Volume 63 Number 4

December 2021

QQ F low resistance in channels with large

emergent roughness elements
QQ  nderstanding the barriers and drivers of sustainable
U
construction adoption and implementation in South
Africa: A quantitative study using the Theoretical
Domains Framework and COM-B model
QQ  n investigation of South African low‑income
A
housing roof anchor systems
QQ  development cost comparison between a multi-storey mass
A
timber and reinforced concrete building in South Africa
QQ L oad spreading in ultra‑thin high-strength
steel‑fibre-reinforced concrete pavements
QQ  riority infrastructure for minibus-taxis: An analytical
P
model of potential benefits and impacts
Volume 63 No 4 December 2021 ISSN 1021-2019

Publisher
South African Institution of Civil Engineering
Block 19, Thornhill Office Park,
Bekker Street, Vorna Valley, Midrand, South Africa
Private Bag X200, Halfway House, 1685, South Africa
Tel +27 11 805 5947, Fax +27 11 805 5971
http://www.saice.org.za
civilinfo@saice.org.za

Contents
Editor-in-chief
Dr Peter Day
Consultant, Jones & Wagener (Pty) Ltd
Adjunct Professor of Geotechnical Engineering
University of Stellenbosch
Tel +27 11 519 0200
day@jaws.co.za
2 Flow resistance in channels with large
joint EditorS-in-chief
Prof Chris Clayton emergent roughness elements
University of Southampton
c.clayton@soton.ac.uk C S James
Prof Gerhard Heymann
University of Pretoria
gerhard.heymann@up.ac.za 11 Understanding the barriers and drivers of sustainable
Editor
Verelene de Koker construction adoption and implementation in
Tel +27 11 805 5947, Mobile +27 83 378 3996
verelene@saice.org.za
South Africa: A quantitative study using the
journal editorial Panel Theoretical Domains Framework and COM-B model
Prof Y Ballim – University of the Witwatersrand
Prof W Burdzik – Retired from the University of Pretoria R J Marsh, A C Brent, I H De Kock
Prof CRI Clayton – University of Southampton
Dr P Day – Jones & Wagener (Pty) Ltd (chairman)
Prof J du Plessis – University of Stellenbosch
Prof GC Fanourakis – University of Johannesburg
24 An investigation of South African low‑income
Prof PJ Gräbe – University of Pretoria housing roof anchor systems
Dr C Herold – Umfula Wempilo Consulting
Prof G Heymann – University of Pretoria R Van der Merwe, J Mahachi
Prof A Ilemobade – University of the Witwatersrand
Prof SW Jacobsz – University of Pretoria

35 A development cost comparison between

Prof EP Kearsley – University of Pretoria
Prof C Roth – University of Pretoria
Prof W Steyn – University of Pretoria
Mr M van Dijk – University of Pretoria a multi-storey mass timber and reinforced
Prof C Venter – University of Pretoria
Prof A Visser – University of Pretoria
concrete building in South Africa
Dr E Vorster – Zutari
S Van der Westhuyzen, J Wium
Prof J Wium – University of Stellenbosch
Prof A Zingoni – University of Cape Town
Prof M Zuidgeest – University of Cape Town
Peer reviewing
45 Load spreading in ultra‑thin high-strength
The Journal of the South African Institution of
Civil Engineering is a peer-reviewed journal
steel-fibre-reinforced concrete pavements
that is distributed internationally M S Smit, E P Kearsley
Design and reproduction
Marketing Support Services, Ashlea Gardens, Pretoria
Papers for consideration should be submitted 53 Priority infrastructure for minibus-taxis: An analytical
online at: http://journal.saice.org.za
model of potential benefits and impacts
The South African Institution of Civil Engineering accepts no
responsibility for any statement made or opinion expressed L R De Beer, C Venter
in this publication. Consequently, nobody connected with
the publication of this journal, in particular the proprietor,
the publisher and the editors, will be liable for any loss or
damage sustained by any reader as a result of his or her action
upon any statement or opinion published in this journal.

