout the structure. These additional
71. Some setbacks may consist of simple one- P-Delta moments reduce the capacity of the struc-
storey penthouses, whereas others may constitute ture to resist storey shears.
substantial portions of the entire building. In view
of the dynamic and also the notch effect phenomena 78. P-Delta effects(53)(54)(55) have only a small in-
that may occur, the lateral load resisting elements fluence on the response of buildings to seismic
should be vertically continuous through the setback forces, provided that storey shear capacities exceed
portion. If the lateral load resisting elements are not certain minimum values. When shear capacities fall
vertically continuous through the setback portion below these minimums, the displacements during
and all the way to the foundations, a special analysis earthquakes can become unacceptably large.

Commentary J
79. Currently, there is no widely accepted the deflections of a properly designed structure of
method for accounting for P-Delta effects under the same type, assuming that the P-Delta effects are
seismic loading. Until an acceptable method is de- small enough to be ignored.
veloped, it is recommended that the earthquake
induced forces, shears, overturning moments and 85. If the stability factor calculated from Equa-
torsional moments calculated at each storey level be tion (9) is less than about 0.10, then P-Delta effects
multiplied by an amplification factor of (1 + x) to can often be ignored. When the stability factor is
allow for P-Delta effects, where x is a stability fac- more than 0.40, the structure should be redesigned
tor. The stability factor at level x is equal to to guard against potentially unstable buildings dur-
ing extreme earthquakes.
86. The procedure described above is similar to

that recommended by Paulay and Priestley.(56) This
reference presents a comprehensive discussion of
(9) the P-Delta problem as it relates to the design of
buildings with moment-resisting frames.
80. In the above expression, is the seismic 87. The procedure described above is also simi-
lar to that used by CAN/CSA-S16.1-94 and CAN/
shear force at the level under consideration, which CSA-A23.3-94 to account for P-Delta effects in struc-
is equal to the sum of the design lateral forces act- tures that respond elastically to lateral loads. In
ing at and above the storey under consideration as these standards, the translational moments obtained
determined in Sentence 4.1.9.1.(13), is that from a first-order elastic analysis are multiplied by
an amplification factor, which is equivalent to 1/(1 -
portion of the factored dead plus live load above x), to obtain an estimate of the translational mo-
1
the storey under consideration, mx is the maximum ments that would be calculated from an elastic
inelastic interstorey deflection as defined in second-order analysis (an analysis with P-Delta ef-
Sentence 4.1.9.2.(2), and hs is the interstorey height. fects included). For seismic design, an amplification
factor of (1 + x) is used since, with the increased
81. The amplification factor of (1 + x) need not seismic shear capacities required by Equation (10),
be applied to displacements. the maximum inelastic interstorey displacements
82. The procedure recommended to allow for P- with an allowance for P-Delta effects are assumed to
Delta effects is equivalent to proportioning the be the same as those calculated by the Code proce-
structure at each level x to resist an increased seis- dures.
mic shear force calculated from 88. Both structural and non-structural compo-
nents should either be isolated or be accounted for
in the design. The non-structural components of the
structure should be detailed so as not to transfer to
the structural system any forces that are not ac-
counted for in the design. If non-structural members
(10) are designed as isolated units, their connections
should be detailed so as to be capable of accommo-
dating the anticipated movement due to drift and
temperature changes.(57) If, on the other hand, these
83. In evaluating , the dead load factor components are rigidly attached to the structure,
should be taken as 1.0, and the live load factor then their effect on the behaviour of the structure
should be taken as 1.0 for storage and assembly should be considered and allowed for in the elastic,
loads or 0.5 for other live loads, including snow. plastic and fracture stages. An example is the stair
The live load may be reduced for large tributary that may act as a stiffening element. Failures of some
areas in accordance with Sentence 4.1.6.9. Accord- buildings in the Caracas 29 July, 1967 earthquake
ingly, is an estimate of the actual gravity load were caused by the partition tile walls, which acted
as shear walls, thus changing the relative rigidity of
acting at the storey under consideration at the time the bents from that assumed in the design.(58)
of an earthquake.
Special Provisions
84. With the seismic shear capacities at each [NBC 4.1.9.3.(1)-(5)]
storey increased to allow for P-Delta effects, the
ability of the strengthened structure to absorb 89. Unreinforced masonry buildings have fared
inelastic energy during an earthquake is also in- badly when subjected to earthquakes.(1) The pres-
creased. The interstorey deflections of the ence of reinforcing embedded in mortar or grout
strengthened structure should be about the same as increases ductility and reduces the likelihood of

Commentary J
brittle failure. Examples and detail sheets for the buildings approximately 60 m high have shown that
seismic design of reinforced masonry can be found if damage is to be minimized in a moderate earth-
in References (57) and (59). quake, the structure must incorporate ductile
features that are associated with these high R val-
90. The NBC specifies that, except in velocity- ues.(21) Alternatively, extra strength is to be provided
or acceleration-related seismic Zones 0 and 1, build- by increasing the design base shear, V. The height
ings shall have structural systems as described by limitations associated with the provisions of NBC
Cases 1–8, 10–14, 16–18 or 20–21 in Table 4.1.9.1.B. Article 4.1.9.3. and the cases described in Table
For velocity-related Zone 4 and higher, buildings 4.1.9.1.B. are summarized in Table J-3.
over 60 m high having a structural system of R =
1.5 or 2.0 shall increase V by 50%. Analyses of
Table J-3
Summary of Cases Permitted by NBC Article 4.1.9.3.
Height of Building Velocity- or Acceleration- Velocity- or Acceleration- Velocity-Related Seismic Zones

Related Seismic Zones 0 and 1 Related Seismic Zones 2 and 3 4 and Higher
Up to 3 storeys in 1–23 1–23, except unreinforced 1–23, except unreinforced
building height masonry masonry
Greater than 3 storeys 1–23 1–8, 10–14, 16–18, or 20–21 1–8, 10–14, 16–18 or 20–21
but not more than 60 m
in building height
Greater than 60 m in 1–23 1–8, 10–14, 16–18 or 20–21 1–8, 10–14, 16–18 or 20–21; if
building height R = 1.5 or 2.0 increase V by
50%
91. The category of “elevated tanks on 4 or Foundation Requirements

more cross-braced legs” in the 1985 NBC was
broadened to include all elevated tanks, since they Mismatch of Strength between Structure
can be classed as inverted pendulum structures, and Foundation [NBC 4.1.9.4.(1)]
which are sensitive to seismic excitation. Elevated
tanks have been separated from Table 4.1.9.1.B., 94. Conventional seismic structural design is
since they are not buildings in the sense of the other based on acceptance of damage without collapse in
structures referred to in that Table. the event of a severe earthquake. This implies that
plastic deformation (defined loosely to include
cracking of reinforced concrete, yielding of steel,
92. The structural configuration should be such etc.) is permitted to occur at suitable locations in the
that elastic-plastic action in the members or failure structure. The maximum earthquake loads that are
of individual elements will not produce instability transmitted by the structure to its foundation are,
or initiate progressive collapse. Abrupt changes in therefore, governed not by the design loading, but
geometry, stiffness and mass are generally undesir- by the load levels at which yielding takes place in
able from the point of view of seismic response; the structural elements that transfer the lateral loads
yielding tends to concentrate at such discontinuities to the foundation, such as bracing members, frame
and, therefore, such configurations require special members and shear walls.
attention.
95. This load transfer can result in an overload
93. Sentence 4.1.9.3.(5) of the NBC requires that of the foundation during an earthquake if the struc-
masonry in velocity- or acceleration-related zones of ture does not yield at the expected lateral load level
2 and higher be reinforced. These elements include because some of its elements are stronger than re-
exterior loadbearing and non-loadbearing walls, quired. In this case, the foundation will be forced to
parapet walls, interior loadbearing walls and non- receive that higher seismic load and consequently
loadbearing partitions that weigh more than suffer additional distress or damage as compared to
200 kg/m2 or are more than 3 m high. Masonry ele- the case where the structure yields at a lower lateral
ments around stair and elevator shafts are required load level. The problem of mismatch in strength be-
to be reinforced to ensure a safe exit from the struc- tween the structure and the foundation requires
ture. continual liaison between the structural and founda-

Commentary J
tion designers to ensure that the foundation 103. Parking garages used for storage of cars or
elements are designed to be compatible with the parking need not be considered as storage areas in
earthquake loads transmitted from the structure. calculating W.
96. If the designer intends to provide energy Evaluation and Ugrading of Existing
dissipation mechanisms in the foundation rather Buildings
than in the superstructure, the areas of yielding in

the foundation must be defined clearly. The ductil-
104. While the seismic provisions of the
ity imposed on the potential plastic hinges in the
National Building Code are primarily intended for
foundation system should be checked.
the design of new buildings, they can also be used
as criteria for the evaluation of the seismic ade-
Soil Pressures on Basement Walls quacy of existing ones. The general principles
[NBC 4.1.9.4.(5)] pertaining to the structural evaluation of existing
buildings are discussed in Commentary K “Applica-
97. Basement wall pressures arising from tion of NBC Part 4 for the Structural Evaluation and
ground shaking may be determined according to Upgrading of Existing Buildings” of this document.
procedures described in References (60) and (61) or
by other generally accepted methods.
References
Design Considerations
(1) The Prince William Sound, Alaska, Earth-
quake of 1964 and Aftershocks. Vol. II,
98. Most failures in structures subjected to
Research Studies, Seismology and Marine
seismic loading can be traced to poor detailing, es-
Geology. Department of Commerce, Environ-
pecially at beam and column connections. This
mental Science Services Administration, Coast
becomes the governing factor in good aseismic be-
and Geodetic Survey, U.S. Government Print-
haviour of buildings built of precast elements.(1)
ing Office, Washington, 1967.
(2) The Niigata Earthquake, 16 June 1964, and
99. Floor systems that act as diaphragms Resulting Damage to Reinforced Concrete
should be studied to ensure that they are capable of Buildings. International Institute of Seismol-
distributing the loads to the various elements. ogy and Earthquake Engineering, Tokyo,
February 1965.
100. When the shear wall contains numerous (3) Canadian Standards Association, General Re-
openings, the design should account for its real be- quirements, Ground Motion Determination,
haviour under lateral loads, i.e., whether the wall Design Procedures, Testing Procedures, and
acts as a unit or as a number of units because of the Seismic Instrumentation Requirements for
reduced rigidities due to the openings. Overstress at Seismic Qualification of CANDU Nuclear
the openings should be examined. This is a com- Power Plants. CSA Standards CAN3-N289.1,
mon cause of damage to lintels above door N289.2, N289.3, N289.4 and N289.5, Rexdale,
openings and to piers between window openings.(1) Ontario.
Suggested design details are also given in Chapter (4) Canadian Standards Association, Liquefied
21 of CAN3-A23.3-94, “Design of Concrete Struc- Natural Gas (LNG) – Production, Storage and
tures for Buildings.” Handling. CSA Standard Z276-M1981, Rex-
dale, Ontario.
101. Special mechanical protection systems such (5) Canadian Standards Association, Code for the
as base isolation or controlled friction damping de- Design, Construction and Installation of Fixed
vices can significantly alter the seismic response of Offshore Production Structures. Preliminary
buildings.(62)-(65) It must be demonstrated through Standard S471, 1987, Rexdale, Ontario.
non-linear analysis and representative experimental (6) P.W. Basham, D.A. Forsyth and R.J. Wetmiller,
data that the building so equipped will perform at The Seismicity of Northern Canada. Can. J.
least equally well in seismic events as the same Earth Sci., Vol. 14, 1977, pp. 1646-1667.
building designed following the NBC seismic re- (7) W.G. Milne et al, Seismicity of Western
quirements. Canada. Can. J. Earth Sci., Vol. 15, 1978, pp.
1170-1193.
102. The fact that the design forces for wind are (8) R.B. Horner and H.S. Hasegawa, The Seismo-
greater than the seismic design forces (i.e., wind tectonics of Southern Saskatchewan. Can. J.
“governs” the design) does not obviate the need for Earth Sci., Vol. 15, 1978, pp. 1341-1355.
seismic detailing. While wind forces govern, the de- (9) J. Adams and P.W. Basham, The Seismicity
sign must provide at least the type of seismic and Seismotectonics of Canada East of the
detailing that corresponds to the seismic forces cal- Cordillera. In Geoscience Canada. Vol. 16, No.
culated for that building. 1, Geological Assoc. of Can. 1989, pp. 3-16.

Commentary J
(10) G.C. Rogers and R.B. Horner, An Overview of (24) S.M. Uzumeri, Development of Canadian
Western Canadian Seismicity. In Decade of Seismic-Resistant Design Code for Reinforced
North American Geology, Vol. GSMV-1, Geo- Concrete Buildings. Thomas Paulay Sympo-
tectonics of North America, Geological Soc. sium. Recent Developments in Lateral Force
Am., 1991, pp. 69–76. Transfer in Buildings, SP157, American Con-
(11) D.H. Weichert and W.G. Milne, On Canadian crete Institute, Detroit, 1995, pp. 301–330.
Methodologies of Probabilistic Seismic Risk (25) J.H. Rainer, Force Reduction Factors for the
Estimation. Bull. Seism. Soc. Am., Vol. 69, NBCC Seismic Provisions. Can. J. Civ. Eng.,
1979, pp. 1549-1566. Vol. 14, No. 4, August 1987, pp. 447-454.
(12) H.S. Hasegawa, P.W. Basham and M.J. Berry, (26) A.C. Heidebrecht and C.Y. Lu, Evaluation of
Attenuation Relations for Strong Seismic the Seismic Response Factor Introduced in the
Ground Motion of Canada. Bull. Seism. Soc. 1985 Edition of the National Building Code of
Am., Vol. 71, 1981, pp. 1943-1962. Canada. Can. J. Civ. Eng., Vol. 15, No. 3, June
(13) P.W. Basham et al., New Probabilistic Strong 1988, pp. 382-388.
Seismic Ground Motion Maps of Canada: A (27) A. Anderson et al., Lateral Forces of Earth-
Compilation of Earthquake Source Zones, quake and Wind. Paper No. 2514, Trans. Am.
Methods and Results. Earth Physics Branch Soc. Civ. Eng., Vol. 117, 1952, pp. 716-780.
Open File Report No. 82-33, Energy, Mines (28) Applied Technology Council, Tentative Provi-
and Resources Canada, 1982, 205 pp. sions for the Development of Seismic
(14) P.W. Basham et al., New Probalistic Strong Regulations for New Buildings. ATC 3–06,
Seismic Ground Motion Maps of Canada. National Bureau of Standards, U.S. Dept. of
Bull. Seism. Soc. Am., Vol. 75, 1985, pp. 43-75. Commerce, Washington, D.C., 1978.
(15) A.C. Heidebrecht et al., Engineering Applica- (29) W.D.L. Finn and A.M. Nichols, Seismic Re-
tions of New Probabilistic Seismic Ground sponse of Long Period Sites, Lessons from the
Motion Maps of Canada. Can. J. Civ. Eng., September 19, 1985 Mexican Earthquake. Can.
Vol. 10, 1983, pp. 670-680. Geotech. J., February 1988, pp. 128-137.
(16) C.A. Cornell, Engineering Seismic Risk Analy- (30) The Caracas Earthquake of July 29, 1967.
sis. Bull. Seism. Soc. Am., Vol. 58, 1968, pp. Venezuelan Official Seismic Commission,
1583-1606. Proc., 4th World Conf. on Earthquake Engi-
(17) R.K. McGuire, FORTRAN Computer Program neering, Santiago, 1969, Session J-2, Vol. 3, pp.
for Seismic Risk Analysis. U.S. Geol. Survey 75-86.
Open File Report 76-67, 1976, 90 pp. (31) H.B. Seed et al., Soil Conditions and Building
(18) L.G. Jaeger and A.D.S. Barr, Parametric Insta- Damage in 1967 Caracas Earthquake. J. Soil
bilities in Structures Subjected to Prescribed Mech. Found. Div., Am. Soc. Civ. Eng., Vol.
Periodic Support Motion. Proc., Symposium 98, No. SM8, August 1972, pp. 787-806.
on Design for Earthquake Loadings, McGill (32) D. Mitchell, J. Adams, R.H. DeVall, R.C. Lo
University, September 1966. and D. Weichert, Lessons from the 1985 Mexi-
(19) G.W. Housner et al., Spectrum Analysis of can Earthquake. Can. J. Civ. Eng., Vol. 13, No.
Strong Motion Earthquakes. Bull. Seism. Soc. 5, 1986, pp. 535-557.
Am., Vol. 43, No. 2, 1953, pp. 97-119. (33) The Mexico Earthquakes – 1985: Factors In-
(20) Earthquake Resistant Regulations, A World volved and Lessons Learned. Proc. of the
List 1984. Compiled by International Associa- International Conf., Mexico City, 1986. M.A.
tion for Earthquake Engineering, Tokyo. Cassaro and E.M. Romero, Eds., Am. Soc. Civ.
(21) R.W. Clough and K.L. Benuska, FHA Study of Eng., New York, 1987.
Seismic Design Criteria for High Rise Build- (34) H.B. Seed and I.M. Idriss, Ground Motions
ings. Report prepared for Technical Studies and Soil Liquefaction During Earthquakes.
Program and Federal Housing Administra- Monograph Series, Vol. 5, Earthquake Engi-
tion, HUD TS-3, August 1966. neering Research Institute, Berkeley, 1982,
(22) J. Penzien, Dynamic Response of Elasto- 134 pp.
Plastic Frames. Trans. Am. Soc. Civ. Eng., (35) H.B. Seed, Soil Liquefaction and Cyclic Mobil-
Paper 3284, Vol. 127, Part II, 1962. ity Evaluation for Level Ground During
(23) S.M. Uzumeri, S. Otani and M.P. Collins, An Earthquakes. J. Geotech. Eng. Div., Am. Soc.
Overview of Canadian Code Requirements Civ. Eng., Vol. 105, No. GT2, 1979, pp. 201-255.
for Earthquake Resistant Concrete Buildings. (36) H.B. Seed, I.M. Idriss, and I. Arango, Evalua-
Can. J. Civ. Eng., Vol. 5, No. 3, September tion of Liquefaction Potential Using Field
1978, pp. 427-441. Performance Data. J. Geotech. Eng. Div., Am.
Soc. Civ. Eng., Vol. 109, No. 3, March 1983,
pp. 458-482.