© South African Institution of Civil Engineering 1

TECHNICAL PAPER
Flow resistance in channels
with large emergent
Journal of the South African
Institution of Civil Engineering

roughness elements
ISSN 1021-2019
Vol 63 No 4, December 2021, Pages 2–10, Paper 1040

PROF CHRIS JAMES is Professor Emeritus in

the School of Civil and Environmental C S James
Engineering at the University of the
Witwatersrand (Wits), Johannesburg.
Before retiring, he lectured in hydrology and
hydraulics, and carried out research in river Reducing uncertainty in flow resistance estimation in natural channels requires elucidation
hydraulics, sedimentation, hydraulic
of contributing influences. Surface shear and form drag are the major contributors in
structures and environmental flows. At Wits
he was also a co-director of the interdisciplinary Centre for Water in the channels containing large roughness elements under emergent flow conditions. The two
Environment. effects can be accounted for in the Darcy-Weisbach and Manning equations by adding their
Contact details:
associated, separate friction factors or taking the square root of the sum of the squares of
School of Civil and Environmental Engineering the corresponding Manning coefficients. The friction factor for form drag can be estimated
University of the Witwatersrand from the drag coefficient and areal density of the roughness elements and the flow depth.
Private Bag 3
Predictions of the combined effect are tested against results of laboratory experiments with
Wits 2050
South Africa different arrangements of emergent cylinders on smooth and rough beds, using experimentally
T: +27 28 341 0060 determined drag coefficients for the cylinders. The variation of the overall resistance coefficient
E: chris.james@wits.ac.za with flow condition depends on the dominant influence, decreasing with flow depth when
Alternative postal address: surface shear dominates, and increasing with depth when form drag dominates.
PO Box 1140
Stanford 7210
South Africa
INTRODUCTION and form types. For natural rivers with
The flow resistance of a river is an expres- multiple contributions to overall resistance,
sion of the effects that its physical features the estimation of n is usually based on
have on flow depth and velocity character- direct site-specific calibration, experience
istics. Various sources of resistance related of similar situations, or synthesis from
to different physical features have been finer resolution descriptions of the differ-
recognised (e.g. Rouse 1965; Yen 2002), ent contributing effects.
the most influential being surface shear Estimation of n by experience is aided
associated with grain roughness and form by previous calibrations with physical
drag arising from flow separation around channel characteristics, as documented in
large roughness elements or irregularities. tables (e.g. Chow 1959), or presented with
The resistance effects of the underlying photographs for matching (e.g. Hicks &
sources can be described at different levels Mason 1991). Although common practice,
of resolution, depending on the purpose the use of tabulated values and photographs
of the application, the information avail- is unreliable because the value of n can vary
able, and the way in which finer resolution significantly with flow magnitude, and it
processes are modelled (James & Jordanova is difficult to select a representative value
2010). At any particular level, the effects of within the range reported, or to predict the
lower-level processes that are not described direction and rate of trend with flow depth
explicitly are accounted for through a or discharge. The comprehensive informa-
resistance coefficient. tion provided by Hicks and Mason (1991)
Many engineering and environmental illustrates this uncertainty more convinc-
river problems are appropriately addressed ingly than it provides confidence in selecting
at the river reach level, using 1D or 2D representative values. For more than a third
models that require descriptions of resist- of the documented reaches, the n values
ance to relate cross-section or cell average vary with flow depth by more than a factor
velocities to flow depths. Commonly of 2, approaching an order of magnitude in
used resistance equations are those of some cases. While n commonly decreases
Darcy-Weisbach (Equation 8), Chézy, and with flow depth, it increases distinctly in
Manning (Equation 13). The Manning some cases and even reverses in trend for
equation is the most widely used, with some. Association of the range and the
its resistance coefficient (n) providing a direction and rate of trend with the physical
Keywords: flow resistance, drag, large-scale roughness, lumped parameter to account for all con- characteristics of the river is not obvious,
low-flow hydraulics, river flow tributing influences, including both shear making estimation of a representative value

James CS. Flow resistance in channels with large emergent roughness elements.
2 J. S. Afr. Inst. Civ. Eng. 2021:63(4), Art. #1040, 9 pages. http://dx.doi.org/10.17159/2309-8775/2021/v63n4a1
for unmeasured sites very uncertain. Such
uncertainty prompted Ferguson (2010) to
question the usefulness of the Manning
equation, but it does provide a convenient
way for lumping the effects of multiple
resistance contributions, whether calibrated
directly or synthesised from explicit expres-
sions of underlying effects. Reducing the
uncertainty in the overall n value requires
explanation of its variability, which is
unlikely to be gained by observation at the
level of resolution of its application; some
elucidation of the processes at a higher level
of resolution is necessary.
Various synthesis approaches for
estimating n have been proposed. These
include (a) the use of composite roughness
equations for combining different local
resistance coefficients across a channel sec-
tion (such as the equations of Horton (1933)
and Pavlovski (1931)), (b) the SCS method
initially proposed by Cowan (1956) for aug-
menting a basic surface n value to account
for effects of surface irregularity, channel
shape variations, obstructions, vegetation
and meandering, and (c) application of a
lateral distribution model such as included
in the Conveyance Estimation System (CES)
software described by Knight et al (2010).
The composite roughness equations and
lateral distribution model really apply only
to transverse variations of surface shear
and flow depth across a section, although
certain formulations have been shown to
apply equally well to areal variations (James
& Jordanova 2010). The SCS method applies
over a channel reach and does allow for con-
sideration of form drag effects, but is flawed
in its linear addition of contributions and in
considering multiple effects to be independ-
ent (James & Jordanova 2010).
Synthesis by explicit combination of
surface shear and form resistance effects
has been used in various applications. For
alluvial resistance estimation, Einstein and
Barbarossa (1952) and many others since
have considered total resistance to be the
sum of surface grain and bedform contribu-
tions. The concept has also been applied to
other types of large-scale roughness in natu-
ral streams (e.g. David et al 2011; Shields &
Gippel 1995) and especially for emergent
vegetation (e.g. James et al 2004; James et
al 2008; Petryk & Bosmajian 1975; Stone
& Shen 2002). Applications to emergent
vegetation have generally been with stem
densities so high that the overall resistance
is almost exclusively form drag, although
Schoneboom et al (2010) have demonstrated
the contribution of bed shear, and James et
Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021
al (2004) quantified the error in the resist- The bed shear force is given by
ance coefficient associated with neglect of
the bed shear contribution. F/ = τoAbf(3)
In this work, synthesis of a combination
resistance coefficient (in terms of either in which Abf is the surface area on which
the Darcy-Weisbach f or the Manning n) the shear stress acts. (Note that Abf is less
is considered for a simple case of distinct than the area of the bed not occupied by
contributions from bed shear and form form roughness elements Ab, because the
drag, such as would occur in a stream com- shear stress is not effective within the sepa-
prising a gravel or cobble bed and dispersed ration zones downstream of each element.)
boulders. The results have immediate The shear stress τo can be expressed in
relevance for low-flow hydraulic analysis terms of the surface friction factor f / as
in environmental flow determinations and
in river rehabilitation design involving ρf /
τo = V 2(4)
the placement of boulders to create suit- 8
able habitat conditions. The resistance of
emergent boulders in a stream is similar in which V is the average velocity.
in nature to that of emergent vegetation The form drag can be quantified
stems, just at a larger scale, and the formu- through the drag equation
lation presented in the following section is
essentially similar to that presented by the 1
F// = CD Ap ρV 2(5)
authors cited above. A similar approach 2
has been followed by Guillén-Ludeňa et al
(2020) to account for the effect of buildings in which CD is the drag coefficient and Ap
on floodplains. It is shown further that the is the projected area of the form roughness
variation of drag coefficient with element elements in the flow direction.
Reynolds number for large elements is Inserting the force expressions in the force
significantly different from that for slender balance equation and rearranging terms gives
stems, and that the relative contributions an equation for the velocity in terms of f /, CD
of shear and form resistance have a marked and the channel characteristics, i.e.
influence on the trend of the overall resist-
ance coefficient with flow depth. 1
V= √2 g S(6)
f / Abf Ap
+ CD
4 Vol Vol
FORMULATION
The combination of surface shear and form Further manipulation of Equation 6 leads to
resistance for a stream with large emergent
roughness elements follows formulations for 8g Vol
V= S(7)
emergent vegetation stems, as presented by Ap Abf
f/ + 4CD
Petryk & Bosmajian (1975) and subsequent Abf
researchers. The development is presented
here to include minor modifications and which can be expressed as
clarify concepts. The relationship is derived
by considering the downslope weight com- 8g
V= √R S(8)
ponent of a volume of water in uniform flow f
to be balanced by the sum of the forces aris-
ing from bed shear and form drag, i.e. i.e. the conventional form of the Darcy-
Weisbach equation, with
F = F/ + F//(1)
f = f / + f //(9)
where F is the downslope weight com-
ponent of the water, F/ is the bed shear where f is the total friction factor and f // is
force and F// is the form drag. The weight the effective friction factor associated with
component is form resistance, given by
F = ρ g Vol S(2) Ap
f // = 4CD (10)
Abf
in which ρ is the water density, g is gravi-
tational acceleration, S is the channel slope and R is a volumetric hydraulic radius
and Vol is the volume of water. given by
3
 Vol Table 1 Experimental conditions
R=  (11)
Abf Flume width Element
Test Bed type Bed slope N
(m) density (%)
Expressing the bed shear in terms of the
A1 2.0 Smooth 0.00075 0 0
Manning resistance coefficient for the sur-
face (n / ) rather than the friction factor, i.e. A2 2.0 Smooth 0.00050 0.55 0.46