Commentary J
(37) J.A. Blume, Structural Dynamics of Cantilever (52) T.Y. Galambos and B. Ellingwood, Serviceabil-
Type Buildings. Proc., 4th World Conf. on ity Limit States: Deflection. J. Struct. Eng.,
Earthquake Engineering, Santiago, 1969, Ses- ASCE, Vol. 112, No. 1, January 1986, pp. 67-84.
sion A-3, Vol. 2, pp. 1-16. (53) D. Bernal, Amplification Factors for Inelastic
(38) E.L. Wilson, A. Der Kiureghian and E.P. Bayo, Dynamic P-Delta Effects in Earthquake Anal-
A Replacement for the SRSS Method in Seis- ysis. Journal Earthquake Eng. Struct. Dyn.,
mic Analysis. J. Earthquake Eng. Struct. Dyn., Vol. 15, 1987, pp. 635-651.
Vol. 9, 1981, pp. 187-194. (54) D. Bernal, Instability of Buildings Subjected to
(39) V. Davidovici, Ed., Génie Parasismique. Earthquakes. Journal Struct. Eng., ASCE, Vol.
Presses de l’école national des points et 118, No. 8, 1992, pp. 2239-2260.
chaussées, Paris, 1985, 1105 pp. (55) C. J. Montgomery, Influence of P-Delta Effects
(40) J.M. Ayres and T.Y. Sun, Criteria for Building on Seismic Design. Can. J. Civ. Eng., Vol. 8,
Services and Furnishings. Building Science 1981, pp. 31-43 with erratum on p. 273.
Series 46, National Bureau of Standards, Feb- (56) T. Paulay and M.J.N. Priestley, Seismic Design
ruary 1973, pp. 253-285. of Reinforced Concrete and Masonry Build-
(41) Anonymous, Design Could Mitigate Disaster ings. John Wiley & Sons, New York, 1992, 744
Results. Engineering News Record, November p.
15, l973, p. 13. See also November 29, 1973, (57) Seismic Design for Buildings. Dept. of Army
p. 64. Technical Manual TM5-809-10, Washington,
(42) C. Nguyen, A. Ghobarah and T.S. Aziz, February 1982.
Inelastic Response of Tuned Equipment- (58) M.A. Sozen et al., Engineering Report on the
Structure Systems, Proc., 5th Canadian Conf. Caracas Earthquake of 29 July, 1967. National
on Earthquake Engineering, Ottawa, July 6-8, Academy of Sciences, Washington, 1968.
1987, pp. 571-576. A.A. Balkema, Rotterdam/ (59) J.E. Amrhein, Reinforced Masonry Engineer-
Boston, 1987, 882 pp. ing Handbook. Masonry Institute of America,
(43) American Society of Civil Engineers, Struc- Los Angeles, 1978, 445 pp.
tural Analysis and Design of Nuclear Plant (60) N. Mononobe and H. Matsuo, On the Deter-
Facilities. Manual No. 58, New York, 1980. mination of Earth Pressures During
(44) Canadian Standards Association, Design Pro- Earthquakes. Proc., World Engineering Con-
cedure for Seismic Qualification of CANDU gress, 1929, Vol. 9, pp. 177-185.
Nuclear Power Plants. CSA Standard CAN3- (61) H.B. Seed and R.V. Whitman, Design of Earth
N289.3-M81, Rexdale, Ontario. Retaining Structures for Dynamic Loads. Am.
(45) Canadian Standards Association, Testing Pro- Soc. Civ. Eng., Specialty Conference – Lateral
cedures for Seismic Qualification of CANDU Stresses in the Ground and Design of Earth
Nuclear Power Plants. CSA Standard CAN3- Retaining Structures, Cornell University,
N289.4-M85, Rexdale, Ontario. Ithaca, NY, 1970.
(46) T.S. Aziz, Coupling Effects for Secondary Sys- (62) Applied Technology Council, Base Isolation
tems in Nuclear Power Plants. ASME Special and Passive Energy Dissipation. Proc. of Sem-
Publication PVP-Vol. 65, American Society of inar and Workshop, San Francisco, March
Mechanical Engineering, July 1982. 12-14, 1986 (ATC-17). Redwood City, CA,
(47) W.K. Tso, Static Eccentricity Concept for Tor- 1986, 478 pp.
sional Moment Estimation. Journal Struct. (63) A. Filiatrault and S. Cherry, Experimental
Eng., ASCE, Vol. 116, No. 5, May 1990, pp. Studies of Friction Damped Braced Steel
1199–1212. Frames. Proc., 5th Canadian Conference on
(48) R.K. Goel and A.K. Chopra, Seismic Code Earthquake Engineering, Ottawa, 6-8 July
Analysis of Buildings Without Locating Cen- 1987, pp. 867-873. A.A. Balkema, Rotterdam/
ters of Rigidity. Journal Struct. Eng., ASCE, Boston, 1987, 882 pp.
Vol. 119, No. 10, October 1993, pp. 3039-3055. (64) A.S. Pall, V. Verganelakis and C. Marsh,
(49) J.I. Bustamante and E. Rosenblueth, Building Friction Dampers for Seismic Control of Con-
Code Provisions on Torsional Oscillations. cordia University Library Building. Proc. 5th
Proc., 2nd World Conf. on Earthquake Engi- Canadian Conference on Earthquake Engi-
neering, Japan, 1960, Vol. 2, pp. 879-892. neering, Ottawa, 6-8 July 1987, pp. 191-200.
(50) W.K. Tso, and K.M. Dempsey, Seismic Tor- A.A. Balkema, Rotterdam/Boston, 1987,
sional Provisions for Dynamic Eccentricity. J. 882 pp.
Earthquake Eng. Struct. Dyn., Vol. 8, 1980, pp. (65) I.D. Aiken, J.M. Kelly and A.S. Pall, Seismic
275-289. Response of a Nine-Story Frame with Friction
(51) C.M. Wong and W.K. Tso, Seismic Loading for Damped Cross-Bracing. Report No. UCB/
Buildings with Setbacks. Can. J. Civ. Eng., Vol. EERC-88/17, Earthquake Eng. Research Cen-
21, No. 5, October 1994. tre, University of Calif., Berkeley, 1988, pp. 1-7

92
Commentary K
Application of NBC Part 4 for the
Structural Evaluation and Upgrading of
Existing Buildings
Introduction into account in the structural criteria of Part 4
and referenced structural design standards.
1. This Commentary concerns the structural
evaluation and upgrading of an existing building to 4. To help overcome these difficulties, this Com-
achieve a level of performance which is appropriate, mentary provides guidance on the application of the
based on the intent of the current National Building requirements of Part 4 to existing buildings, includ-
Code requirements. Buildings which satisfy the ing relaxations where appropriate, and alternatives
guidelines provided here should generally be con- where available (usually by reference to other docu-
sidered acceptable. More stringent criteria may be ments). NBC Subsection 2.5.2. allows structural
appropriate for buildings used for post-disaster ser- alternatives which are equivalent to Part 4 but, ex-
vices. cept for load testing, they are directed primarily to
new construction. Except as recommended in this
2. This Commentary does not apply to new Commentary, structural equivalence should comply
additions to an existing building structure or to a re- with the requirements of NBC Subsection 2.5.2. and
view of newly constructed work which was required Appendix note A-2.5.2.
to be in conformance with the current codes and
standards. In both of these applications, NBC Part 4 5. Earthquake requirements provide the greatest
applies without any of the relaxations described in difficulty in the application to existing buildings of
this Commentary. New additions, however, may in- Part 4 and referenced structural design standards.
crease loads on the existing building structure. More specific guidelines to address the seismic eval-
uation and upgrading of existing buildings have
3. Part 4 of the National Building Code and the
been developed separately from this Commentary,
structural standards referenced by Part 4 are written
as discussed in Paragraphs 38-42.
primarily for the design of new buildings (or new
additions), not for the evaluation and upgrading of 6. This Commentary does not specify the cir-
existing buildings. As a consequence, difficulties cumstances which would require a structural
have arisen: evaluation of an existing building. Typical situations
• Many current requirements specify quantities where structural evaluation becomes necessary in-
and arrangements of materials (such as rein- clude change of use of the building, damage or
forcing details in masonry and concrete deterioration, and where the safety of the building
structures) which are economical and practical is a concern because of known or potential defects.
to implement during initial construction but
impractical after a structure is completed. In 7. After the evaluation and before any upgrad-
such cases, alternative solutions are needed. ing, any life-safety implications of the conclusions
• Many older buildings consist of structural sys- of the evaluation should be discussed with the
tems, components or materials which are not owner and authority having jurisdiction to establish
addressed by the structural design standards the timetable for the work to be done. Each case
referenced by Part 4. When properly con- must be dealt with taking into account its specific
nected, however, these old systems can be circumstances and the degree of urgency in the re-
made to work effectively. Information on the quirements for upgrading. Actions to be taken may
structural properties of such systems is range from immediate evacuation of the building, to
lacking, making evaluation and upgrading dif- a phased repair program, to monitoring or further
ficult. This is especially important for heritage evaluation, or to acceptance of the building “as is.”
buildings.
• Despite their lack of compliance with some as- Basic Considerations
pects of current codes, many old buildings
have performed satisfactorily over the years 8. Structural requirements in Part 4 and refer-
without distress or failure. In addition, some enced structural design standards include general
structural parameters, such as dead load and performance requirements and design criteria.
material properties, can be ascertained by mea- These requirements are based on the following fun-
surement or test. Such information is not taken damental considerations:

Commentary K
• life safety Quality Assurance

• comfort of occupants
• function of the building for its intended use 13. The structural criteria contained in Part 4
• durability and referenced CSA structural design standards are
• economics. based on a level of quality assurance corresponding
to the requirements contained in Parts 2 and 4 of
9. The structural requirements of Part 4 and ref- the Code and in the referenced structural design
erenced CSA standards address life safety first and standards. The most important of these are NBC
foremost, but they also address comfort, function Sentence 4.1.1.2.(2), which requires that the designer
and economics. Life safety is addressed by criteria be a professional engineer or architect skilled in the
for the ultimate limit states (strength, stability, work concerned, and NBC Section 2.6., which
integrity). Comfort, function and economics are ad- requires that construction be reviewed for confor-
dressed by criteria for the serviceability limit states. mance to the design.
Economics are also taken into account by basing the
criteria on appropriate levels of structural reliability, 14. These quality control requirements also
thereby helping to avoid unnecessary consumption apply to structural evaluation and upgrading of ex-
of materials. isting buildings. The quality assurance may have to
be greater for the evaluation and upgrading of an
10. The basic considerations of safety and ser- existing building because the uncertainties concern-
viceability apply equally to existing or renovated ing structural properties of an existing building can
buildings and to new construction. However, other be considerably greater than for new construction.
basic considerations related to construction costs, More engineering judgment is generally required
user disruption and conservation (heritage value, for structural evaluation and upgrading of existing
reduction of waste and recycling) may be more criti- buildings than for design of new buildings. For
cal for existing buildings than for new construction. these reasons the following recommendations are
These other basic considerations usually result in a based on the prerequisite that:
requirement for minimised structural intervention • an appropriate structural evaluation of the
for the continued use or renovation of an existing building has been carried out, and the engi-
building. Therefore, where it can be shown that the neering evaluator has examined construction
resultant life safety (defined as an appropriately low details which are considered critical by the
probability of death or injury due to structural fail- evaluator,
ure) is generally equivalent to that required by the • field review by the designer will be carried out
National Building Code, and the building is known during any upgrading work.
to be functional, some departure from current code
design criteria may be appropriate. Recommended Code or Standard
11. Structural criteria in Part 4 and referenced 15. Recommendations on the code or standard
CSA structural design standards are based primarily which may be applied to the evaluation and upgrad-
on the limit states methodology [NBC 4.1.3.]. The ing of existing buildings are summarized in Table
working stress methodology [NBC 4.1.4.] is still K-1. Sometimes the standard used for the design of
used as an alternative for some design procedures the building may be preferred to a current standard;
(masonry, foundations). Criteria recommended in for example, some old buildings were made with
this Commentary are based on the limit states products no longer used, such as undeformed rein-
methodology. Appropriate adjustments will be forcement. Restrictions on the use of earlier versions
indicated for the working stress methodology, of standards are given in Notes 1 and 2 of Table K-1.
which was applicable at the time of the design of
most existing buildings.
16. Buildings designed and built in accordance
with previous codes may be considered acceptable
12. This Commentary addresses principally cri- provided:
teria for the ultimate limit states, because these limit • the previous code or standard essentially satis-
states directly affect life safety. Criteria for the ulti- fies the life-safety requirements of the current
mate limit states include loads, load factors and code or standard, and
load combinations specified in NBC Section 4.1., • the building or its use is not altered in such a
and resistances and resistance factors specified in way as to affect its structural behaviour or to
the CSA structural design standards. Serviceability increase the loadings on the structure.
and durability problems may also occur as a
consequence of renovation or change of use or envi-
ronment, and these are discussed.

Commentary K
Table K-1
Recommended Codes/Standards
1995 Code/Standard Commentary K Code/Standard when Built

Load Material Material
Loads Load Factors Loads
Factors Standards Standards

Evaluation
- no change in use or occupancy loads (1) (1)(2)
- change in use or occupancy loads X (2)
Design of Upgrading (3) (3) X X

acceptable X unacceptable
Notes to Table K-1:

(1) Acceptable provided the following conditions are met:
• no significant damage, distress or deterioration
• designed and built in accordance with recognized codes
• no changes that could impair the performance of the structure
• excludes seismic considerations
(2) Acceptable provided experience does not show serious deficiences in the Standard.
(3) NBC 1995 load factors are preferred (see Paragraph 24).
17. A benchmark version of a code or standard structural component was designed prior to the
is the earliest version which satisfies the life-safety benchmark version in Table K-2, then the current
intent of the current requirement. Use and occu- (1995) version should be applied using the load fac-
pancy loads, with one or two exceptions, are tors recommended in the 1995 NBC or in this
essentially unchanged over the years. On the other Commentary, or the evaluation may be based on
hand, earthquake requirements have changed con- satisfactory past performance under the conditions
siderably over the years and consequently buildings given in Paragraph 18. In the future, each structural
designed to earlier codes often do not provide a design standard referenced in NBC Section 4.3. will
level of life safety that meets the intent of current provide benchmark versions of the standard for
requirements. Table K-2 identifies benchmark ver- structural resistance. For further guidance see Para-
sions of NBC Section 4.1. for structural loads. If a graphs 54-56.
Table K-2
Benchmark Versions of NBC 4.1. (Structural Loads and Procedures)
Load Benchmark Year As Modified (with year of modification)

Use and occupancy 1941 guards (1975 and 1995(1))
interior walls over drops (1985)
Snow, ice, rain 1960 snow drifts (1965)
ground snow loads (1990)
large flat roofs (1995)
rain loads – blocked drains (1970)
Wind 1960 flexible structures and canopies (1970)
Earthquake 1970 seismic zones (1985)
Notes to Table K-2:

(1) The 1995 NBC guard loads, which are less stringent than those in the 1975 to 1990 editions of the NBC, should be used for evaluation of all
guards and their supports.

Commentary K
Evaluation Based on Satisfactory for life safety. The requirement for life safety, as
Past Performance distinct from structural safety, is based on an accept-
able maximum annual probability of death or
18. Buildings or components designed and built serious injury as a result of a structural failure in a
to earlier codes than the benchmark codes or stan- building. This probability is equal to the probability
dards, or designed and built in accordance with of structural failure (corresponding to a reliability
good construction practice when no codes applied, index of approximately 3 for buildings conforming
may be considered to have demonstrated satisfac- to Part 4) times the likelihood of death or serious in-
tory capacity to resist loads other than earthquake, jury if failure occurs. If the likelihood is high, there
provided: should be no relaxation in the load factors specified
• careful examination by a professional engineer in Sentence 4.1.3.2.(4). Where the likelihood is low,
does not expose any evidence of significant as in the case of storage buildings of low human
damage, distress or deterioration; occupancy, the load factors may be reduced. This is
• the structural system is reviewed, including recognised in Sentence 4.1.3.2.(7) by means of an
examinations of critical details and checking importance factor, . For post-disaster buildings, the
them for load transfer; loads and load factors of NBC Section 4.1. should be
• the building has demonstrated satisfactory applied.
performance for 30 years or more;
• there have been no changes within the past 30 21. Reduced load factors for structural evalua-
years that could significantly increase the tion, incorporating the principle of an importance
loads on the building or affect its durability, factor, are recommended in Table K-3. These factors
and no such changes are contemplated. are based on maintaining the level of life safety im-
plied by Part 4 by using the principle described in
19. If these conditions are not satisfied, the eval- Paragraph 20.(1) The load factors in Table K-3 are de-
uation should be based on the recommendations in termined by the evaluator based on consideration of
Paragraphs 20-37. three factors which affect life safety – the behaviour
of the structure (system behaviour), the likelihood
of people being at risk and their number (risk cate-
Load Factors and Load gory) and the evidence of safety indicated by past
Combinations Recommended for performance. The risk category is addressed in
Use in Evaluations [NBC 4.1.4. or Table K-5.
4.1.3.]
20. Criteria for the ultimate limit states should

be applied in conformity with the basic requirement
Table K-3
Load Factors for Structural Evaluation(1)
Load Factor Load Combination

Reliability Level(2)
Dead, D Variable,(3) L or W Earthquake,(4) E Factor,
5 1.25 (0.85)(5) 1.50 0.6 0.70
4 1.20 (0.88)(5) 1.40 0.6 0.70
3 1.15 (0.91)(5) 1.30 0.6 0.75
2 1.11 (0.93)(5) 1.20 0.6 0.75
1 or 0 1.08 (0.95)(5) 1.10 0.6 0.80
Notes to Table K-3:

(1) This table does not apply to post-disaster buildings.
(2) Reliability Level = sum of the 3 indices for system behaviour, risk category and past performance in Table K-4.
(3) A reduction in load factor may also be justified because of control of the load, as for example, liquid in storage tanks (see Commentary F, Para-
graph 18). This may be taken into account in the application of Table K-3, provided that the load factor is not less than the minimum in Table K-3.
(4) See Paragraph 38 and Reference (2) for more specific guidance on the load factor for earthquakes.
(5) The value in brackets applies when dead load resists failure.

Commentary K
Table K-4
Indices for the Calculation of Reliability Level
Items to be Considered for Reliability Level Index

System Behaviour
failure leads to collapse, likely to impact people 2

failure unlikely to lead to collapse, or unlikely to impact people 1
failure local only, very unlikely to impact people 0
Risk Category (See Table K-5)
high 2
medium 1(1)
low 0(1)
Past Performance
no record of satisfactory past performance 1
satisfactory past performance(2) or dead load measured(3) 0
Notes to Table K-4:

(1) Increase by 1 for loads in assembly areas or for wood structures.
(2) At least 20 years, no significant deterioration.
(3) Apply to dead load factor only.
Table K-5
Risk Category(1)
Category Description
High Schools and other occupancies where many people are likely to be exposed to risk associated with
the failure (N(2) = 100 or more), buildings of major heritage importance, or industrial or other facilities
with hazardous occupancies
Medium Other occupancies where fewer people are likely to be exposed to risk associated with the failure
(N(2) = 5 to 100)
Low Other occupancies where the floor area or adjacent outside area exposed to the failure is not likely
to be occupied by people and, when occupied, by a small number of people only (N(2) < 5)
Notes to Table K-5:

(1) This Table does not apply to post-disaster buildings.
(2) The estimated maximum number of people exposed to risk associated with the failure, N, may be estimated as follows:
N = Occupied area exposed to risk, in m2 1 occupancy density 1 duration factor
where
• for building occupants the occupancy density and duration factor may be estimated using Table K-6,
• duration factor = average weekly hours of human occupancy/100 ≤ 1.0, and
• for people outside adjacent to the building these parameters should be assessed approximately, using the same concepts as for
building occupants.