A3 2.0 Smooth 0.00050 1.05 0.87

ρ g n/2
τo = V2  (12) A4 2.0 Smooth 0.00050 1.59 1.54
R⅓
A5 2.0 Smooth 0.00050 3.19 3.00

similarly leads to the conventional Manning B1 2.0 Rough 0.00050 0 0

equation B2 2.0 Rough 0.00050 0.55 0.46

1 B3 2.0 Rough 0.00050 1.05 0.87

V= R⅔ S½(13)
n B4 2.0 Rough 0.00050 1.59 1.54

B5 2.0 Rough 0.00050 3.19 3.00

with
C1 1.0 Smooth 0.00058 0 0

n = √n/2 + n//2(14) C2 1.0 Smooth 0.00058 2.75 2.61

C3 1.0 Smooth 0.00058 6.08 5.78

where n// is the resistance coefficient
C4 1.0 Smooth 0.00058 8.50 8.08
accounting for the form resistance compo-
nent, given by C5 1.0 Smooth 0.00058 14.3 13.6

D1 1.0 Rough 0.00058 0 0

R⅓ Ap
n// = 4CD (15) D2 1.0 Rough 0.00058 2.92 2.77
8g Abf
D3 1.0 Rough 0.00058 5.83 5.54
The combined resistance of bed shear and D4 1.0 Rough 0.00058 8.25 7.84
form roughness presented by Equations
D5 1.0 Rough 0.00058 14.3 13.6
8 to 11 or 13 to 15 accounts for the two
contributions at a mutually consistent level
of resolution: bed shear is represented by a by the elements associated with each N Each setup was tested with five
friction factor which may be related to the value. The tests with no cylinders in place discharges (over the ranges indicated in
Nikuradse roughness (ks) of the bed materi- (N = 0) were carried out to establish bed Figure 4) with the cylinders emerging
al, and form drag is represented by a similar friction factor relationships for the two bed through the water surface in all cases. The
coefficient related to CD, which depends on types. Figure 1 shows the flume setup for discharge was measured by a turbine meter
the form element size, shape and spacing. Test B5. in the supply line and controlled by a valve

EXPERIMENTAL PROCEDURE
Experiments to test Equations 8 to 11 were
carried out in a 12.0 m long, 2.0 m wide
tilting flume. The form roughness elements
were 0.110 m diameter circular cylinders
placed in staggered arrangements with
equal longitudinal and transverse spacing.
Two sets of experiments were conducted,
one with the full width of the flume (Nkosi
2007) and the other with the width reduced
to 1.0 m, to enable a wider range of flow
conditions and element densities to be test-
ed. For each set, experiments were carried
out with different cylinder density arrange-
ments on both the original smooth vinyl
flume bed and on a rough bed made with a
single layer of crushed gravel with a nomi-
nal diameter of 19 mm. Stage-discharge
measurements were taken for each of the
test conditions listed in Table 1, which also
indicates the areal density of bed coverage Figure 1 Experimental setup for Test B5

4 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
 investigation (Red = Vd/ν, with ν being the
kinematic viscosity of water and d the cylin-
Pivot 320 mm
der diameter). As the drag force on cylinders
Pointer is known to depend strongly on size (Turcotte
et al 2015), a series of tests were carried out to
obtain estimates of CD for the cylinders used
in these experiments under emergent condi-
tions. Drag forces were measured using the
apparatus shown schematically in Figure 2,
comprising two lever arms connected at right
angles with a pivot at their vertex. A cylinder
Balance
was attached to the vertical arm, with its base
just clear of the flume bed, and a pointer on
the horizontal arm was placed in contact with
the pan of an electronic balance. Drag on the
840 mm cylinder produced a measurable force on the
Cylinder
balance which could be related to the drag
force by moment equilibrium. Measurements
were taken with a wide range of discharges
and weir settings for single cylinders in isola-
tion and in groups.

Flow RESULTS

Drag coefficients
Values of CD were calculated from the
measured forces by Equation 5, with V
Figure 2 Apparatus for measuring drag force on emergent cylinder taken as the cross-section average velocity
calculated from the measured discharge
and flow depth. In the moment equilibrium
10 calculations, it was assumed that the drag
Data Equation 16
force was applied at mid-depth (repeating
Standard Kothyari et al (2009)
Tanino & Nepf (2008)
the calculations with the force assumed to
8 be exerted at the bottom of the cylinder
and at the water surface showed a maxi-
Red = 18 000 mum difference in calculated CD within
6 about ±7%). The value of CD was found to
vary strongly with Red over the lower range
CD

tested (Figure 3). The relationships derived

4
for single, isolated cylinders are

–0.64
CD = 722Red for Red < 18 000
2 (16)
CD = 1.32 for Red > 18 000

with R2 = 0.90 for the lower Red equa-

0
10 100
Figure 3 Drag coefficients for emergent cylinders
in the supply line near the flume entrance.
Uniform flow was ensured by adjusting a
horizontal weir at the downstream end of
the flume. The height of the water surface
was measured at 1.10 m intervals along the
centre line of the flume using piezometers,
and the weir level was adjusted until the
water surface slope matched the bed slope.
The average of the flow depths at all the
1 000 10 000
Red
measurement locations was used as the
uniform flow depth.
Calculation of the form friction factor by
Equation 10 requires knowledge of CD. Values
of CD for emergent cylinders have previously
been determined (e.g. Kothyari et al 2009;
Tanino & Nepf 2008), but only for slender
vegetation stems and lower values of element
Reynolds number (Red) than obtained in this
tion. These results are consistent with
those obtained independently for similar
cylinders by Mahdi and Bismilla (2014),
and Jackson (2016). Also shown in Figure 3
is the “standard” curve for infinitely long
cylinders, as presented in many textbooks
(e.g. Albertson et al (1960), as well as the
upper and lower limits of the data obtained
by Tanino and Nepf (2008), and Kothyari
et al (2009) for ranges of slender cylinder
densities. No clear effect of cylinder density
was apparent in the results for the large
cylinders, and Equations 16 were used for
the following resistance predictions.

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 5
 0.20 0.20

0.15 0.15
A1 B1
Flow depth (m)

Flow depth (m)

A2 B2
0.10 A3 0.10 B3
A4 B4
A5 B5
0.05 0.05

0 0
0 0.02 0.04 0.06 0.08 0 0.02 0.04 0.06
(a) Discharge (m3/s) (b) Discharge (m3/s)

0.25 0.25

0.20 0.20
C1 D1
Flow depth (m)

Flow depth (m)

0.15 C2 0.15 D2
C3 D3
0.10 C4 0.10 D4
C5 D5
0.05 0.05

0 0
0 0.02 0.04 0.06 0 0.01 0.02
(c) Discharge (m3/s) (d) Discharge (m3/s)