Commentary K
Table K-6
Parameters for Estimation of N
Primary Use Occupancy Density, Average Weekly Hours of

Persons per m2 Human Occupancy
Assembly 1.0 5 – 50
Mercantile and personal services 0.2 50 – 80

Offices, care or detention, manufacturing 0.1 50 – 60
Residential 0.05 100
Storage 0.01 to 0.02 100
22. The choice of the reduced load factor in 25. The combinations of loads other than earth-
Table K-3 is made by the evaluator for the specific quake to be used in the application of Table K-3
component addressed by the calculation. The evalu- should, in accordance with Sentence 4.1.3.2.(3), be:
ator must consider what will happen if the
component fails. Are there protective features of the
structural system (including non-structural compo-
nents) that, given structural failure, reduce the
likelihood of people (both outside and inside the
where the values of the load factors D, L , W
building) being injured or killed? Are many people
and are given in Table K-3, and D, L and W are
likely to be within the region affected by the failure?
defined in Sentence 4.1.2.1.(1). Loads due to move-
For example, failure of exterior building compo-
ments, T, have not been included for reasons given
nents (such as masonry parapets) overlooking exits
in Paragraphs 28 and 29. For load combinations in-
or busy streets are a greater risk than failure of com-
cluding earthquake, E in Table K-3 should be used
ponents overlooking rarely used areas. Those which
in conjunction with Sentence 4.1.3.2.(8).
fail during earthquake are generally a greater risk
than those which fail in very high winds, when 26. For the application of Table K-3 to working
fewer people are outside. Finally, if the building is stress criteria in Subsection 4.1.4., the specified loads
old and its past performance is satisfactory, this evi- can be adjusted by the ratio of the load factor deter-
dence of its safety can be taken into account, except mined from Table K-3 to that corresponding to
for seismic hazards. Reliability Level 5 in Table K-3. For example, if the
Reliability Level in Table K-3 is 2 for a floor girder,
23. Table K-5 provides guidance for determining the dead load is multiplied by 1.11/1.25 and the
the risk category used in Table K-3, including a pro- live load is multiplied by 1.2/1.5 before checking
cedure for estimating the number of people exposed the safety in accordance with Subsection 4.1.4.. An
to risk associated with the failure. In applying this upward adjustment of the load combination factor
procedure, the engineer should estimate the area of 0.75 in load combination (b) of Article 4.1.4.2.
the building which is likely to be affected by the should also be made in increments of 0.05 similar to
failure mode of the component being evaluated. For that in Table K-3.
example, a punching shear failure of a flat slab
building may be likely to cause a major total col- Loads Recommended for Use in
lapse, whereas a floor joist failure usually affects Evaluations
only a small area.
27. Loads specified in Part 4 concern primarily
the ultimate limit states and life safety, and there-
24. Minimum load factors in Table K-3, while fore relaxations are generally not recommended.
maintaining low risk to life safety, infer an increased Sometimes, however, as discussed in the following,
risk of building damage due to structural failure. it may be possible to determine loads for evaluation
They should be considered as a minimum to require more accurately than for design. Earthquake loads
upgrading. They may not be appropriate for use in are discussed in Paragraphs 38 to 42.
the design of the upgrading. Where the difference in
upgrading cost due to increasing the minimum load Loads Due to Movements, T [NBC
factor is small, and the loss due to failure is large, 4.1.2.1.(1) and 4.1.2.2.(2)]
higher load factors, such as those specified in
Sentence 4.1.3.2.(4), are recommended for structural 28. Loads due to movements caused by temper-
design of the upgrading. The level of upgrading ature change, moisture change and sustained stress
should be determined in consultation with the (e.g., shrinkage, creep, differential settlement) may
owner. usually be neglected for structural evaluation of an

Commentary K
existing building provided an inspection of compo- 35. For all other use and occupancy loads, it is
nents and connections indicates no damage recommended that Part 4 be followed.
affecting the safety of the building. This is because
past experience with the existing building will show Live Loads Due to Snow, Ice and Rain, L
whether such movements cause local damage or [NBC 4.1.7.]
displacements which may affect the strength or in-
36. It is generally difficult to justify a reduction
tegrity of the building. Ten years of experience is

usually sufficient except for differential settlements in snow, ice or rain loads from those specified in
of footings on materials such as clay, which can take Subsection 4.1.7. and recommended in Commentary
approximately 30 years. H. Despite apparent structural deficiencies accord-
ing to current Code requirements, however, many
29. For upgrading, consideration should be years of satisfactory roof performance may indicate
given to differential movements between new and a need to better assess actual snow loads on the
old materials. building. Special studies, including a comparison of
local records of ground snow accumulation at the
Dead Loads, D [NBC 4.1.5.] building site with those determined at the Atmo-
spheric Environment Service weather station, as
30. Where dead loads are determined from field well as special model or analytical studies of snow
measurements, the uncertainty of dead load is re- accumulation on the building in its location, may be
duced compared to design. Tables K-3 and K-4 take used to estimate more closely the site-specific snow
this into account by means of a reduction in the load. The assumptions of such studies may not
dead load factor. Similarly, Note 3 to Table K-3 al- apply, however, if there will be a change in roof ge-
lows a reduction in dead load factor where the load ometry or in wind exposure (e.g., due to new
is highly controlled, as for example hydrostatic load buildings). Also a change in snow or ice loads on an
in storage tanks. existing building can occur due to changes in insu-
31. Due to difficulty in the control of future in- lation or indoor heating, or it can occur due to snow
stallations of partitions in office buildings, it is sliding off a sloping roof as a result of a change in
recommended that 1kPa, as called for in Sentence roofing material. See Commentary H for further
4.1.5.1.(3), be maintained in those occupancies. guidance.
Live Loads Due to Use and Occupancy, L Live Loads Due to Wind, W [NBC 4.1.8.]
[NBC 4.1.6., NBC 4.1.10.]
37. It is equally difficult to justify a reduction in
32. Loads due to people, such as those for wind load from that specified in Subsection 4.1.8.
assembly, access and exit areas, have a direct effect and recommended in Commentary B. Despite calcu-
on life safety. Note 1 of Table K-4 therefore allows lated structural deficiencies in a building (according
less of a reduction in load factor for loads in such to current Code requirements), many years of satis-
areas than for all other loads. factory performance may indicate the need to better
assess actual wind loads on the building. Special
33. It may be possible in an existing building to studies, including measurements of wind speeds at
control some floor loads to a value less than that the building site (as compared to those measured at
specified in Subsection 4.1.6. If the analysis of the the Atmospheric Environment Services weather sta-
projected use of the floor clearly indicates that the tion), as well as model or analytical studies of wind
NBC load, including dynamic effects, will not be loads on the building in its location, may be used to
approached, then a reduction may be warranted, estimate the site-specific wind load more closely.
provided that any future change from the use con- The assumptions of such studies may not apply,
templated is controlled. For example, Article 4.1.6.6. however, if there is a future change in building
allows a reduction in specified loads for dining shape or local topography. See Commentary B for
areas from 4.8 to 2.4 kPa, provided the floor area is further guidance.
100 m2 or less and the floor will not be used for
other assembly uses, such as dancing. Generally, Earthquakes [NBC 4.1.9.]
however, future use is difficult to control and this
provision should be used with caution and only 38. Current earthquake requirements in Part 4
with the approval of the authority having jurisdic- and referenced structural design standards can
tion. present major difficulties for rehabilitation, particu-
larly for heritage or other buildings of unreinforced
34. The requirements of Sentence 4.1.10.5.(1) masonry.
concerning dynamic analysis of floors supporting
rhythmic activities need not be applied if past 39. Specification-type clauses which cause diffi-
experience indicates that vibration has not been dis- culties include restrictions on structural systems for
tinctly noticeable and that a change of use of the buildings more than 3 storeys high [NBC 4.1.9.3.(1)],
floor area is not contemplated. requirement for reinforcement in masonry walls

Commentary K
[NBC 4.1.9.3.(5)], requirements that certain portions floor structure should be evaluated for such a use
of the structure fail before others [NBC 4.1.9.3.(4) either by means of a performance test or by calcula-
and 4.1.9.4.(1)], restrictions related to lateral deflection procedures (see Commentary A for further
tions and pounding [NBC 4.1.9.2.(3) and (4)], as guidance). An evaluation is also recommended for
well as restrictions on detailing for earthquakes con- intended uses such as the installation of reciprocat-
tained in the referenced structural design standards. ing machinery or the use of equipment which is
sensitive to vibration, floor smoothness, slope, etc.

40. To help overcome these difficulties it is rec-
ommended that the NRC Guidelines for the Seismic 46. An alteration of building components affect-
Evaluation of Existing Buildings(2) be followed. In- ing the properties of the structure, and therefore its
formation on techniques for seismic upgrading is response to loading, might include the removal of
contained in Reference (3). Reference (4) contains a partitions, which reduces the damping and stiffness
method of screening buildings prior to detailed seis- of the floor system and increases its sensitivity to
mic evaluations and is essentially a management vibration induced by footfalls. In this case, it is rec-
tool for an owner or authority responsible for a ommended that the floor construction be reviewed
large building inventory. for the intended use before removing the partitions.
Similar alterations which may affect structural ser-
41. The reduced load factor for earthquakes of
0.6 in Table K-3 should be considered suitable as a viceability include alterations to cladding and
triggering criterion for seismic upgrading. For de- partitions in tall buildings, which affects wind sway
sign of the upgrading, the load factor should be motions, and the addition of heavy components,
increased, preferably to the NBC value, based on which results in increased deflection.
considerations of future building use, control of
seismic damage (to the building and contents) and 47. In the case of earthquakes, the deflection cri-
the differential in upgrading costs with earthquake teria of Sentence 4.1.9.2.(3) are intended to control
force level. An exception is the upgrading of unrein- damage of non-structural components. This will
forced masonry buildings covered by the special usually not have been tested by experience. For
procedure contained in Appendix A of Reference guidance see Reference (2).
(2), for which the criteria of Appendix A apply.
Durability
42. For many buildings in low to medium seis-
mic zones (Za or Zv of 3 or less), life safety can often 48. Durability is a major factor affecting service-
be greatly improved at relatively low cost by pro- ability and safety requirements which, although not
viding lateral support to masonry and other heavy addressed in the general requirements of Part 4, is
non-structural components. addressed in Section 4.2. and in the structural de-
sign standards referenced in Sections 4.3. and 4.4.
Serviceability (often by reference to other standards, such as CSA
A23.1). The CSA standard on parking garages refer-
43. Serviceability requirements in Part 4 enced in Subsection 4.4.2. is concerned essentially
(Articles 4.1.1.5., 4.1.1.6., 4.1.3.3. and much of with durability, as is a CSA standard (CSA S448.1-
Section 4.2.) and referenced structural standards 93) on repair of reinforced concrete in buildings.
concern human comfort and the function of the
building structure for its intended use (operation of 49. Corrosion failures of unbonded post-
equipment, drainage, protection function of the tensioned beams and slabs, reinforced concrete
building envelope, etc.). parking structures, supports and connections for
44. The serviceability criteria contained in Part 4 precast or other wall panels, masonry wall ties and
and referenced standards are intended for the design deep foundations may result in unsafe structures
of new buildings. For existing buildings, in many without visible deterioration. References (5) and (6)
cases demonstration of satisfactory performance provide guidance for assessment of such conditions.
eliminates the need to apply the serviceability crite-
ria given in Part 4 and referenced structural 50. Change of use (e.g., internal environmental
standards for structural evaluation. Unacceptable conditions) or alteration of the building components
deformation, settlement, vibration or local damage (e.g., insulation) may result in future deterioration
will usually be evident to the occupants within a pe- where none had occurred in the past, particularly to
riod of 10 to 30 years from construction. Examples exterior wall components. Such potential deteriora-
where serviceability evaluations may be required tion should be considered in the evaluation.
include change of use, or alteration of building com-
ponents affecting the properties of the structure. Structural Integrity
45. A change of use, for example, might include 51. In the structural evaluation of an existing
introduction of activities such as aerobics or jogging building, the ability of the structure to absorb local
into an existing building. In such cases, the existing failure without widespread collapse is an important

Commentary K
property, which should be considered by the engi- 58. Most load tests of existing building struc-
neering evaluator. This property can be assessed by tures consist of proof tests to establish safety.
considering the likelihood of specific failures due to Occasionally it may be useful to carry out destruc-
overloading, accidental damage, defects and deteri- tive ultimate tests of isolated structural components
oration and, if there is such a likelihood, the ability to determine their capacity and mode of failure.
of the building (both structural and non-structural Load tests can also be used to determine component
components) to provide alternative paths of sup- forces in a structure where it is difficult to apply a
port. This consideration, however, is not easily conventional structural analysis.
quantifiable and therefore involves considerable en-
gineering judgment. Tables K-3 and K-4 take 59. In some situations a load test may not pro-
alternative paths into account by means of a reduc- vide sufficient evidence concerning the future safety
tion in load factors based on a consideration of of the structure. An example is a post-tensioned
system behaviour. See also Commentary C. structure with very little normal reinforcement,
where there is hidden corrosion of prestressing. Al-
Foundations though such a structure may pass a load test, further
deterioration may result in a sudden brittle failure.
52. The adequacy of spread footings can gener-
ally be demonstrated by satisfactory performance in 60. It is important that in a load test, the struc-
the past. Consideration should, however, be given ture be exposed and accessible for visual inspection
to spread footings that will be subjected to a signifi- before, during and after the test.
cant increase in loading. Consideration should be
given to deep foundations in situations where they 61. For proof tests, the loads should be applied
may have been weakened by deterioration. to the structure in a pattern representative of the ex-
pected loading and to produce the maximum effects
53. Guidance concerning earthquake effects on for the critical modes of potential failure as ascer-
foundations is given in Reference (2). tained by the evaluator. The proof test loads should
be representative of the effect of factored loads spec-
Referenced Structural Design ified in Section 4.1., or some multiple thereof,
Standards depending on the type of failure (gradual versus
sudden) and whether the whole structure is tested
54. In the future each material design standard or only a representative portion. For concrete or
referenced in Section 4.3. will provide guidance on composite concrete and steel structures, the require-
its application to existing buildings. Until such ments of Chapter 20 of CAN/CSA A23.3-94 should
guidance is available, the evaluator is advised to be followed. In the case of non-composite steel
follow the ultimate limit state requirements for re- frame structures, an evaluation can normally be
sistance (including resistance factors) contained in done by measurement and calculations. For other
each standard referenced in Section 4.3. In the materials, a test load (including the weight of the
meantime, information contained in Reference (7) structure tested) representing 1.3 times the total
may be helpful. dead load of the renovated building plus 1.6 times
the live load should be applied for a minimum of 24
55. Alternatively, the building may be consid- hours. The test should include the measurement of
ered adequate on the basis of satisfactory past deflections and recovery after the load is removed.
performance, provided the conditions described in
Paragraph 18 are met.
62. In general, the structure is considered to
56. Paragraphs 57 to 63 also provide guidance pass the load test if there is no evidence of impend-
for determining resistance by means of load tests as ing failure during the test. In addition, there may be
an alternative to that determined by structural anal- an indication of serviceability problems under speci-
ysis. fied loads if there is excessive cracking or deflection
(short-term or long-term). This should be evaluated
considering past experience with the structure and
Load Testing the contemplated future change of use.
57. Load testing can be used for structural eval-
uation where safety is in doubt (due to lack of 63. For more guidance on load testing see Refer-
drawings or design information, deterioration, fire ence (9).
or possible inherent deficiencies). In some cases,
load testing may be used to monitor the effects of Further Guidance on Methods of
deterioration [see Reference (8) for guidance]. Load Structural Evaluation
testing is generally used in the structural evaluation
process as a last resort, because it is usually disrup- 64. Further guidance on methods of structural
tive and costly. evaluation is contained in References (10) and (11).

Commentary K
References
(1) D.E. Allen, Criteria for Structural Evaluation
and Upgrading of Existing Buildings. Can. J.
Civ. Eng., Vol. 17, No. 6, December 1991.
(2) D.E. Allen, J.H. Rainer and A.M. Jablonski,
Guidelines for the Seismic Evaluation of Ex-

isting Buildings. Institute for Research in
Construction, National Research Council,
Ottawa, 1992. NRCC 36941.
(3) NEHRP, Handbook of Techniques for the Seis-
mic Rehabilitation of Existing Buildings.
Federal Emergency Management Agency, Re-
port FEMA-172, Washington, D.C., June 1992.
(4) J.H. Rainer, D.E. Allen and A.M. Jablonski,
Manual for Screening of Buildings for Seismic
Investigation. Institute for Research in
Construction, National Research Council,
Ottawa, 1992. NRCC 36943.
(5) ASCE Standard ASCE11-90, Guideline for
Structural Condition Assessment of Existing
Buildings. Am. Soc. Civ. Eng., New York,
New York, 1991.
(6) N.R. Webster, Evaluation of Unbonded Post-
Tensioned Structures. Proc. of Second
Canadian Symposium on Cement and Con-
crete. (Editor Sidney Mindess). University of
British Columbia, Vancouver, 1991, pp. 230-
241.
(7) Existing Bridge Evaluation. Supplement No. 1
to CSA Standard CAN/CSA S6-88 (Design of
Highway Bridges). Canadian Standards Asso-
ciation, Rexdale, Ontario, 1990.
(8) CEB Bulletin 192. Diagnosis and Assessment
of Concrete Structures. Comité Euro-
International du Béton, Lausanne,
Switzerland, 1989.
(9) Load Testing of Structures and Structural
Components. Institution of Structural Engi-
neers, London, England, 1989.
(10) Evaluation, Maintenance and Upgrading of
Wood Structures: A Guide and Commentary.
Am. Soc. Civ. Eng., New York, New York,
1982.
(11) Structural Renovation of Traditional Build-
ings. Construction Industry Research and
Information Association, CIRIA Report III,
London, England, 1986.