Figure 4 Measured (markers) and predicted (lines) stage-discharge relationships

Surface friction factors
Tests A1, B1, C1 and D1 were carried out to
enable formulation of surface friction fac-
tor relationships for the two flume widths
and two surface types.
For the smooth beds (Tests A1 and
C1) the flow conditions were around the
threshold between hydraulically smooth
and transitional turbulent flow. The shear
Reynolds number (Re* = u* ks/ν, with the
shear velocity u* = (gRS)0.5) was between
3.7 and 5.2 for Tests A1 and between 7.3
and 8.1 for Tests C1. Friction factor relation-
ships were therefore obtained by calibrating
the laws presented by the ASCE Task Force
to give values of b = 6.17 for the 2.0 m
wide flume and b = 9.55 for the 1.0 m
wide flume. Using Equation 18 the default
values for a, b and c were retained, and
values of ks found as 0.22 mm for the 2.0 m
wide flume and 0.46 mm for the 1.0 m
wide flume.
The flow conditions for the rough
bed (Tests B1 and D1) were well within
the hydraulically rough range and the
resistance was characterised by a ks value
in the hydraulically rough flow equa-
tion presented by the ASCE Task Force
on Friction Factors in Open Channels
(ASCE 1963), i.e.
The best fit factor value for the 1.0 m wide
flume was found to be 0.16, and for consist-
ency this was used for the 2.0 m flume as
well. The resulting values of ks were found
to be 0.040 m for the 1.0 m flume and
0.049 m for the 2.0 m flume.
Friction factors may also be expressed
as equivalent Manning coefficients accord-
ing to
R¹⁄₆ √f
n= (20)
√8 g
Stage-discharge predictions

= c log ⎫⎪ ⎫⎪ (19)
on Friction Factors in Open Channels Equations 8 to 11 were tested by predicting
1
⎭ ks ⎭
(ASCE 1963) for both conditions, i.e. aR the stage-discharge relationships for the

= c log ⎫⎪
Re √f ⎫
√f / experimental conditions. For a channel

⎪ (17)
1 /

⎭ b ⎭
with rectangular cross-section and circu-
√f / with a = 12 and c = 2.0. The friction factor lar-cylindrical emergent form roughness
for the rough bed f b was calculated by the elements
for hydraulically smooth flow, and sidewall correction procedure proposed

= –c log ⎫⎪ s +
b ⎫
by Vanoni and Brooks (1957) and ks then Vol = Ab D(21)

⎪ (18)
1
⎭ aR Re √f / ⎭
k calculated from Equation 19. The value of
√f / ks depends on the height within the bed π d2
Ab = 1 – N (22)
from where the flow depth is measured. 4
for transitional flow. In these equations Estimates of the distance below the tops of
Re = 4VR/ν and the Task Force recommend bed grains to the theoretical bed, defining Ap = N Dd(23)
default coefficient values of a = 12, b = 2.51 the origin of the vertical velocity profile,
and c = 2.0. Using Equation 17 the value of range from 0.15 to 0.35 times the bed 2D π d2
Abf = 1 + –αN (24)
c was fixed at 2.0 and b was recalibrated particle size (as reviewed by James (1990)). W 4