Commentary L
Foundations
Introduction excavation are possible. The length of time the cut

will remain unsupported must also be considered.
1. This Commentary provides guidance, com-
patible with sound engineering practice, for the 5. Guidelines for treatment of open cuts in
design of Foundations and Temporary Excavations broad soil categories are included in Table L-1. The
in accordance with the provisions of Section 4.2. selection of stable slope angles for Categories C and
(Foundations) of the National Building Code of D requires that stability analyses be carried out. The
Canada 1995. Subsections 4.1.3. and 4.1.4. allow the selection of appropriate design shear strength
use of limit states design or working stress design parameters for such analyses requires a careful as-
for the design of buildings and their structural sessment of imposed shear stress levels, time effects,
components. The Commentary deals with both ap- soil directional properties and uniformity, and
proaches for the design of Shallow and Deep should be carried out by a person qualified in this
Foundations. The material herein is intended as a work. The influence of groundwater conditions
first approximation dealing with routine problems within the slope, or piezometric levels at or below
of foundation design and construction. Neither this the toe of the proposed slope, should also be inves-
material nor the papers or texts to which it refers tigated, as the resisting shear strength along a
should substitute for the experience and judgement potential failure surface may be greatly reduced by
of a person competent in dealing with the complexi- hydrostatic pressures.
ties of foundation practice.
2. The Commentary falls into three principal Supported Excavations
parts: Temporary Excavations, Shallow Foundations
and Deep Foundations. Design of shallow or deep 6. Temporary shoring support of vertical exca-
foundations may be carried out by limit states de- vation faces requires an assessment of a number of
sign. Limit states design of temporary excavations factors, including the length of time the excavation
has not yet been introduced and such excavations is to be supported, earth pressures, pressures from
are designed by the traditional allowable stress pro- frost action or corrosion from aggressive soil or
cedures. groundwater. The shoring wall elements may be
either open, permitting full drainage, or closed, pro-
3. The Commentary does not deal specifically viding an impermeable barrier, depending mainly
with the identification and classification of soils and on the soil permeability and groundwater condi-
rocks, with subsurface investigations, with swelling tions. Closed systems are designed for soil and full
and shrinking clay, with frost action as related to groundwater pressures, whereas hydrostatic pres-
foundations, with soil and hydrostatic pressures, sures are not included in open systems, where
nor with retaining walls; these topics are included seepage through the wall can take place.
in the “Canadian Foundation Engineering Manual,
1992.”(1) Design earth pressures
Temporary Excavations 7. For flexible and semi-flexible shoring walls
Unsupported Excavations commonly used for support of vertical faces of exca-
vations, and which may have a variety of support
4. The safety and stability of unsupported exca- conditions, no satisfactory general theoretical solu-
vations depends on the soil and groundwater tions for prediction of earth pressures are available.
conditions and on the depth and slope of the cut. In The design earth pressure must take into account
granular materials slope failure will generally be the method and sequence of construction and the
fairly shallow; in clays, deep rotational failures in- tolerable deformation limits of the sides or faces of
volving not only the sides but also the base of the the excavation.

Commentary L
Table L-1
Open Cut Excavation Guidelines(1)(2)
Category Typical Failure Time to

Soil Type Groundwater Remarks References
Mode Failure
A Free-draining, Below cut or Shallow Generally Rarely a problem if groundwater (2)

granular, non- controlled by surface or rapid under control and slope angle
plastic silts advance de- slope wedge does not exceed friction angle
watering of soil. Unsaturated temporary
steeper cuts rely on apparent
cohesion and may slough with

time; cuts steeper than 45 are
not recommended; vertical cuts
more than 1.2 m in depth
should never be used.
B As for Cut below Sloughing to Rapid Uniform fine soils may flow for As for
Category A groundwater flow considerable distances if Category A
pumping from within excavation
is attempted. Slopes are
controlled by hydraulic effects
and may range from 1/3 or less
to full value of friction angle.
C Non-sensitive Saturated(3) Rotational. Rapid or Analytical methods generally As for
clays. Plastic Plane of delayed reliable for prediction of stability Category A
and cohesive weakness or depending on in soft to firm clays.
silts composite per cent of
surface operational
soil shear
strength
mobilized
D Sensitive Saturated(3) Rotational. As per (C): Extreme caution required; once —
clays Retrogressive little advance initial failure is provoked,
slides and as warning retrogressive action may affect
per (C) wide area; reliability of
analytical prediction methods
generally poor.
Notes to Table L-1:

(1) Mixed soils such as glacial tills should be classified into Category A, B, or C, depending on grain size, plasticity and permeability, and treated ac-
cordingly.
(2) The stability of an open cut slope which is only marginally stable at the end of excavation may be adversely affected by such factors as the na-
ture and magnitude of crest loading, vibrations, rainfall, the length of time the cut remains open or disturbance of the soil in the vicinity of the toe
of the slope.
(3) Excavations through alternate layers of cohesive and granular soils or excavations terminated within a cohesive soil underlain by granular strata
require an investigation of groundwater conditions in each layer, and the factor of safety against excavation base heave or slope failure as a re-
sult of upward water pressure should be assessed.
8. The yield of one part of a flexible wall throws represented in triangular, trapezoidal or rectangular
pressure onto the more rigid parts. Hence, pressures form, and the applicable design earth pressure coef-
in the vicinity of supports are higher than in unsup- ficients will range between the active KA* case and
ported areas, and the loads on individual supports the earth pressure at rest KO,** depending on per-
vary, depending largely on the stiffness characteris- missible wall and soil movements.
tics of the supports themselves and the construction
technique. * K = (1 – sin′)/(1 + sin′), where ’ = effective friction
A
angle of soil and the ground surface is horizontal.
9. The pressure envelopes representing the pres- ** K = 1 – sin ’
O
sures that would normally be anticipated can be

Commentary L
10. Non-cohesive (granular) soils. As a first (12 kPa < Cu < 25 kPa) to firm (25 kPa < Cu < 50
approximation the guidelines in Table L-2 are sug- kPa) clays, reference should be made to Table L-4.
gested in essentially granular soil such as fills,
sands, silts, sandy silts, gravelly sands, sands and Movements Associated with Excavations
gravels or alternate layered conditions composed of
such strata. 13. Movements associated with excavations are
primarily related to construction technique and

11. Cohesive soils. For cohesive soils a distinc- commonly consist of lateral yield of the soil and
tion must be made between soft to firm clays and support system towards the excavation, with
stiff to very stiff clays. The effects of clay sensitivity corresponding vertical movement adjacent to the ex-
and the factor of safety against base heave must cavation walls. Both lateral and vertical movements
also be accounted for. due to yield are generally of the same order of
magnitude; however, if very flexible vertical wall el-
12. For stiff clay soils (Cu > 50 kPa)*** including ements are used, lateral movements can be grossly
silty clays, sandy clays and clayey silts, the guide- increased. Where construction technique is poor, er-
lines in Table L-3 are suggested. Similarly, for soft ratic movements can also occur due to loss of
ground or erosion behind the wall.
*** Cu = 1/2 unconfined compressive strength.
Table L-2
Envelope of Earth Pressure for Design of Temporary Supports for Granular Soils
Design Total Envelope of Pressure

Restraint Ability to Restrict Adjacent Soil Movements(3)
Pressure(1) Distribution(2)
Cantilever 1.0 PA Triangular Generally very poor unless wall extremely stiff and
embedded in dense soil
Braced 1.2 to 1.3 PA Rectangular or Generally poor where control of groundwater
trapezoidal inadequate or where workmanship poor; can be
moderate to good where these factors are properly
controlled and bracing properly designed and tightly
wedged or preloaded
Tied-back 1.1 to 1.4 PA Rectangular or Generally good where high total pressures are
trapezoidal used; movements usually less than for braced walls
and dependent on degree of prestressing,
workmanship and wall stiffness
Notes to Table L-2:

(1) PA = theoretical total active pressure = 0.5gH2 2 KA
where
= density of soil (submerged if below groundwater), kg/m3,
H = depth of cut, and
g = acceleration due to gravity, m/s2.
The figure of 0.2 is suggested as a lower bound for KA even in dense soils. Surcharge pressures and hydrostatic water pressures should be
added where appropriate.
(2) After increasing PA by the appropriate multiplier, distribute total pressure over depth of cut as indicated in this column: triangular limits of trape-
zoid generally taken as 0.2H to 0.25H at top and bottom.
(3) Where greater control of adjacent ground movements is required, earth pressure should be computed using the at rest Ko earth pressure coeffi-
cient with prestress in struts or tie-backs to the full design load. Additional measures would include choice of a stiff wall and close vertical
spacing of struts or tie-backs.

Commentary L
Table L-3
Envelope of Earth Pressure for Design of Temporary Supports for Stiff Cohesive Soils
Restraint Envelope of Pressure

Design Total Pressure Ability to Restrict Adjacent Soil Movements
Distribution(1)
Cantilever 1.0 PA but not less than Triangular May be poor depending on length of cantilever, wall
0.15gH2 (2) stiffness, embedment conditions and clay
sensitivity(3)(4)
Braced or tied back 0.15gH2 to 0.4gH2 (5) Rectangular or Depends on soil strength, sensitivity, effective
trapezoidal preloading or prestressing and wall stiffness
Notes to Table L-3:

(1) Surcharge pressures should be added where appropriate; hydrostatic pressures need not be included; total density of soil, , is to be used in
calculations.
(2) PA may be computed using short term strength, i.e., PA = gH – 2Cu if excavation open for limited period. Regardless of whether pressures are
negative or zero, minimum positive pressures indicated must be used.
(3) Computed passive pressures below the base of the excavation should be reduced by 50% to account for unavoidable disturbance due to strain
effects and stress release.
(4) The factor of safety against base heave in stiff overconsolidated clays, as a result of high locked-in lateral stresses, should also be investigated.
(5) Use higher range where clay is of high sensitivity. If the construction sequence or workmanship allow significant inward movement during any
stage of excavation, pressures may build up to essentially fluid soil values in very sensitive clays. With good workmanship, clay pressures are
similar to those given in Table L-2. Strength tests taken on intact samples of stiff clays which are jointed or fissured may over-estimate the
strength characteristics and thus lead to an under-estimation of earth pressures.
Table L-4
Envelope of Earth Pressure for Design of Temporary Supports for Soft to Firm Clays
Restraint Envelope of Pressure

Design Total Pressure Ability to Restrict Adjacent Soil Movements(2)(3)
Distribution(1)
Cantilever 1.0 PA but not less than Triangular Very poor; this type of support generally to be
0.15gH2 (4) avoided in soft, sensitive clays
Braced or tied back 0.4gH2 to 0.8gH2 (5) Rectangular Depends on clay shear strength and stability(6)
Notes to Table L-4:

(1) Essentially fluid soil pressures in very sensitive clays may be realized as a result of unavoidable wall movements prior to insertion of restraint
supports.
(2) Computed passive pressures below the base of the excavation should be reduced by at least 50% to account for unavoidable disturbance due to
strain effects.
(3) Additional precautions in soft to firm sensitive clays would include (a) insertion of the top strut or anchor prior to excavation beyond 1.5 to 3 m
depth, and (b) where the excavation area is of limited size, placing of a 150- to 300-mm thick concrete mat at the base of the excavation, where
practical, immediately on completion of excavation.
(4) PA may be computed using short term strength, i.e., PA = gH – 2Cu if the excavation is open for a limited period. Regardless of whether pres-
sures are negative or zero, minimum positive pressures indicated must be used.
(5) Higher range should be used where clay is of soft consistency, and lower range where clay is of firm consistency. This value may be conserva-
tive for non-homogeneous, non-sensitive sandy-silty cohesive soils of firm consistency. If stability number N = (gH + surcharge)/Cu approaches
5 to 6, use the higher range. At this depth base heave may also take place and therefore suitable precautions should be taken.
(6) Design of a suitable shoring and bracing system in soft to firm clay conditions is not a routine matter, and the advice of a specialist should be ob-
tained to establish design pressures, to check overall stability and base heave and to predict adjacent soil movements.

Commentary L
14. Movements due to yield of cantilever walls 16. The yield movements of anchored walls are
are related to the wall and soil stiffness. For most controlled by design methods more than is the case
flexible or relatively flexible wall types the lateral with strutted walls. The number of anchors and the
deformations will exceed values required for mobi- vertical spacing of such anchors, play a significant
lization of active soil pressures. For most soils and part in controlling the degree of lateral deformation.
particularly cohesive soils, therefore, there is a dan- In normal practice, movements due to the yield of
ger that a further build-up of lateral pressures anchored diaphragms, sheeted or soldier pile walls
beyond active values will take place as a result of are usually less than for strutted walls for the same
loosening due to strain effects. An exception would depth of excavation.
be where design soil pressures of an at rest
magnitude or greater are used in design, and an 17. For general guidance Table L-5 summarizes
appropriately stiff wall, such as large diameter cylin- the approximate range of vertical and lateral move-
der piling, is provided, embedded in competent soil. ments to be expected. In certain cases, more
favourable results may be achieved with proper de-
15. Movements due to yield in strutted excava- sign, good construction workmanship and careful
tions are, to a large extent, unavoidable, since they field supervision, including monitoring the be-
are controlled not by design assumptions but by haviour of the excavation.
construction details and procedures. Such move-
ments develop in each excavation phase before the
next level of struts is installed.
Table L-5
Vertical and Lateral Movements Associated with Excavation(1)(2)
Restraint(3) Granular Stiff Clay, % Soft to Firm

Wall Details Remarks
Soils, % depth depth Clay,(4) % depth
Cantilever Conventional Moderate to Moderate May collapse Movements related to wall, soil
stiffness large stiffness and embedment condition
Braced Soldier piles or 0.2 to 0.5 0.1 to 0.6(5) 1 to 2(5) Struts installed as soon as support
sheet piles level reached and prestressed to
100% design load
Rakers or struts 0.5 to 1.0 0.3 to 0.8 >2 Poor workmanship would result in
loosely wedged greater values
Tied-back Soldier piles or 0.2 to 0.4 0.1 to 0.5 1 to 2 Prestressed to pressure between
sheet piles active and at-rest
Concrete < 0.2 < 0.1 to 0.5 < 1 to 2 Prestressed as above, since wall
diaphragm walls stiffness and design earth
pressures are normally greater,
movements generally are less than
for soldier piles or sheet piling; little
data available
Notes to Table L-5:

(1) Movements indicated apply directly behind wall; for granular soils and stiff clays, movements would feather out in approximately linear fashion
over horizontal distance of 1.0 to 1.5 depth of excavation (H). For soft to firm clays, and assuming average workmanship, this distance increases
to 2.0 to 2.5H, and with poor workmanship to greater than 3H.
(2) If groundwater is not properly controlled in granular strata, movements may be much larger than indicated, and loss of ground could also result.
(3) Experience indicates that movements are reduced by using close vertical spacing between strut or tie-back levels and by careful attention to pre-
stress details.
(4) If the factor of safety against base heave for soft to firm clays is low, large deformations will result.
(5) Upper range of movements usually applies for highly sensitive clays in either stiff or soft to firm category.

Commentary L
Underpinning should be considered in design, together with the

weight of the vertical member itself.
18. Structures adjacent to excavations frequently
need to be supported. The need for underpinning 22. Forces which resist downward movement
depends on the location of the structure, the details due to the inclined anchor load are skin friction and
of its foundation support, its sensitivity to settlement the reaction at the base of the vertical member.
and lateral deformations, the cost of underpinning When soldier piles are used, vertical forces are con-
or provision of extra excavation face support and centrated in the piles. Only minimal friction, if any,
other precautions, and the cost of repairs or the con- can be mobilized. Such vertical forces must, there-
sequences if the structure is not underpinned. fore, be supported at the base of the pile. The
vertical and horizontal base capacity of the pile
19. The geometry of zones within which sup- must be checked; otherwise, unacceptable vertical
port for adjacent structures is usually considered and horizontal deformation may take place.
necessary, as a result of adjacent excavation through
soil, is shown in Figure L-1. Where adjacent struc- 23. Settlement of vertical members produces
tures are founded on bedrock and excavation is some reduction in anchor loads, with a consequent
through rock, less underpinning and more face sup- tendency for outward displacement of the sup-
port should be considered. ported face. Therefore, vertical and horizontal
movements at the top and bottom of the excavation
must be monitored at regular intervals throughout
the course of the work.
1 2
1 1 tightly braced/tied 24. The performance of soil and rock anchors
A excavation wall
depends not only on minor variations in soil and
B
groundwater conditions but also on construction
C techniques and details. Consequently, the prediction
base of excavation of anchor capacity by theoretical calculations is not
reliable. Anchorage capacities must be established
by test, taking into account the load deformation
600 mm and “creep” properties of the soil, and each anchor
must be proofloaded during construction.
25. The overall stability of a soil anchorage sys-

tem should be checked by analyzing the stability of
Zone A
Foundations within this zone often require underpinning. Horizontal the block of soil lying between the wall and the
and vertical pressures on excavation wall of non-underpinned anchorages. In general, the anchors should be ex-
foundations must be considered. tended beyond a 1:1 line drawn from the base of the
excavation, and no allowance for any load carrying
Zone B
Foundations within this zone often do not require underpinning. support should be assumed within this line.
Horizontal and vertical pressures on excavation wall of non-under-
pinned foundations must be considered. Design and Installation of Members
Zone C 26. Members such as walers, struts, soldier piles
Foundations within this zone usually do not require underpinning.
and sheeting should be sized in accordance with the
structural requirements of Part 4 of the National
Building Code of Canada 1995.
Figure L-1
27. The depth of penetration of the vertical wall
Requirements for underpinning member should be 1.5 times the depth required for
moment equilibrium about the lowest strut.
20. The general order of magnitude of move-
ments as a result of excavation with various support 28. For driven soldier piles, the maximum hori-
methods in different soil conditions has been sum- zontal force on the flange of the soldier pile below
marized in Table L-5. This table may also be used to the bottom of the excavation may be taken as 1.5
assist in judging the necessity for underpinning. times the values computed for the width of the
flange, providing that the pile spacing is not less
Factors to be Considered with Soil and than five times the flange width.
Rock Tie-Back Anchors
29. For piles placed in a concrete base, the di-
21. Anchors are usually inclined downwards, ameter of the concrete filled hole may be used in
transmitting the vertical component of the anchor place of the flange width as discussed in the preced-
force into the anchored vertical member. This force ing paragraph.

Commentary L
30. The selection of material and sizes of timber (d) Adequate pumping and standby pumping ca-
planks or lagging should conform with good prac- pacity must be provided.
tice, and the lagging should be of good quality (e) Pumped water must be discharged in a man-
hardwood. Lagging is installed by hand after a ner that will not interfere with the excavation
depth of about a metre is excavated. The maximum or cause pollution.
depth made each time before a section of lagging is (f) For most soils the groundwater table during
placed depends on the soil characteristics. Soft clay construction must be maintained at least 600 to
and cohesionless soils must be planked in short 1 500 mm below the bottom of the excavation
depths to reduce the amount of soil moving into the so as to ensure dry working conditions. It
excavation. The depth of excavation below any lag- should be maintained at a somewhat lower
ging boards that have not yet been placed should level for silts than for sands in order to pre-
not exceed 1.2 m. Lagging must be tightly backfilled vent traffic from pumping water to the surface
or wedged against the soil. and making the bottom of the excavation wet
or “spongy.”
31. To minimize the possibility of erratic loss of (g) Adequate monitoring of groundwater levels by
ground in local areas when excavating sands and piezometers or by observation standpipes
silts below original groundwater, straw packing, should be maintained.
burlap or in extreme conditions, grouting must be (h) Where impermeable strata are underlain by
used behind the lagging as it is installed. pervious water-bearing layers, depending on
the depth of excavation and the hydrostatic
32. The design of all members including struts, head in the pervious strata, it may be neces-
walers, sheetpiling, walls and soldier piles should sary to lower the head in the pervious stratum
be checked for several stages of partial excavation in advance of excavation, to prevent a “blow”
when the wall is assumed to be continuous over the or excessive disturbance of the base as a result
strut immediately above the excavation level and of upward hydrostatic pressure.
supported some distance below the excavation level (i) Pumping from sumps or ditches inside the ex-
by the available passive resistance. This condition cavation is normally carried out where dense
could produce the maximum loading in struts and low permeability soils, such as certain glacial
walers. tills or cohesive soils, are present or where the
excavation is in bedrock; this method is not
33. Where excessive stresses or loads would re- recommended for excavation in semi-pervious
sult from interim construction conditions using or pervious soils, such as silts or fine sands,
regular construction procedures, trenching tech- since it often leads to extensive sloughing of
niques can be employed to advantage. the excavation sides and disturbance of the
bottom.
34. The design of members should also be
checked for the condition when portions of the Shallow Foundations
building within the excavated area are completed
and lower struts are removed. Consideration must General
be given to the possible increase in loading on the
upper struts remaining in place; also the span be- 36. A shallow foundation means a foundation
tween that portion of the building that has been unit which derives its support from the soil or rock
completed and the lowest strut then in place must close to the lowest part of the building which it
be considered in relation to flexural stresses. supports. The depth of the bearing area below the
adjacent ground is usually governed by the require-
Control of Groundwater in Excavations ment to provide adequate protection against
climatic or frost effects; vertical loads on the sides of
35. Good practice requires that the following the foundation due to adhesion or friction are nor-
conditions must be fulfilled when dewatering exca- mally neglected.
vations:
(a) A dewatering method must be chosen that will Bearing Capacity and Settlement
not only assure the stability of the sides and
bottom of the excavation but will also prevent 37. The design of a foundation unit normally re-
damage to adjacent structures, such as by set- quires that both bearing capacity and settlement be
tlement. checked. While either bearing capacity or settlement
(b) The lowered water table must be kept con- criteria may provide the limiting condition, settle-
stantly under full control, and fluctuations ment normally governs. Distress from differential
liable to cause instability of the excavation settlement, as evidenced by cracking and distortion
must be avoided. of doors and window frames, is common. The dras-
(c) Effective filters must be provided where neces- tic effects of a bearing capacity failure are rare,
sary to prevent loss of ground. except perhaps during construction, where shallow

Commentary L
temporary footings are frequently used with false- Allowable bearing pressure means the maximum
work. pressure that may be safely applied to a soil or rock
by the foundation unit considered in design under
38. The bearing capacity of both cohesive and expected loading and subsurface conditions.
non-cohesive soils can be determined with reason-
able reliability by assuming that the strength Allowable load means the maximum load that
parameters for the bearing soil are accurately may be safely applied to a foundation unit consid-
known within the depth of influence of the footing. ered in design under expected loading and
subsurface conditions.
39. Cohesive soil. The settlement of a structure
on cohesive soil can be calculated with less accuracy Design bearing pressure means the pressure ap-
than the bearing capacity. Such a calculation is af- plied by a foundation unit to a soil or rock and
fected by a number of complicating factors usually which is not more than the allowable bearing pres-
requiring judgement to assess. The most important sure.
of these is an estimate of the preconsolidation pres-
sure, that is, the maximum past consolidation Design load means the load applied to a founda-
pressure on the in situ soil. Because of the various tion unit and which is not greater than the
uncertainties, errors of a factor of 2 should be ex- allowable load.
pected in the calculation of settlement.
Design Bearing Pressure
40. Non-cohesive soil. The settlement of a
structure on non-cohesive soil can normally only be 45. The design bearing pressure is limited by
estimated by empirical methods. Such an estimate is two considerations:
usually taken to mean the settlement directly (a) the foundation must be safe against shear fail-
related to the load, but this settlement generally oc- ures of the supporting soil, and
curs quite rapidly, often during the construction (b) settlement must not be excessive.
period. Post construction settlement in such a case
will be negligible and may be considerably less than 46. The design bearing pressure is the lesser of
predicted. the values dictated by these two requirements.
41. Post construction settlement can occur for a 47. A detailed flow diagram for the design of
considerable period after construction, even after a shallow foundations is shown in Figure L-2. In
period of successful performance of the structure, as many cases this can be simplified; however, it illus-
the result of vibrations or changes in the groundwa- trates the factors affecting the choice of design
ter conditions, whether natural or man-made, due bearing pressure for most structures
to earthquake or blasting, flooding or groundwater
lowering. Settlement of this nature is not usually in- Estimates of Allowable Bearing Pressure
cluded in an empirical estimation, but should
be assessed. 48. Universally applicable values of allowable
bearing pressure cannot be given. Many factors af-
Bases for Design of Shallow Foundations fect bearing capacity, and the allowable load will
frequently be controlled by settlement criteria. Nev-
42. Either working stress design, described in ertheless, allowable bearing pressure for the
Paragraphs 43 and 44, or limit states design, de- preliminary design can usefully be estimated on the
scribed in Paragraphs 50 to 70, may be used to basis of the material description; such values should
design shallow footings. Guidance is only given for be recalculated for the final design in keeping with
footings supporting vertical loads. good geotechnical practice and normal analytical
procedures.(2)
Working Stress Design
49. Estimated values of presumed allowable
43. The traditional design method for shallow bearing pressure and notes are given in Tables L-6,
foundations is referred to here as working stress de- L-7, L-8 and L-9. Such pressures should be consid-
sign. Design is carried out using unfactored loads ered as the maximum permissible under the total
and design bearing pressures. The design bearing dead and live loading and treated as first approxi-
pressures represent the pressure required to cause mations only.
bearing failure divided by a factor of safety in the
order of 3.
44. The following terms should be used for ex-

pressing recommended geotechnical criteria for
working stress design of foundations:

Commentary L
Flow diagram for foundation design

Shallow foundations
Assemble information Assemble information

regarding proposed regarding site for prop.
structure: structure: geology,
type, function topography, climatic

factors
Formulate specification of
required foundation
performance
Field investigation
Examine number of – boreholes
possible foundation – trial pits
configurations and make – in situ tests
tentative economic – groundwater
evaluation of each conditions
Laboratory
investigations
soil
rock } properties
Selection of foundation
type
Proposed arrangement *
of footings in plan Factors affecting depth
of footing: frost
protection, slope stability,
Decide depths for each erosion, topography,
footing soil conditions, water
level, swelling(?)
Geotechnical / structural design of footing
Select tentative Calculate applied loads

dimensions for on each footing
footings
Check if total settlement yes

yes Check if bearing capacity
for each footing is
of soil directly underlying
excessive
footing is exceeded at
fs = 3 or φ = 0.5 no
no
yes Check if soil strata at Check if differential yes
greater depth is over- settlement between
stressed footings is excessive
no no
Check stability against no

horizontal forces
o.k.
Check uplift on each no
footing
o.k.
Proposed design and Construction methods
predicted performance and possible difficulties
– dewatering yes
– drainage
Construction
poor Modifications
– observe rock/soil
Check with predicted to original
conditions
performance design
– monitor behavior
required?
o.k.
no
Construction complete
*These factors frequently control foundation design
Figure L-2
Flow diagram for design of shallow foundations
Commentary L
Table L-6
Estimates of Allowable Bearing Pressure on Rock
Allowable
Rock Type Bearing
Rock Conditions(1) Remarks
Pressure, (2)
MPa
(a) Massive igneous and Discontinuities (joints, minor cracks) at 10 —
metamorphic rocks in sound wide spacing (> 1 m)
condition; granite, diorite,
basalt and gneiss Discontinuities at moderate spacing (300 2 to 5
mm to 1 m)
(b) Foliated metamorphic rocks (i) Discontinuities at wide spacing (> 1 m) 3 Foliations approximately horizontal
in sound condition: slate and
schist (ii) Discontinuities at moderate spacing <1 Foliations approximately horizontal
(300 mm to 1 m)
(iii) Foliations tilted to the horizontal — Potential sliding along foliations.
Potential lack of support adjacent to
cuts on excavations. See
Reference (3).
(c) Sedimentary rocks in sound Discontinuities at wide spacing (> 1 m) 1 to 4 Strata approximately horizontal
condition: cemented shale
or siltstone, sandstone, — Potential solution cavities in
limestone, dolomite and limestone, dolomite. Variability in
heavily cemented cementation of conglomerates. See
conglomerate (b)(iii)
(d) Compaction shale and other Discontinuities at wide spacing (> 1 m) 0.5 to 1 Strata approximately horizontal
argillaceous rocks in sound
condition — Argillaceous shales are subject to
some swell on release of stress. All
shales tend to soften on exposure
to water and certain shales swell
markedly
(e) All closely jointed rocks Discontinuities at spacing less than 300 — Can only be assessed by detailed
including thinly bedded mm apart. Random joint or crack investigations and examination in
limestones and shales patterns situ, including loading tests if
necessary
(f) Heavily shattered or — — See (e)
weathered rocks
Notes to Table L-6:

(1) Spacing of discontinuities is critical to the bearing pressure allowable on a rock mass. Discontinuities, such as joints or cracks, are considered
widely spaced if greater than 1 m apart and moderately spaced when greater than 300 mm. The thickness or width of such discontinuities is pre-
sumed to be less than 5 mm (or less than 25 mm if completely filled with soil or rock debris). Where such conditions do not exist, Types (e) or (f)
must be assumed.
(2) Values of bearing pressures given above, except for (f), are based on the assumptions that the foundations are close to the rock surface but car-
ried down to unweathered rock with adequate frost protection and that the foundation is greater than 300 mm wide.

Commentary L
Table L-7
Estimates of Allowable Bearing Pressure on Non-Cohesive Granular Soils
Soil Type and Conditions(1) Allowable Bearing

Potential Problems(3) Remarks
Pressure, (2) kPa
(a) Dense well-graded sands, 400 to 600 Density of sands containing large sizes or For general reference
dense sand and gravel gravels is frequently overestimated when see References (1)
inferred from standard or cone penetration and (5)
(b) Compact well-graded sands, 200 to 400 tests only, see Reference (4).
compact sand and gravel
(c) Loose well-graded sand, 100 to 200 Potential settlement when subject to shock
loose sand and gravel or vibrations. See (f)
(d) Dense uniform sands 300 to 400 Density usually better defined by standard See References (6) –
or cone penetration tests, as compared to (8)
(e) Compact uniform sands 100 to 300 (a) to (c). Considerable caution required in
interpretation of test data
(f) Loose uniform sands < 100 Even where very low bearing pressures are See Reference (9)
used, settlement can occur due to
submergence, vibrations from blasting
machine operation or earthquake
(g) Very loose uniform sands, — Subject to possible liquefaction. Should
silts never be used for support of foundations
Notes to Table L-7:

(1) Density condition of the soil is assumed to be established in conformance with good geotechnical practice.
(2) Values of bearing pressure are based on the assumptions that the foundation (B) is not less than 1 m wide and that the groundwater level will
never be higher than a depth, B, below the base of the foundation. When the groundwater level is, or could be, higher than such depth, the val-
ues listed should be divided by a factor of 2.
(3) Long term settlement of foundations on compact to dense non-cohesive deposits is normally modest, provided such deposits are not underlain
by compressible cohesive deposits at depth.
Table L-8
Estimates of Allowable Bearing Pressure on Cohesive Soils (for sensitive clays, see Table L-9)
Allowable
Applicability for Support of Shallow
Soil Type and Conditions(1) Bearing Settlement(2)
Foundations(2)
Pressure, (2) kPa
(a) Very stiff to hard clay, 300 to 600 Good Normally estimated on the basis of
heterogeneous clayey deposits investigations, sampling and
or mixed deposits such as till laboratory test data
(b) Stiff clays 100 to 200 Fair to good
(c) Firm clays 50 to 100 Poor — except for minor structures For general reference see
little affected by distortion References (1) and (10) – (12)
(d) Soft clays 0 to 50 Very poor — not recommended
(e) Very soft clays — Not permitted
Notes to Table L-8:

(1) Strength of cohesive soils is assumed to be established in conformance with good geotechnical practice.
(2) Cohesive soils are susceptible to long term consolidation settlement. For Types (b) to (d) inclusive, such settlement often governs the applied
pressure rather than bearing pressures based on soil strength. In the case of Type (a) soils, heave can take place with excavation and conse-
quent relief of stress.

Commentary L
Table L-9
Problem Soils, Rocks or Conditions(1)
Type or Condition Examples References

Organic soils Muskeg terrain: estuarine organic silts and clays (13)
Normally consolidated clays Lacustrine deposits and varved glacio-lacustrine deposits in (14)
Manitoba, Northern Ontario, Northern Quebec
Sensitive clays Marine clay deposits in St. Lawrence River Valley, Eastern (15) – (17)
Ontario, Quebec
Swelling/shrinking clays Clay-rich deposits in Alberta, Saskatchewan, Manitoba (18)
Metastable soils British Columbia loess (19)
Expansive shales Western Canada – Bearpaw and Cretaceous deposits (20)(21)
Eastern Canada – weathering of sulphide minerals
accelerated by oxidizing bacteria
Permafrost Northern Canada, Arctic (22)(23)
Notes to Table L-9:

(1) No bearing pressure can be presumed without detailed investigations.
Limit States Design 53. The following limit states terms should be
used for expressing recommended geotechnical
50. Limit states refer to those conditions of a criteria for the design of the building structure, in-
structure in which the structure ceases to fulfil the cluding its foundations:
function for which it was designed. The limit states
are classified into two main groups: Bearing pressure for settlement means the
• Ultimate limit states bearing pressure beyond which the specified ser-
• Serviceability limit states. viceability criteria are no longer satisfied.
51. Ultimate limit states are primarily concerned Factored bearing resistance means the calculated
with collapse mechanisms for the structure and, bearing resistance, obtained using characteristic soil
hence, safety. For foundation design, ultimate limit parameters, multiplied by the recommended resis-
states consist of: tance factor.
• exceeding the load carrying capacity of the
foundation (i.e., ultimate bearing capacity);
• sliding; Factored sliding resistance means the calculated
• uplift; sliding resistance, obtained using characteristic soil
• large deformation of foundation, leading to an parameters, multiplied by the recommended resis-
ultimate limit state being induced in super- tance factor.
structure or building;
• overturning; Factored pull out resistance (i.e., against uplift)
• loss of overall stability. means the calculated pull out resistance, obtained
using characteristic soil parameters, multiplied by
52. Serviceability limit states consider mecha- the recommended resistance factor.
nisms that restrict or constrain the intended use or
occupancy of the structure. They are usually associ- 54. In limit states design, the relevant limit
ated with movements that interrupt or hinder the states are identified and through the design process
purpose (i.e., serviceability) of the structure. For shown in Figure L-2, it is verified that no limit state
foundation design, serviceability limit states can be is exceeded. The design process may be simplified
categorized as: in many cases as experience will often show which
• excessive movements (e.g., settlement, differ- type of limit state will govern the design and the
ential settlement, heave, lateral movement, and other limit states are checked to ensure that they are
tilt or rotation); not exceeded.
• unacceptable vibrations.

Commentary L
55. The basic design equation for limit states de- mean to nominal of 1.1) typical for shallow and
sign is: deep foundations. The advantage of Figure L-3 is
that can be readily interpreted by geotechnical en-
gineers who have considerable experience in using
the traditional values of global safety factor. This
where 'Rn is referred to as the factored resistance. can assist in bridging the gap, during the transi-
The resistance factor, ', accounts for variability in tional stage, between the use of working stress and
the soil strengths and as-built dimensions, and limit states design.
variabilities introduced by inaccuracies in the calcu-
lation model. It also indirectly allows for ductile
and catastrophic failures. The nominal resistance, Table L-10
Rn, is the engineer’s best estimate of the resistance Resistance Factors for Shallow and Deep Foundations
of the foundation. The value of Rn should allow, at
least partially, for variabilities resulting from
geotechnical uncertainties. It is based on the Description Resistance
nominal strengths of the soil, nominal (specified) di- Factors
mensions and the normal calculation model. 1. Shallow foundation
56. The term Sni is the nominal value of the (a) Vertical resistance by semi-empirical 0.5
forces on the foundation resulting from the ith load. analysis using laboratory and in-situ
These forces are obtained from the specified loads test data
by structural analysis. The term i is the load factor
for the ith load. It accounts for the variability in the (b) Sliding
load itself, approximations in the loading model
given in the code and variability introduced by the (i) based on friction (c = 0) 0.8
structural analysis. (ii) based on cohesion/adhesion (tan 0.6
57. The load factors and load combinations are f = 0)
as given in Subsection 4.1.3. of the National 2. Deep foundation
Building Code.
(a) Bearing resistance to axial load
58. The recommended resistance factors are
given in Table L-10. The resistance factors in Table (i) semi-empirical analysis using 0.4
L-10 have been derived by direct calibration to cur- laboratory and in-situ test data
rent working stress design. This means that the
dimensions of foundations governed by bearing ca- (ii) analysis using static loading test 0.6
pacity are not significantly different using the limit results
states design procedure as compared to the working (iii) analysis using dynamic 0.5
stress design procedure. The derivation of the resis- monitoring results
tance factor in Table L-10 is described in more detail
in Reference (24), where it is shown that the esti- (iv) uplift resistance by semi- 0.3
mated reliability index () for shallow foundations empirical analysis
using the resistance factors in Table L-10 ranges
from 2.8 to 3.5, a range which is consistent with val- (v) uplift resistance using loading 0.4
ues commonly used for the design of the building test results
structure. Figure L-3, taken from Reference (24), (b) Horizontal load resistance 0.5
shows the relationship between global safety factor,
resistance factor and reliability index, , using
statistical assumptions for variability in bearing re-
sistance (coefficient of variation 0.3 and ratio of

Commentary L
1.0 sures due to the specified loads. The allowable

Bearing resistance 0.9 bearing pressures in Sections 10.2 and 10.3 of
the CFEM are generally based on a maximum
Resistance factor, φ
kR = 1.1, VR = 0.3 0.8 settlement of approximately 25 mm. The em-
0.7 pirical method is convenient for initial design
of foundations of buildings as well as for final
0.6
design for most ordinary buildings. The foun-