6 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering

0.08
Predicted discharge (m3/s)
0.06
0.04
0.02
0
0 0.02 0.04
Measured discharge (m3/s)
A B C D
Figure 5 C
 omparison of measured and
predicted discharges for all tests
in which Ab is the area of the bed not occu-
pied by form roughness elements, D is the
flow depth, N is the number of elements per
unit area, d is the element diameter, W is
the channel width and α is a factor applied
to the element plan area to incorporate the
areas of separation over which bed shear is
ineffective. Thompson and Roberson (1976)
suggest values of α in the range 2.0 to 3.0.
Note that if α is assumed to be 1.0 (for a
negligible separation zone) and the channel
is wide (2D << W), then Abf ≈ Ab and R ≈ D.
Values of CD were calculated by
Equations 16, and the surface friction
factors by Equations 17 to 19 with the
calibrated coefficients (for the smooth
channels the results using Equations 17
and 18 were indistinguishable). For the
rough channels f / was determined by
calculating f b from Equation 19 and then
applying an inverse Vanoni and Brooks
(1957) sidewall correction procedure to
obtain a composite f / including the rough
0.20
0.15
Manning n
0.10
0.05
0 0
0 0.10
(a)
Figure 6 M
 easured (markers) and calculated (lines) variations of Manning resistance coefficient with flow depth for tests in the 1.0 m wide (a) and
2.0 m wide (b) rough bed flumes
Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021
bed and smooth side walls. Discharge
was calculated by multiplying the velocity
by the cross-sectional area adjusted to
account for the volume occupied by the
cylinders. Predictions were found to be
insensitive to the value of α, and a value of
2.5 (in the middle of the range suggested by
Thompson and Roberson (1976)) was used
throughout. The measured and predicted
results are presented in Figure 4. The
average absolute prediction errors for each
experimental series were 9.9% for Series A,
3.9% for Series B, 10.2% for Series C and
0.06 0.08 9.5% for Series D. Predicted and measured
discharges for all tests with cylinders in
Perfect place are compared in Figure 5. This shows
generally good agreement, except for the
smooth channel cases with the lowest ele-
ment densities.
DISCUSSION
The relative contributions of surface shear
and form drag to overall resistance depend
on the areal density of elements, the flow
depth and the roughness of the surface
between the elements. Under conditions
where bed shear dominates, i.e. where f //
(or n// ) is small in comparison with f / (or
n/ ), the average velocity increases with flow
depth and the resistance coefficient (either
f or n) decreases. Where form drag domi-
nates, the velocity is approximately constant
with flow depth (expanding Equation 7 with
f / = 0 shows V to be independent of D), and
the effective resistance coefficient therefore
increases with flow depth to compensate
R in the resistance equation. As shown in
Figure 6, the presence of form resistance in
a channel considerably increases the rate of
rise of flow depth with discharge.
Variations of the overall resistance
coefficient with flow depth are shown in
0.06
D1 0.04
Manning n
D2
D3
D4
D5 0.02
0.20 0 0.10
Flow depth (m) (b)
Figures 6(a) and 6(b) for the 1.0 m and 2.0
m wide flumes with rough beds. (Manning’s
n, as the more common representation
for overall resistance, is plotted here; the
relationship is similar in terms of f.) Shown
with the data are curves of n as predicted
using Equation 14 with n // from Equation 15
and n / from Equations 19 and 20, and the
inverted Vanoni and Brooks (1957) sidewall
correction procedure. The average absolute
deviation predicted from experimental
n values was 0.0064 for test series D and
0.0013 for test series B. The form resistance
component was dominant in the 1.0 m wide
rough-bed flume and the resistance coef-
ficient consequently increased with depth
(Figure 6(a)). (Form resistance also domi-
nated in all the smooth bed experiments,
resulting in similar trends.) For the 2.0 m
wide rough-bed flume, the overall resistance
coefficient represents a varying balance of
the bed shear and form drag contributions
(Figure 6(b)). For the relatively sparse ele-
ment arrangements in the 2.0 m wide flume
with the rough bed (Tests B2 and B3) the
bed surface contribution is considerable, and
the overall resistance coefficient reduces
with flow depth. As the element density
increases the form roughness contribution
increases, eventually producing a consist-
ent increase of resistance coefficient with
depth (Test B5). For an intermediate density
(Test B4) the resistance coefficient shows
a distinct dip in the mid-range of depths,
with bed shear dominating at low depths
and form drag dominating at high depths.
Figure 7 shows the variations of the pre-
dicted n/ and n// values with flow depth for
Test B4, and their combination compared
with the measured values. Similar variations
and trends of resistance coefficient may be
expected in other situations in natural chan-
nels where form roughness is significant,
B1
B2
B3
B4
B5
0.20
Flow depth (m)
7
such as associated with woody debris, bank
irregularities or large bank vegetation, as
well as buildings on floodplains.
These results show that uncertainty
in estimating resistance coefficients, and
especially their variation with discharge,
can be reduced by considering surface
shear and form resistance characteristics
separately, rather than estimating overall
resistance directly. Practical estima-