0.5 dations of tall buildings or towers, special
buildings sensitive to movements or buildings
resistance factor 0.4 on sensitive ground, however, should be eval-
0.3 uated using the appropriate direct procedures
described in (a) involving both limit states and
8 0.2
including an estimation of foundation move-
7 ments described in Paragraphs 62 to 67.
Reliability index, β
6
reliability index 61. Selection of characteristic design values of
5 soil and rock properties shall be based on the re-
4 sults of laboratory and field tests and shall take
account of the following:
3 (a) geological and other background information,
2 such as data from previous projects,
(b) the variabilities of the property values,
1 (c) the extent of the zone of ground governing the
0 behaviour of the geotechnical structure for the
1.5 2.0 2.5 3.0 3.5 4.0 limit state considered,
Safety factor, F (d) the influence of workmanship on artificially
placed or improved soils,
Figure L-3 (e) the effect of construction activities on the
Relation between safety factor, resistance factor and properties of in situ ground.
reliability index for bearing resistance, KR = 1.1, VR = 0.3
Account shall be taken of the possible difference
between the properties measured in the tests and
59. Another approach to deal with the ultimate the soil and rock properties governing the behaviour
limit states is contained in the Canadian Foundation of the geotechnical structure due to factors such as:
Engineering Manual (CFEM)(1) and the Danish Code (a) the presence of fissures, which may play a dif-
of Practice for Foundation Engineering.(25) In this ferent role in the test and in the geotechnical
approach the soil properties are modified using par- structure,
tial factors. (b) the time effects, and
Limit States Design Procedure for Shallow (c) the brittleness or ductility of the soil and rock
Foundations tested.
60. When designing a spread foundation one of 62. In many cases, the variabiltiy of a mean
the following design methods shall be used: value of a soil or rock property should be investi-
(a) Direct Method, in which separate analyses are gated, rather than the variability of an individual
carried out for each limit state using calcula- value resulting from a test. The extent of the zone of
tion models recommended in the CFEM and ground governing the behaviour of a geotechnical
appropriate load factors and resistance factors structure for a limit state is usually much larger
described in this Commentary. In the case of than the extent of the zone involved in a soil or rock
serviceability limit states related to settlement, test and consequently the governing parameter is
the settlement under the service loads is deter- often a mean value over a certain surface or volume
mined in accordance with the methods given of the ground.
in Chapter 12 of the CFEM. In the case of ulti-
mate limit states related to bearing capacity, 63. The governing zone of ground may also de-
the foundation forces due to factored loads (in- pend on the behaviour of the supported structure.
cluding wind or earthquake) are compared For instance, when considering a bearing resistance
with the resistances (ruA’ given in Sections for a building resting on several spread footings, the
10.4 to 10.6 of the CFEM) multiplied by the re- governing parameter is the mean strength over each
sistance factors given in Table L-10. individual zone of ground under a footing, if the
(b) Empirical Method, in which allowable bearing building is unable to resist a local failure. If instead
pressures estimated empirically in Sections 10.2 the building is stiff and strong enough, the govern-
to 10.3 of the CFEM are compared to the pres- ing parameter may be the mean of these mean

Commentary L
values over the entire zone or part of the zone of to keep in mind that differential settlement of iso-
ground under the building. lated footings will always occur because of the
natural variability of soils.
Total and Differential Settlements
70. In situations where calculation models are
64. The total and differential settlements and not available or are considered to be unnecessary,
relative rotations for foundations shall be estimated limit states may be avoided by the use of prescrip-
to ensure that these do not lead to the occurrence of tive measures. Prescriptive measures may be used,
an ultimate limit state or a serviceability limit state, for example, to ensure durability against frost action
such as unacceptable cracking or jamming of doors, and chemical or biological attack. These measures
in the supported structure. This requires attention by involve conventional and generally conservative de-
both the geotechnical and the structural engineers. tails in the design, and attention to specification and
control of materials, workmanship, protection and
65. The maximum acceptable relative rotations maintenance procedures.
for open frames, infilled frames and loadbearing or
continuous brick walls are likely to range from Frost Penetration
about 1/2000 to about 1/300 to prevent the occur-
rence of a serviceability limit state in the structure 71. The best indication of frost penetration in a
and about 1.5 to 2 times these values for long term particular locality is local experience. In the absence
movements over many years because of creep of of local experience, however, daily air temperature
building materials. A maximum relative rotation of measurements can be used to estimate the com-
1/500 for short term movements and 1/300 for long bined effects of both depth and duration of freezing.
term movements is acceptable for many structures. The cumulative total of the difference between daily
The relative rotation likely to cause an ultimate limit mean air temperatures and the freezing point is
state is about 1/150. known as the “Freezing Index,” and is expressed in
degree days. Freezing indices for a large number of
66. For normal structures with isolated founda- weather stations in Canada have been published by
tions, the maximum acceptable differential the Atmospheric Environment Service.(26) As a gen-
settlement is about 20 mm between adjacent eral guideline the variation in freezing indices
columns. On sand the differential settlement of across Canada is illustrated in Figure L-4. Informa-
foundations is unlikely to exceed 75% of the maxi- tion on how the “Freezing Index” may be used to
mum settlement and the maximum total settlement estimate depth of frost penetration is given in Refer-
should not exceed about 25 mm. For a raft founda- ences (27) to (30).
tion, the maximum total settlement may be
increased to 50 mm. The maximum allowable total Insulated Shallow Foundations
and differential settlement may be increased in the
case of foundations on clay soils provided the rela- 72. In recent years lightweight plastic insulation
tive rotations remain within acceptable limits and has been used to reduce the loss of ground heat and
provided the total settlements do not cause prob- thereby reduce the depth of frost penetration. Insu-
lems with the services entering the building, tilt, etc. lation should be used for this purpose only after
The above guides concerning limiting settlements careful examination of the pertinent conditions and
apply to simple routine buildings. They should not with a thorough understanding of its effect on the
be applied to buildings which are out of the ordi- temperature at the soil-foundation interface.(30) Insu-
nary or for which the loading intensity is markedly lation is of particular advantage in the design of
non-uniform. unheated buildings such as warehouses, garages
and refrigerated buildings used for food storage. It
67. Differential settlements calculated without is also used to restrict the depth of frost penetration
taking account of the stiffness of the structure tend beneath artificial ice surfaces.
to be overpredictions. An analysis of ground-
structure interaction may be used to justify reduced 73. Insulation can be obtained with relatively
values of differential settlements. high compressive strengths, so that slabs of these
materials can be placed directly below the bearing
68. Differential settlement caused by variability surfaces of foundations. Substantial economic ad-
of the ground should be considered unless it is pre- vantages may accrue where such designs are used,
vented by the stiffness of the structure. For spread because foundations can be located closer to the
foundations on natural ground, its magnitude may ground surface, thereby reducing the costs of pro-
typically be up to 10 mm, but it does not usually viding granular fill to replace frost-susceptible
exceed 50% of the calculated total settlement. soil.(30)
69. Calculation models for settlement analyses

are given in Chapter 12 of the CFEM. It is important

Commentary L
50
72 70
00
50
CANADA
67
5 0 kilometres
72
00 100
65
00
200 0 200 400 600 800
0
4 75 70
50
0 0 67
5 00 6 50 50
62
45 0
50 00
62
40
60
0 0 4 25
0 3 750
00
5 25
0
00 60
55 50
0
57
35
50
57
00
32 00
50
5 5 50
52
00
50
2 750
3 000
3 2 00
2 50
Dawson
35
22 0
50
2 0 50 50
47
1 7500 00
1 50 0 45 50
1 250 4 2 000
1 000 4
0
500 Whitehorse
250 50
750 37
00
35
Yellowknife 32
50
00
30
50 00
27 25
50 0
2 2 00
2
50
17
00
1 00 15
0
50
1 000 Churchill 1 2 00
10
0
75
Edmonton 50
0
Saskatchewan
Victoria Vancouver St. John’s
0
1 00
Calgary Saskatoon
250
500
500 Lake
500
750
750
500
Regina Winnipeg
750
1 00
0 12
50
Winnipeg Sydney
Fredericton
1 500 Thunder Bay 1 250
1 750
Québec
50 1 000
Halifax
17 1 500 0 500
1 25 1 000
750 Montréal 750
Ottawa 1 000
750
500
500
Toronto
Lake
250 250 Erie
Figure L-4
Normal freezing index in degree-days Celsius, based on the period 1931 to 1960
Deep Foundations shafts of diameter greater than about 750 mm are

frequently referred to as caissons in Canada.)
Introduction
76. Loads which may be applied to a deep
74. A deep foundation is a foundation unit that foundation depend not only on the properties of the
provides support for a building by transferring foundation as a structural unit (e.g., the shaft
loads either by end-bearing to a soil or rock at con- strength of a drilled shaft determined on the basis of
siderable depth below the building, or by adhesion CAN3-A23.3-94, “Design of Concrete Structures”),
or friction, or both, in the soil or rock in which it is but also on the properties of the foundation soil (or
placed. Piles are the most common type of deep rock) and of the soil/foundation system (e.g., pile
foundation. capacity as a function of soil strength, settlement of
a drilled shaft as a function of contact pressure).
75. Piles can be premanufactured or cast-in- Thus, the designer must distinguish the structural
place; they can be driven, jacked, jetted, screwed, from the geotechnical capacity of a deep foundation
bored, drilled or excavated. They can be of wood, unit or system, analyze each very carefully and de-
concrete or steel or a combination thereof. (Drilled fine application of loads which may be carried

Commentary L
safely, both from a structural and a geotechnical and are commonly referred to as drilled shafts. In
point of view. In many applications, geotechnical this case, the area of end-bearing contact is known
considerations limit the permissible loads to levels and, provided this area and the character of the
well below those which might be arrived at on the foundation stratum can be defined by inspection,
basis of structural considerations alone. the geotechnical capacity of the deep foundation
can be evaluated on the basis of the allowable bear-
77. Geotechnical criteria for assessing permissi- ing pressure of the foundation stratum. (Refer to
ble loads on a deep foundation are determined on Tables L-6, L-7 and L-8, on shallow foundations.)
the basis of site investigations and geotechnical
analyses. However, in most cases, the quality of a 81. Rock sockets. Frequently, cast-in-place foun-
deep foundation is highly dependent on dations are socketed into rock, either to obtain
construction technique, equipment and workman- higher end bearing capacity at depth or to transfer
ship. Such parameters cannot be quantified nor load to the rock by adhesion or bond along the
taken into account in normal design procedures. walls of the socket. Adhesion is highly dependent
Consequently, as implied in NBC Subsection 4.2.7., on the rock type and on the socket wall condition
deep foundations should be designed on the basis after drilling. Design values used for adhesion in
of in situ load tests on actual foundation units. sound rock lying below weathered or shattered rock
range from 0.7 MPa to 2.0 MPa; however, much
78. Criteria relating to structurally permissible lower values have been observed in practice, where
loads are defined in the design sections of the the construction methods used have produced a
National Building Code applicable to the structural poor contact area. Careful inspection of all rock
materials used in the deep foundation unit. How- sockets prior to concreting is essential. Socketing
ever, the standards referenced in the NBC were may also be employed to provide base fixity and re-
written mainly for the purpose of designing ele- sistance to horizontal movement.(31)(32)
ments and assemblies in the superstructure. A
structural consultant involved in the design of deep 82. Deep foundations may also be driven to rock
foundations must recognize that installation and or into dense basal deposits. In this case, which in-
quality control conditions below grade differ from cludes H-piles, pipe piles driven closed-end or
those above grade; the permissible loads deter- precast concrete piles, the exact area of contact with
mined by the usual structural design methods may the foundation stratum, the depth of penetration
have to be reduced, sometimes to a marked degree, into it or the quality of the foundation stratum are
to account for these differences. Permissible loads largely unknown. Consequently, the load capacity of
can only be selected on the basis of close co- such driven deep foundations should be determined
operation between the geotechnical and structural on the basis of observations during driving, load
consultants for the project. tests and local experience. (Refer to Table L-11.)
79. In this section of the Commentary, sug-
gested values of permissible loads are given for
several kinds of foundation units. These values are
listed solely to provide a first approximation of the
probable loads which, under routine conditions,
might be applied safely to a given kind of unit. In
each case, both geotechnical and structural evalua-
tion and analysis is mandatory. However, as
discussed above, since construction procedures
often have a dominant influence on the load/defor-
mation behaviour of the deep foundation, the choice
of a permissible load is always subject to judgment
and experience and to the provision that appropri-
ate review is carried out as specified in Article
4.2.2.3. of the NBC. Review must be considered an
integral part of the design process.
Geotechnical Requirements of Deep

Foundations
Deep foundations end-bearing on rock

or highly competent deposits
80. Deep foundations which are placed on rock

or on a dense basal deposit, such as till or hard clay,
are bored, drilled or excavated and cast-in-place,

Commentary L
Table L-11
Load Capacities of Driven Piles
Pile Type Load Capacity Recommendations References

(a) End-bearing on rock, High to very high, but dictated by Ultimate pile capacity usually high —
dense till or other driving conditions, conditions of but load/deformation can only be
similar materials basal deposits, pile types and assessed by load test (ASTM
stiffness D1143-81, Method A)
(b) Piles driven into See (a) See (a) (33) (34)
dense sand, sand
and gravel
(c) Piles driven into loose Medium to high, part point First approximation to load capacity, See references above.
to compact sand, resistance, part skin friction use skin friction (kPa) = 50 ± 25. See also References (35)
sand and gravel Define by load test (ASTM D1143- and (36)
81, Method A)
(d) Piles driven into Medium, but “relaxation” effects See (c). Essential to define by load (37)
compact to dense must be checked test
silts
(e) Piles driven into Low to medium, susceptible to long First approximation, use skin (38) (39)
cohesive soils term settlement friction. Soft cohesive soil,
0 – 30 kPa. Firm to stiff cohesive
soil, 30 – 60 kPa. Define by load
test (ASTM D1143-81, Method B)
Piles in granular soils technique and can only be assessed on the basis of
load tests and detailed local experience.
(Refer to Table L-11)
Piles in cohesive soils
83. Piles which are driven into granular soils
derive their load carrying capacity from both point (Refer to Table L-11)
resistance and shaft friction. The relative contribu-
tions of point resistance and shaft friction to the total 86. The load capacity of piles driven into cohe-
capacity of the pile depend essentially on the den- sive materials is governed by the adhesion between
sity of the soil and on the characteristics of the pile. the pile and the soil and, to a much lesser extent
than in granular soils, by the point resistance. This
84. It is commonly assumed that pile driving in is particularly true for soft to firm clays.
granular soils increases the density of the deposit.
Because of this, piles in granular soils should be 87. The adhesion is not always equal to the
driven to the maximum depth possible, without undrained shear strength of the soil since, in some
causing pile damage, in order to obtain the maxi- circumstances, the effect of pile driving markedly
mum working load on the pile. However, in some changes the character of the soil. In soft sensitive
granular soils, such as fine sands or cohesionless clays, complete remoulding of the soil may occur on
silts, the pile capacity may decrease after driving. driving. This effect diminishes with time following
This effect is known as “relaxation.” In contrast, in driving, as the soil adjacent to the pile consolidates.
some coarse sands or other coarse grained deposits, In some cases, soil strength has not returned to the
the load capacity of piles may increase after driving. original undisturbed value even after a considerable
This effect is known as “freeze.” Neither of these ef- period of time.(40)
fects can be assessed quantitatively, except on the
basis of redriving and load testing. 88. Because of the slow rate of regain of
strength in certain cohesive soils, load testing
85. Compacted concrete piles. Compacted or should sometimes be delayed until several weeks
rammed concrete piles in granular soils derive their have elapsed after driving.
load capacity mainly from the densification of the
soil around the base. The capacity of such piles is, 89. In stiff to very stiff cohesive soils, evidence
therefore, entirely dependent on the construction indicates that, in driving, a gap is formed between

Commentary L
the pile and soil; this gap is not always fully closed ations of group action and, for cohesive soils, long
with time, thus minimizing the adhesion to the pile term consolidation settlement. The actual capacity
relative to the high shear strength of the soil. For and load/deformation characteristics of individual
this reason, an approximate limit of 60 kPa has been piles are not significant in this case. The purpose of
suggested for the adhesion value, even for stiff clays friction piles in the upper part of a deep deposit of
(Table L-11). cohesive soils or of granular soils (or silts) is to re-
duce the intensity of pressure acting at ground level

90. Drilled shafts in cohesive soils. Except for and to shift the zone of maximum stress to the
shafts drilled through stiff or very stiff cohesive de- lower levels, where less settlement will result.
posits, the major portion of drilled shaft capacity is
derived from the bearing capacity of a hard or
dense stratum at the base. For a first approximation 95. In the case of an individual pile or where
of bearing capacity, Tables L-6 and L-7 should be the building is narrow in relation to the depth of
used. For a more detailed assessment of bored piles, piles, the zone of pressure increase is spread over a
see Reference (41). large area in comparison with the width of the
foundations; in contrast, where the building is wide,
Spacing and arrangement of piles and friction piles spread the load out very little, and the
drilled shafts effect of the pile foundation on the soil is practically
the same as that of a raft foundation without piles.
91. The following should be considered during In this case, the total bearing value of the piles in
the spacing and arrangement of piles and drilled the foundation bears no relation to the carrying
shafts: value of an individual pile by itself; the settlement
(a) the overlap of stresses between units, which of the foundation is, therefore, governed by the
influences total load capacity and settlement, character of the subsoil, not by the load capacity of
(b) overstressing of weaker zones at depth, and the piled foundation.
(c) installation difficulties, particularly the effects
on adjacent piles or drilled shafts. Load tests on deep foundations
92. In most cases the spacing, D, between the
centres of driven piles of average diameter, d, 96. Use of load tests. As previously indicated,
should not be less than 2.5 d. load testing of piles, as specified in NBC Sentence
4.2.7.2.(2), is the most positive method of determin-
Settlement and group effects in piles ing load capacity. Depending upon the type and size
of the foundation, such load tests may be performed
93. In practice, piles are frequently used in at different stages during design and construction.
groups; however, most of the published literature
deals with the behaviour of single piles. Leonards
(42) states that, “there is no consistent relationship 97. Load tests during design. The best method
between the settlement of a single pile and the set- of designing a pile foundation is to perform pile
tlement of the pile group at the same load per pile. driving and loading tests. The number of tests, the
Therefore, selecting a design load on the basis of the type of pile tested, the methods of driving or of in-
load at a given gross or net deflection, or at a given stallation and of test loading should be selected by
fraction of the ultimate pile capacity, is equivalent to the engineer responsible for the design. The follow-
accepting an unknown factor of safety with respect ing points should be considered:
to satisfactory performance of the foundation.” This (a) The test program should be carried out by a
statement is certainly valid for all piled foundations person competent in this field of work.
where the piles derive their support from skin (b) Adequate soil information should be obtained
friction, or from combined skin friction and end- at the test location.
bearing; however, group effects may be less critical (c) The piles, the equipment used for driving or
where piles derive all of their support or the major other method of installation and the procedure
portion of it from end-bearing on a relatively in- should be those intended to be used in the
compressible stratum. An example of such support construction of the foundation.
is where piles are driven through weak deposits to (d) As a minimum, the head of a pile should be
end-bearing on rock. For this case, the engineer instrumented to record the total pile and soil
normally relies on some means of assessing the dy- deformation. Where possible, deformation
namic resistance during pile driving complemented measurements should also be made at the tip
by load tests to define the deformation characteris- of the pile and at intermediate points to allow
tics of the piles under load. for a separate evaluation of point resistance
and skin friction.
94. In contrast to true end-bearing pile founda- (e) The driving process should be observed in de-
tions, where the load/deformation characteristics of tail and, wherever possible, stress levels in the
individual piles are significant, the use of friction pile assessed (e.g., by means of the wave equa-
pile foundations is generally governed by consider- tion method of analysis).

Commentary L
(f) The piles should be loaded to at least twice the Installation and Structural Requirements
proposed working load and preferably to fail- of Deep Foundations
ure.
101. In most cases, the maximum allowable load
98. Routine load tests during construction. on a deep foundation unit is governed by geotech-
Load tests should be performed on representative nical considerations. The design capacity of a deep
deep foundation units at early stages of
foundation unit determined from structural consid-

construction. The purpose of such tests is to ascer- erations represents the maximum axial load which
tain that the allowable loads obtained by design are theoretically could be carried; however, this load is
appropriate, and that the installation procedure is generally less than could be applied to a comparable
satisfactory. unit used in the superstructure of a building because
99. The selection of the test piles should be (a) the actual placing of deep foundations fre-
made by the engineer responsible for the design on quently deviates from the position and
the basis of observed driving behaviour or installa- alignment assumed in design,
tion features. (b) once in place, deep foundation units often can
neither be inspected nor repaired, and
100. Load tests for control. Where full advan- (c) the placement of concrete in cast-in-place deep
tage is to be taken of Clause 4.2.4.1.(1)(c) and foundations frequently cannot be done with
Sentence 4.2.7.2.(2) of the NBC, a sufficient number the same degree of control as in structural
of load tests must be carried out on representative columns.
units to ascertain the range of the pile performance
under load. Load tests for control should be per- 102. In Tables L-12 to L-14 guidelines are given
formed on one out of each group of 250 units, or to assist in determining a reasonable axial design
portion thereof, of the same type and capacity. Load capacity for deep foundation units under common
tests should also be performed on one out of each conditions. These tables are not a substitute for
group of units where driving records or other obser- structural analysis and design, but only provide a
vations indicate that the soil conditions differ conservative guide for routine situations which may
significantly from those prevailing at the site. Selec- confront a designer, where a unit may be considered
tion of the deep foundation units to be load tested as a short column and where axial load governs the
is the responsibility of the design engineer. design.

Commentary L
Table L-12
Guidelines for Driven Piles
Type of Pile Normal Size Typical Pile Installation

Structural Considerations Notes
Range Load, kN Considerations
(a) Timber 180 to 180 to 450 Must be checked in Cannot be inspected. Preservative treatment
250 mm tip accordance with NBC Susceptible to normally required.
Subsection 4.3.1. damage during hard (CAN/CSA O80-M89)
driving. Tip
reinforcement
recommended where
driven to end bearing
(b) Steel 200 to 350 to 1 800 • Must be checked in May be damaged Tip points often
sections 350 mm accordance with NBC during driving but load required for hard
(H, WF) Subsections 4.3.3. and capacity not driving. Average
4.3.4. necessarily reduced thickness of flange or
• End bearing: allowable web, t ≥ 1 cm.
working stresses usually Projection of flange
0.3fy when driven to 14t
end bearing refusal on
(c) Pipe 200 to 350 to 1 800 rock or dense strata; Suitable for inspection Normally driven
sections 600 mm diam higher stresses possible after driving. Concrete closed-end. Tip
under specific controlled quality highly reinforcement or drive
conditions. dependent on shoe required when
• Friction: working placement method driven open-end. Pipe
stresses usually thickness > 5 mm, but
governed by 10 mm recommended
geotechnical
considerations and
rarely exceed 80 MPa.
In pipe piles, concrete
strength does not
normally contribute to
pile capacity unless the
pile is driven to end
bearing.
(d) Precast 200 to 350 to 1 000 • End bearing: capacity Cannot be inspected. Refer to ACI
concrete 300 mm must be checked in Careful selection and 543R-1980. Possible
sections accordance with NBC driving method tensile stresses in
300 to 900 to 2 500 Subsection 4.3.3. required to prevent concrete during ‘soft’
900 mm Normally f’c > 27.5 MPa. damage driving. High
• Friction: the capacity of compressive stresses
friction piles is normally in concrete during
governed by both ‘hard’ driving. Tip
installation method and reinforcement usually
geotechnical essential
considerations; average
compressive stress
under load rarely
exceeds 10 MPa

Commentary L
Table L-13
Guidelines for Compacted, Expanded Base Piles
Type of Pile Normal Size Typical Load, Structural Installation

Notes
Range, mm kN Considerations Considerations
(a) Rammed 350 to 600 450 to 1 350 Concrete quality highly Cannot be inspected. Allowable load
shaft dependent on technique Contamination of frequently determined
concrete. ‘Necking’ of on the basis of energy
shafts. Possible damage required to expel
by adjacent piles. measured volumes of
concrete at base. Highly
dependent on
judgement and
experience. Possible
heave of all piles must
be continuously
monitored.
(b) Steel pipe 300 to 500 450 to 1 550 Where the pipe wall Less subject to damage See (a) above
shaft, thickness < 5 mm, the than (a) above. Shaft
concrete structural contribution of can be inspected prior
filled the pipe should be to filling
disregarded
Table L-14
Guidelines for Drilled Shafts
Normal Size Typical Structural Installation

Type of Shaft Notes
Range Load, kN Considerations Considerations
(a) Uncased plain 30 to 700 mm 250 to 450 Good concrete quality Where shaft diameter Not recommended for
concrete diam shaft not always possible < 700 mm normally normal application
cannot be inspected, where caving can occur
thus permitting only
low allowable loads
< 450 kN.
(b) Uncased. 750 to 450 to Generally good Can be inspected. Usually under-reamed
Reinforced or 1 500 mm 45 000 concrete quality Where temporary to provide belled base.
plain concrete. diam shaft possible with f’c > 20 casing is used to retain Bell sides typically at
Under-reamed or MPa < 35 MPa. Can wet, caving soil, high 2(V) to 1(H). Often not
straight normally be designed slump concrete may be under-reamed where
in accordance with required. Precautions bearing on sound rock.
NBC Subsection 4.3.3. should be taken to
(CAN3-A23.3-94) prevent contamination
of concrete.
(c) Cased. 450 to 450 to See (b) above. Must be Can be inspected Usually not under-
Permanent steel 1 500 mm 45 000 checked as composite reamed. Generally
pipe lining diam shaft unit in accordance with socketed where taken
NBC Subsections to rock. Design for
4.3.3. (CAN3-A23.3- complete load transfer
94) and 4.3.4. through socket.
Essential to seat liner
on rock bearing surface.
Drive shoe usually fitted
to pipe liner

Commentary L
103. The flexural capacity and ductility of piles 108. Piles may also be damaged by driving
should be considered when, under certain soil con- through obstructions, such as boulders or fill mate-
ditions, the soil either does not provide lateral rial, or by sloping rock surfaces which may deflect
support or could cause lateral loads to be applied to the pile or create high local stresses leading to seri-
the piles. ous deformation or breakage.
104. Frequently, economies can be made by us- 109. Excessive bend or sweep may be experi-
ing higher capacities or different techniques. Such enced when driving long piles (30 m or more). A
higher capacities should only be used in conditions discussion of allowable bending of piles is given in
where they can be justified as suitable and when Reference (43).
quality can be ensured by an adequate program of
inspection and load tests. 110. The use of steel reinforcing tips is strongly
recommended whenever ends may be damaged. Tip
Driven piles reinforcement may also reduce damage incurred
through overdriving.
105. This type of deep foundation unit may suf-
fer structural damage when being driven. 111. Movement of adjacent piles during driv-
Determination of capacity is generally made by com- ing. Where a group of piles is to be placed through
paring driving resistance (blows per 30 cm) with the silt or clay, measures shall be taken to indicate any
energy or size of hammer blow and relating these movement of each pile during the installation of ad-
figures to previous experience or to the behaviour jacent piles. Horizontal and vertical movement
of similar piles subjected to static load tests. For this should be recorded.
purpose, observations of pile driving must include:
(a) pile length and weight, 112. Piles which have suffered vertical move-
(b) hammer type (e.g., drop, diesel, ram weight), ment should generally be redriven. Piles which
(c) hammer energy applied, have suffered horizontal displacement must be in-
(d) type and thickness of packing, and vestigated for structural damage.
(e) blows per 30 cm and elastic rebound of pile, or
(f) acceleration and stress at head of pile. 113. Jetting or pre-excavation. When jetting,
predrilling or other pre-excavation methods have
106. The assessment of pile stresses during driv- been used during pile installation, the pile tip
ing by the theory of wave propagation or by the should be driven below the depth of pre-excavation
“wave equation” method of pile analysis is useful. to the required bearing. Care must be taken to avoid
By assigning appropriate elastic properties to such jetting, pre-driving or pre-excavating to a depth or
parameters as the pile/cushion system and the in a manner that will affect the design capacity of
pile/soil system, the penetration per blow and pile piles previously placed. This is discussed in detail
stresses for a given hammer energy can be com- in ACI 543R-1980.(44)
puted; however, these results and the extrapolation
of the penetration per blow to a definition of ulti- Cast-in-place deep foundations
mate pile capacity are, at best, only approximations.
The “wave equation” method, in common with all 114. Cast-in-place deep units can be divided
empirical dynamic pile formulae, calls for the exer- into two main categories: compacted expanded base
cise of judgment and experience. No method, in piles (Table L-13) and drilled shafts (Table L-14).
itself, can provide definitive values either for driv-
ing criteria or load/deformation characteristics of a 115. The placement of the materials forming
driven pile. Pile load tests are essential to confirm such units is crucial. It is difficult, if not impossible,
the driving criteria used and to assess load/defor- to ensure the same level of quality in placing con-
mation performance. crete in such units as in a building superstructure.
Careful attention must be given to the methods of
107. Damage to driven piles. Piles may be installation, concrete mix proportions and place-
damaged by attempting to drive to an excessively ment methods, and to the degree of inspection
small “set” per blow or to an excessively large num- possible. The allowable loads on such units must be
ber of blows at high resistance. This is known as adjusted accordingly, in keeping with sound design,
“overdriving.” The driving set should be established engineering experience and judgement.
so as to achieve a reasonable performance under
load without incurring the risk of serious damage. 116. Concrete cast in place. The placing of con-
Driving stresses depend upon the hammer, blows, crete in pipe piles, expanded base pile shafts and in
size and type of pile, length of pile, cushion mate- drilled shafts can be classified in two categories:
rial and soil conditions. These factors must be (1) Concrete placed in the dry should be placed by
examined for each situation and acceptable “set” guided free fall, bucket or chute. Segregation
criteria determined on the basis of previous experi- may occur if concrete is allowed to fall through
ence and load testing. a reinforcing cage or similar obstruction.

Commentary L
Concrete of more than 100 mm slump Location and alignment

placed by free fall of 5 m or more in non-
reinforced or lightly reinforced shafts receives 119. The exact location of each deep foundation
adequate compaction and does not require vi- unit should be staked in advance and checked im-
bration. mediately prior to installation of each unit. After
Placement by tremie methods is necessary completion of the installation the location of each
when a considerable inflow of groundwater is unit should be checked against design location and
present or when there is standing water in the permissible deviation as indicated on the design
hole. documents.
(2) Concrete placed under water should be placed
through a tremie pipe or by pump in such a 120. As required in NBC Article 4.2.7.3., permis-
way as to eliminate any contamination, wash- sible deviations from the design location shall be
ing or dilution of the concrete by the water. It determined by design analysis. In practice, piles and
should have a 150 to 200 mm slump and vi- shafts can usually be positioned within a tolerance
bration should not be applied. (Refer to CAN/ of 80 mm; for practical reasons smaller tolerances
CSA-A23.1-94, “Concrete Materials and Meth- should not be specified.
ods of Concrete Construction.”)
121. As required in NBC Article 4.2.7.4., where
117. Reinforcing steel for cast-in-place units. a deep foundation unit is wrongly located, the con-
Reinforcing steel is generally placed pre-assembled dition of the foundation shall be assessed by the
as in a cage. During placement, the steel may be person responsible for the design and the necessary
subjected to severe handling and placement stresses changes made.
and to impact. Placement cannot be made with as
high a degree of accuracy as in a superstructure, nor 122. During and after installation of any deep
can it be easily checked. foundation unit, its alignment should be checked
against the design alignment and the permissible
118. For the design of cast-in-place foundations, deviation as indicated on the design documents.
the provisions of CAN3-A23.3-94, “Design of Con-
crete Structures for Building” should, therefore, be 123. Current practice is to limit the total devia-
amended in the following respects: tion from design alignment to a percentage of the
(a) Reinforcing steel assemblies should be de- final length of the deep foundation unit; 2% is a
signed and constructed so as to withstand all common value. However, such practice does not en-
handling and placing stresses without defor- sure proper structural behaviour of the unit since it
mation which would impair the structural does not take into account the length over which
performance of the unit. this deviation is distributed.
(b) Weldable steel should be employed, in most (a) The total deviation from alignment of a deep
cases, to permit construction of rigid and foundation unit has little influence on its geo-
strong assemblies. technical capacity unless it reaches values
(c) The clear distance between longitudinal bars greater than 10% of the length of the unit.
should not be less than 75 mm. (b) Practically all piles, particularly when driven,
(d) Ties or spirals may be welded to the longitudi- are more or less out of design alignment. A
nal bars. Welding should be in accordance straight pile is a theoretical concept seldom
with CSA W59-1989, “Welded Steel achieved in practice.
Construction (Metal-Arc Welding).” Welded (c) Only the radius of curvature of a deep founda-
spirals or ties should be of wire not less than tion unit is important for its structural and
7.0 mm in diam, with pitch not more than geotechnical behaviour. The maximum allow-
300 mm and with not less than 75 mm clear able radius of curvature should be determined
space between ties or spirals. by design whenever such radius is required to
(e) The possibility of misplacing the reinforcing be measured during inspection. A discussion
bars should be allowed for in the design, and of allowable bending of piles is given in Refer-
reasonable tolerances established for field per- ence (43).
formance: e.g., ± 75 mm of correct bar location
in plan, ± 150 mm of correct bar location in el- Limit States Design
evation.
124. The limit states to be considered are as dis-
(f) Generally, longitudinal steel should be uni-
cussed in Paragraph 50 of this Commentary.
formly distributed around the cross-section, as
an assembly may become twisted during 125. The ultimate limit states for pile founda-
placement. tions that should be considered include:
(a) vertical bearing capacity failure
(b) uplift (pull-out)
(c) lateral resistance failure of the soil

Commentary L
(d) structural failure of the pile in compression, from 3.5 to 4.5 and are higher than those for shallow
tension, bending, buckling, or shear. foundations. This increase most likely reflects in-
creased uncertainties in the calculation of resistance
126. The serviceability limit states that should for deep foundations which are not taken into ac-
be considered include: count in the values of VR assumed in Reference (24).
(a) excessive movements (settlement, differential
settlement, heave, lateral movement, and tilt or 131. The characteristic values for the soil param-
rotation) eters are selected as discussed in Paragraphs 80 to
(b) unacceptable vibrations. 82.
127. The discussion in Paragraphs 74 to 123 Permafrost
dealing with the working stress design method for
deep foundations also applies to the limit states de- 132. The lines on Figure L-5 indicate the ap-
sign of deep foundations. proximate southern limit of permafrost and the
128. The design of pile foundations shall be boundary between the discontinuous and continu-
based on one of the following methods: ous permafrost zones in Canada. The distribution of
(a) the results of static load tests which have been permafrost varies from continuous in the north to
demonstrated, by means of calculations or oth- discontinuous in the south. In the continuous zone
erwise, to be consistent with other relevant permafrost occurs everywhere under the ground
experience; surface and is generally several decametres thick.
(b) empirical or analytical calculation models as Southward, the continuous zone gives way gradu-
recommended in the Canadian Foundation En- ally to the discontinuous zone, where permafrost
gineering Manual (CFEM) in which separate exists in combination with some areas of unfrozen
analyses are carried out for each limit state material. The discontinuous zone is one of broad
with appropriate values for the loads and transition between continuous permafrost and
characteristic soil parameters and appropriate ground having no permafrost. In this zone, per-
load and resistance factors as described in this mafrost may vary from a widespread distribution
Commentary. The methods used to determine with isolated patches of unfrozen ground to pre-
the ultimate limit state with respect to bearing dominantly thawed material containing islands of
capacity are given in Chapter 20 of the CFEM. ground that remain frozen. In the southern area of
In the case of the serviceability limit state, the this discontinuous zone, permafrost occurs as scat-
methods to determine settlement are described tered patches and is only a few metres thick.
in Chapter 20 of the CFEM.
133. The lines on this map must be considered
129. The load factors and load combinations are as the approximate location of broad transition
as outlined in Section 4.1. of the National Building bands many kilometres wide. Permafrost also exists
Code. at high altitudes in the mountains of Western
Canada a great distance south of the southern limit
130. The resistance factors for the design of shown on the map. Information on the occurrence
deep foundations are given in Table L-10. These fac- and distribution of permafrost in Canada has been
tors were obtained from direct calibration to the compiled by the Institute for Research in
working stress design approach. Calculated b values Construction, National Research Council
for deep foundations given in Reference (24) vary Canada.(45)(46)

Commentary L
2
2
2
1 1
Figure L-5
Permafrost region. 1 – Discontinuous zone, 2 – Continuous zone.
References (8) F.A. Tavenas, R.S. Ladd and P. LaRochelle,

The Accuracy of Relative Density Measure-
(1) Canadian Foundation Engineering Manual ments: Results of a Comparative Test
3rd Edition, Canadian Geotechnical Society, Programme. ASTM, STP 523, 1973.
1992, 46 pp. (Available from BiTech Publish- (9) K. Terzaghi, Influence of Geological Factors
ers, Suite 173, 11860 Hammersmith Way, on the Engineering Properties of Sediments.
Richmond, B.C., V7A 5G1.) Economic Geology, 5th Anniv. Volume, 1955,
(2) K. Terzaghi and R.B. Peck, Soil Mechanics in pp. 557-618.
Engineering Practice. J. Wiley & Sons, New (10) L. Bjerrum, Engineering Geology of Norwe-
York, 1967. gian Normally-consolidated Marine Clays as
(3) E. Hoek and J.W. Bray, Rock Slope Engineer- Related to Settlements of Buildings, Seventh
ing. Inst. of Mining and Metallurgy, 1972. Rankine Lecture. Geotechnique, Vol. 17, 1967,
(4) G.F.A. Fletcher, Standard Penetration Test: Its pp. 83-117.
Uses and Abuses. Journal of Soil. Mech. (11) C.B. Crawford, Interpretation of the Consoli-
Found. Div., Proc., Am. Soc. Civ. Eng., Vol. 91, dation Test. Journal of Soil Mech. Found. Div.,
SM4, 1965, pp. 67-75. Proc., Am. Soc. Civ. Eng., Vol. 90, SM5, 1964,
(5) R.B. Peck, W.E. Hanson and T.H. Thornburn, pp. 87-102.
Foundation Engineering. J. Wiley & Sons, (12) J.H. Schmertmann, Estimating the True Con-
New York, 1974. solidation Behavior of Clay from Laboratory
(6) J.W. Gadsby, Discussion of the ‘The Correla- Test Results. Proc., Am. Soc. Civ. Eng., Vol. 79,
tion of Cone Size in the Dynamic Cone Separate 311, 1963.
Penetration Test with the Standard Penetra- (13) I.C. MacFarlane, Ed., Muskeg Engineering
tion Test.’ Geotechnique, Vol. 20, 1971, pp. Handbook. Univ. of Toronto Press, Toronto,
315-319. 1969.
(7) F.A. Tavenas, Difficulties in the Use of Rela- (14) V. Milligan, L.G. Soderman and A. Rutka, Ex-
tive Density as a Soil Parameter. ASTM, STP perience with Canadian Varved Clays. Journal
523, 1973. of Soil Mech. Found. Div., Proc., Am. Soc. Civ.
Eng., Vol. 88, SM4, 1962, pp. 31-67.

Commentary L
(15) C.B. Crawford, Engineering Studies of Leda (32) F.A. Tavenas, Contrôle du roc de fondations
Clay. In Soils in Canada. R.F. Legget, Ed., Roy. de pieux forés à haute capacité. Can. Geotech.
Soc. Can., Spec. Publ. No. 3, 1961, pp. 200-217. J., Vol. 8, 1971, pp. 400-416.
(16) C.B. Crawford, Quick Clays of Eastern (33) G.G. Meyerhof, Penetration Tests on Bearing
Canada. Eng. Geol., Vol. 2, No. 4, 1968, pp. Capacity of Cohesionless Soils. Journal of Soil
239-265. Mech. Found. Div., Proc., Am. Soc. Civ. Eng.,
(17) P. LaRochelle, J.Y. Chagnon and G. Lefebvre, Vol. 82, SM1, Paper No. 866, 1956.
Regional Geology and Landslides in Marine (34) V.G. Berezantsev, V.S. Kristoforov and V.N.
Clay Deposits of Eastern Canada. Can. Golubkov, Load Bearing Capacity and Defor-
Geotech. J., Vol. 7, No. 2, 1970, pp. 145-156. mation of Pile Foundations. Proc. Intern.
(18) J.J. Hamilton, Shallow Foundations on Conf. Soil Mech. Found. Eng., Paris, Vol. 2,
Swelling Clays in Western Canada. Proc. In- 1961, pp. 11-15.
tern. Res. Eng. Conf. Expansive Clay Soils, (35) A.S. Vesic, Tests on Instrumented Piles,
Texas A & M Univ., Vol. 2, 1965, pp. 183-207. Ogeechee River Site. Journal of Soil Mech.
(19) R.M. Hardy, Construction Problems in Silty Found. Div., Proc., Am. Soc. Civ. Eng., Vol. 96,
Soils. Eng. Journal, Vol. 33, No. 9, 1950, SM2, 1970, pp. 561-584.
pp. 775-782. (36) E.E. De Beer, The Scale Effect in the Transpo-
(20) R.M. Quigley and R.W. Vogan, Black Shale sition of the Results of Deep Sounding Tests
Heaving at Ottawa, Canada. Can. Geotech. J., on the Ultimate Bearing Capacity of Piles and
Vol. 7, No. 2, 1970, pp. 106-112. Caisson Foundations. Geotechnique, Vol. 13,
(21) R.M. Hardy, Engineering Problems Involving 1963, pp. 39-75.
Preconsolidated Clay Shales. Trans. Eng. Inst. (37) N.C. Yang, Relaxation of Piles in Sand and In-
Can., Vol. 1, 1957, pp. 5-14. organic Silt. Journal of Soil Mech. Found. Div.,
(22) R.J.E. Brown, Permafrost in Canada. Univ. of Proc., Am. Soc. Civ. Eng., Vol. 96, SM2, 1970,
Toronto Press, Toronto, 1970. pp. 395-410.
(23) F.J. Sanger, Foundation of Structures in Cold (38) M.J. Tomlinson, The Adhesion of Piles Driven
Regions. Cold Reg. Res. Eng. Lab., Cold Reg. in Clay Soils. Proc. Intern. Soc. Soil Mech.
Sci. Eng. Monogr., Vol. 111-C4, 1969. Found. Eng., London, Vol. 2, 1957, pp. 66-71.
(24) D.E. Becker, D.E. Allen, K.S. Ho and K.T. Law. (39) P. Eide, J.N. Hutchinson and A. Landva, Short
Development of Limit States Design for Foun- and Long Term Loading of a Friction Pile in
dations in the National Building Code of Clay. Proc. Intern. Conf. Soil Mech. Found.
Canada. Proc. International Symposium on Eng., Paris, Vol. 2, 1961, pp. 45-53.
Limit States Design in Geotechnical Engineer- (40) M.J. Tomlinson, Foundation Design and
ing, Copenhagen, Vol. 2, 1993, pp. 479-490. Construction. John Wiley & Sons, New York,
(25) Danish Code of Practice for Foundation Engi- 1963.
neering, 3rd Edition. DS 415(1984), Danish (41) J.D. Burland, F.G. Butler and P. Dunican, The
Technical Press, Copenhagen, 1984. Behavior and Design of Large Diameter Bored
(26) Normal Freezing and Thawing Days for Piles in Stiff Clay. Proc. Symp. Large Bored
Canada 1931-1960. Atmospheric Environment Piles, Inst. Civil Eng., London, 1966, pp. 51-71.
Service, 4905 Dufferin Street, Downsview, (42) G.A. Leonards, Summary and Review of Part
Ontario M3H 5T4. II of the Symposium on Pile Foundations.
(27) U.S. Army Corps of Engineers. Report on Hwy. Res. Record No. 333, Highway Research
Frost Investigations, 1944-1945. Corps Engrs., Board, Washington, 1970, pp.55-59.
New England Division, Boston, 1947. (43) B.H. Fellenius, Bending of Piles Determined
(28) G.H. Argue, Frost and Thaw Penetration of by Inclinometer Measurements. Can. Geotech.
Soils at Canadian Airports. Can. Dept. Trans., J., Vol. 9, 1972, pp. 25-32.
Air Services, Constr. Eng., Arch. Branch, Rep. (44) ACI Committee 543. Recommendations for
CED-6-163, 1968. Design, Manufacture and Installation of Con-
(29) W.G. Brown, Difficulties Associated with Pre- crete Piles. ACI 70-50, ACI Manual of
dicting Depth of Freeze or Thaw. Can. Concrete Practice, Part 3, Detroit, 1974.
Geotech. J., Vol. 1, pp. 215-226, 1964. (Also (45) Permafrost Map of Canada (a joint production
NRC 8276, Division of Building Research, of the Geological Survey of Canada and
National Research Council Canada, Ottawa.) DBR/NRC). August 1967. NRC 9769.
(30) L. Robinsky and K.E. Bespflug, Design of In- (46) R.J.E. Brown, Permafrost Map of Canada.
sulated Foundations. Journal of Soil Mech. Canadian Geographical Journal, February
Found. Div., Proc., Am. Soc. Civ. Eng., Vol. 99, 1968, pp. 56-63. NRC 10326.
SM9, 1973, pp. 649-667.
(31) D.F. Coates, Rock Mechanics Principles. Mines
Branch Monograph 874, Queen’s Printer,
Ottawa, 1967, p. 358.

130
Commentary M
Structural Integrity of Firewalls
1. Sentence 3.1.10.1.(1) of the National Building which would allow premature fire spread through
Code requires that, where framing members are the wall.
connected to or supported on a firewall and such
members have fire-resistance ratings less than that 8. To assess the potential for such damage, ther-
required for the firewall, the connections and sup- mal expansion of the structure should be estimated
ports for such members must be designed so that on the basis of a 500C temperature increase in com-
the collapse of the framing members during a fire bination with the thermal coefficients given in Table
will not cause collapse of the firewall. Sentence D-1 of Commentary D. The expansion of the struc-
4.1.10.4.(1) requires that the firewall be designed to ture toward the firewall can be assumed to begin at
resist a factored lateral load of 0.5 kPa under fire a vertical plane in the fire compartment 20 m from
conditions. the firewall or half the width of the fire compart-
ment, whichever is less.
2. These requirements, along with others in
Subsection 3.1.10., form part of a general require- 9. In assessing thermal effects, attention should
ment that the fire not spread between compartments be given to the effect on the stability of the firewall
separated by a firewall within the required fire- from distortion due to temperature differential
resistance rating for the wall (4 h for high fire through the wall.
hazard occupancies and 2 h for other occupancies).
To achieve this the wall must not be damaged to the 10. If thermal movements are sufficient to dam-
extent that it allows fire spread during this time. age the firewall, either adequate clearances should
be provided or the firewall and structure on both
3. In order to meet the requirement for struc- sides should be detailed to prevent wall damage.
tural integrity of firewalls the following loading
conditions apply. Design Approaches
Lateral Loads on Firewalls 11. Design approaches to satisfy the general
requirements for structural integrity of firewalls in-
4. To prevent collapse of the firewall during the
clude the following.
fire from explosion of unburned gases, glancing
blows from falling debris, force and thermal shock Double Firewall (3.1.10.1.(2))
of fire-hose stream and wind pressure, Sentence
4.1.10.4.(1) requires that the firewall be designed for 12. Here the structure on each side is tied to a
a factored lateral load of 0.5 kPa. If the structure separate wall in such a way that, when the structure
exposed to the fire has less fire resistance than re- exposed to fire fails, only one wall collapses without
quired for the firewall, it is assumed to have failed damaging the remaining wall. A schematic example
and therefore to provide no lateral support to the is shown in Figure M-1. Each wall should have at
firewall. least half the total required fire-resistance rating.
The separation between the walls must satisfy the
5. Sentence 4.1.10.4.(1) also requires that the
requirements for thermal expansion in Paragraphs
firewall be designed for the normal structural
7–10 and for earthquakes in Commentary J.
requirements for interior walls for wind and earth-
quake, including that for pounding damage.
6. The building structure, including the firewall,
should also be designed to provide structural in-
tegrity in accordance with the recommendations of
Commentary C Structural Integrity.
Thermal Effects
7. Thermal expansion of the structure exposed
to a fire must not cause damage to the firewall

Commentary M
beam or joist beam or joist

beam or truss
column
beam or truss
(1) column
firewall
(1)
firewall
separation (2) separation (2)
Figure M-1
Schematic example of double firewall
Figure M-2
Schematic example of cantilever firewall
Notes to Figure M-1:
(1) Each firewall must be tied to the adjacent structure in accordance Notes to Figure M-2:
with Paragraph 11 and reinforced in accordance with Paragraphs 4 (1) Firewall is not tied to the structure and is designed as a cantilever
and 5. from the foundation with reinforcement and pilasters in accordance
(2) Firewalls must be separated in accordance with Paragraphs 5 and with Paragraphs 4, 5, 10 and 11.
10. (2) Separation may be required in accordance with Paragraphs 5 and
10.
Tied Firewall
Cantilever Firewall
14. Here the structure to each side of the wall
13. Here the structure on either side is not con- provides lateral support to the wall and is tied to-
nected to the wall, so that collapse of the structure gether in such a way that lateral forces resulting
exposed to the fire does not collapse the firewall. A from collapse of the structure exposed to the fire are
schematic example is shown in Figure M-2. Rein- resisted by the structural framework on the other
forcement of the cantilever wall and foundations for side. Lateral forces are recommended in Paragraphs
overturning will generally be required to resist the 12 and 13. Suitable provisions must be made to
lateral loads specified in Sentence 4.1.10.4.(1). Pi- transmit these forces to members on opposite sides
lasters will frequently be needed to provide the of the firewall. A schematic example is shown in
requisite lateral load capacity. Figure M-3.

Commentary M
structural diaphragm (1)
beam or joist beam or joist

elongated hole
(for wood shrinkage)
beam or bond beam
truss (for masonry)
tie rod
ties (2)
(1)
blocking
firewall (3)
(2)
firewall
Figure M-3
Figure M-4
Schematic example of tied firewall Example of a weak-link connection used in wood frame
construction
Notes to Figure M-3:
(1) Structural diaphragm resistance may be required in accordance Notes to Figure M-4:
with Paragraphs 11, 16 and 17. (1) Blocking connection to wood frame must be detailed to act as a
(2) Ties must be located and detailed in accordance with Paragraphs weak link in accordance with Paragraph 11.
11, 16 and 17. (2) Firewall must be reinforced and detailed in accordance with
(3) Firewall must be reinforced and detailed in accordance with Paragraphs 4, 5, 10 and 11.
Paragraphs 4, 5 and 10.
Tied Firewalls: Horizontal Forces

from Collapsing Structure
Weak-link Connections
16. Where a structure of fire resistance less than
15. Here structural components are supported that required for the firewall is tied through the
by the firewall in such a way that the failing struc- wall to the structure on the other side of the wall,
ture collapses without causing the firewall to be the supporting structure and tie should be designed
severely damaged. As with a tied firewall, the struc- for a factored horizontal force equal to wBL2/8S,
ture may also provide lateral support to the wall. If where w is the dead weight plus 25% of the speci-
a weak link is provided on each side of the firewall, fied snow load, B is the distance between ties, L is
the link on the fire side must break away while the the span of the collapsing structure between
link on the non-fire side does not. This approach columns perpendicular to the wall and S is its sag,
has traditionally been used in timber construction, assumed to be 0.07 L for steel open-web beams and
where timber beams or joists bear without anchors 0.09 L for steel solid-web beams. The supporting
into pockets of firewalls and can twist free when structure should be capable of resisting the recom-
they collapse.(1)(2) Figure M-4 shows a more recent mended forces for ties within a 10 m length of
technique for a weak link connection to a block fire- firewall; the other ties are assumed to carry no force
wall. If this technique is used, care must be taken to (see Figure M-5). The factored resistance of the tie
provide adequate anchorage for wind uplift and should include a reduction factor of 0.5 to take ac-
earthquake. count of reduced yield strength at high temperature.

Commentary M
Plan
firewall
supporting
structure
area of
10 m
equal
heat
intensity
Section
supporting
structure fire
Figure M-5
Horizontal forces on a tied firewall
17. Alternatively, if the firewall is located so

that the roof structure has the same resistance to
horizontal forces on either side of the firewall (e.g.,
the firewall is located mid-way between end walls
or expansion joints of a structurally symmetric
building), only the tie need be designed for the fac-
tored horizontal force wBL2/8S.
Reference
(1) Canadian Wood Council. Wood and Fire
Safety, Chapter 5, Structural Fire Protection,
Ottawa, 1991.
(2) Canadian Concrete and Masonry Codes
Council. Firewalls – A Design Guide, Ottawa,
1992.

Appendix A
List of Referenced Standards
Issuing Agency Document Number Title of Standards Reference

ACI 543R-1980 Recommendations for Design, Manufacture Commentary L Table L-12, Para.
and Installation of Concrete Piles 113
ASTM D1143-81 (1994) Test Methods for Piles Under Static Axial Commentary L Table L-11
Compressive Load
CSA A23.1-94 Concrete Materials and Methods of Commentary K Para. 48
Concrete Construction Commentary L Para. 116
CSA A23.3-94 Design of Concrete Structures Commentary A Para. 2, Table A-1
Commentary D Table D-1
Commentary J Para. 87, Para. 100
Commentary K Para. 61
Commentary L Para. 76, Table L-
14, Para. 118
CSA CAN/CSA-O80 Series- Wood Preservation Commentary L Table L-12
M89
CSA O86.1-94 Engineering Design in Wood (Limit States Commentary A Table A-1
Design)
CSA CAN/CSA-S16.1-94 Limit States Design of Steel Structures Commentary A Table A-1
Commentary J Para. 87
CSA S136-94 Cold Formed Steel Structural Members Commentary D Table D-1
CSA CAN3 S304-M84 Masonry Design for Buildings Commentary B Para. 67
CSA CAN/CSA S304.1-94 Masonry Design for Buildings (Limit States Commentary B Para. 67
Design) Commentary D Table D-1
CSA S448.1-93 Repair of Reinforced Concrete in Buildings Commentary K Para. 48
CSA W59-M1989 Welded Steel Construction (Metal Arc Commentary L Para. 118
Welding)
CSA Z240.10.1-94 Site Preparation and Anchorage of Mobile Commentary B Para. 67
Homes

1995UGP3-V3 - User S - Guide - National - Building - Code - Part - 4 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1995UGP3-V3 - User S - Guide - National - Building - Code - Part - 4 PDF

Uploaded by

Copyright:

Available Formats

Copyright © NRC 1941 - 2019 World Rights Reserved © CNRC 1941-2019 Droits réservés pour tous pays

User’s Guide – NBC 1995

First Edition 1996

© National Research Council of Canada 1996

Aussi disponible en français :

Commentary A Serviceability Criteria for

Commentary B Wind Loads ........................... 9

Commentary C Structural Integrity.............43

Commentary D Effects of Deformations in

Commentary E Load Combinations.............49

Commentary F Limit States Design ............ 51

Commentary G Tributary Area.....................55

Commentary H Snow Loads ........................ 57

Commentary I Rain Loads .......................... 73

Commentary J Effects of Earthquakes.......75

Commentary K Application of NBC Part 4 for

Commentary L Foundations ...................... 103

Commentary M Structural Integrity of

Appendix A List of Referenced

D.E. Allen W.R. Schriever

Commentary J (Effects of Earthquakes) was pre-

Commentary K (Glass Design) has been deleted.

The new Commentary K (Application of NBC

CAN/CSA-O86.1-94 CAN3-A23.3-94 CAN/CSA-S16.1-94

Notes to Table A-1:

User’s Guide – NBC 1995 Structural Commentaries (Part 4) 1

Table A-1 (Continued)

(5) Includes a warning clause on vibrations.

Vibrations Floor Vibrations in Assembly

7. The undesirable effects of continuous

2 User’s Guide – NBC 1995 Structural Commentaries (Part 4)

Overloading and Fatigue dancing, concerts

people, the most critical range is 1.5 to 3 Hz. Typical

is large compared to the forcing frequency, f, the dy-

Activity Forcing Frequency f, Weight of Dynamic Load Dynamic Load wp,

Notes to Table A-2:

User’s Guide – NBC 1995 Structural Commentaries (Part 4) 3

For a floor with many people on it, the damping where

5 1 S = the elastic shortening of the column due

dynamic effect girders continuous over supports, the elastic deflec-

4 User’s Guide – NBC 1995 Structural Commentaries (Part 4)

structure is shared with a more sensitive occupancy,

ao/g = acceleration limit,

Acceleration Limit, 15. Table A-4 contains examples of the applica-

User’s Guide – NBC 1995 Structural Commentaries (Part 4) 5

Activity and Construction Forcing Frequency f, Effective Weight of

Notes to Table A-4:

References (6) A.W. Skempton and D.H. MacDonald, The

6 User’s Guide – NBC 1995 Structural Commentaries (Part 4)

(13) G. Pernica, Dynamic Live Loads at a Rock

tional Association for Bridge and Structural

User’s Guide – NBC 1995 Structural Commentaries (Part 4) 7

User’s Guide – NBC 1995 Structural Commentaries (Part 4) 9

10. The exposure factor, Ce, reflects changes in

10 User’s Guide – NBC 1995 Structural Commentaries (Part 4)

Ce(z) = the exposure factor over flat terrain

1Smax = the relative speed-up factor at the crest

17. Thirdly, Ce is used in the calculation of pres- 1

Notes to Table B-1:

User’s Guide – NBC 1995 Structural Commentaries (Part 4) 11

effect of the loading. The dynamic response in-

12 User’s Guide – NBC 1995 Structural Commentaries (Part 4)

Simple Procedure 22. Suggested values for must be based