tion of an overall resistance coefficient
therefore requires both contributions to
be evaluated.
The friction factor of the bed surface
between form roughness elements ( f / ) can
be related to its grain size characteristics in
terms of a ks value. The various approaches
available are well reviewed by Ferguson
(2007), and include the forms of Equations
17 to 19. For these, various relationships
have been proposed for ks in terms of per-
centile bed particle sizes (e.g. Millar 1999),
and a value of 3.5d84 has become widely
accepted (James 2010) (d84 is the size of
bed particle for which 84% of particles
are smaller).
For description at a consistent level of
resolution, the friction factor of the form
roughness elements f // should be estimated
through Equation 10. For field calibration it
would be easier to evaluate f // (= 4CDAp/Abf )
directly rather than attempting to quantify
CD, Ap and Abf separately. For preliminary
estimates without calibration, it is noted
that the term Ap/Abf is linearly related to
the relative flow depth D/d, with the coef-
ficient depending on the areal density of the
emergent elements. Assuming a value for
α = 2.5 and a circular cylindrical shape for
the elements, this term can be approximated
for densities less than about 20% by
Ap D high Red.
= 0.014C1.1 (25)
Abf d For areal densities below about 20% a
in which C is the per cent areal density.
The value of CD varies with Red
(Figure 3) and does so differently for circu-
lar cylinders with different diameters. This
variation has yet to be described reliably for
all sizes, but its approach to a constant value
for Red greater than about 18 000 can be
reasonably accepted. This limit represents
rather low flow velocities, for example about
0.18 m/s for a cylinder diameter of 0.10 m
and 0.018 m/s for a diameter of 1.0 m. CD
also depends on the element shape. Jackson
(2016) measured CD for 0.11 m diameter
circular and 0.10 m width square cylinders
placed square and diagonally to the flow
8 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
0.050
0.045
Manning n
0.040
0.035
0.030
0.025
0.020
0.015
0.010
0 0.05
Flow depth (m)
n'
Figure 7 Variation of resistance coefficients with depth for Test B4
direction. For Red > ~10 000 he measured
CD values of about 1.3, 2.0 and 3.0 for the
circular, square and diagonal cylinders
respectively, which gives an indication of
the influence of element angularity. For
Red < ~10 000 he found similar variation
with Red as shown by the data in Figure 3.
Guillén-Ludeňa et al (2020) measured CD
values for 54 mm square cylinders for Red
in the range 3 500 – 24 000 and areal densi-
ties between 0.2% and 20%. They propose a
single average value of 1.32 as representative
for all these conditions. CD has also been
found to vary with element density, espe-
cially with slender elements at low Red (e.g.
Kothyari et al 2009; Tanino & Nepf 2008).
The effect was insignificant for the present
experimental conditions, but requires
further investigation for large elements and
reasonable estimate of f // can therefore
be obtained from Equation 10 with Ap/‌Abf
given by Equation 25 and, for Red > ~18 000,
a value of CD varying from about 1.3 for
smooth rocks up to about 3 for very angular
rocks. For Red < ~18 000 the value should
be increased according to the variation
indicated by Equation 16.
Both f / and f // can be converted to
Manning n values using the equivalence
relationship, Equation 20.
CONCLUSIONS
The variation of resistance coefficient with
flow magnitude in channels with rough
0.10 0.15 0.20
n'' n B4
beds and large form roughness elements can
be explained by elucidating their respective
contributions. The overall resistance coef-
ficient can be obtained by combining inde-
pendently determined bed shear and form
resistance contributions. The overall Darcy-
Weisbach friction factor f is the sum of the
component friction factors (Equation 9); the
overall Manning resistance coefficient n is
the square root of the sum of the squares of
the component values (Equation 14).
The friction factor f / or resistance
coefficient n/ for bed shear depends on the
roughness size ks of the bed material. The
friction factor f // or resistance coefficient
n// for form roughness depends on the ele-
ment drag coefficient CD, the areal density
of the elements and the flow depth, accord-
ing to Equations 10 and 15 respectively.
The variation of CD with Red for large
emergent objects is significantly different
from that for slender cylinders (as repre-
senting vegetation stems, for example).
Under very low-flow conditions the drag
coefficient can be expected to vary consid-
erably with Red. For the areal densities of
emergent elements investigated (less than
about 15%) CD was found to be unaffected
by the density; the effect for greater densi-
ties requires further investigation.
The variation of overall resistance coef-
ficient with flow depth depends on the rela-
tive dominance of the two contributions. If
bed shear dominates (e.g. with a very rough
bed and/or sparse form element distribu-
tion) f and n decrease with flow depth; if
form drag dominates (e.g. with high form
element density and/or relatively smooth ν = Kinematic viscosity of water James, C S & Jordanova, A A 2010. Reach-scale
bed) f and n increase with flow depth. For ρ = Water density resistance of distributed roughness in channels. In
intermediate form element densities, the τo = Boundary shear stress Dittrich, A, Koll, K, Aberle, J & Geisenhainer, P (Eds),
trend of resistance coefficient can reverse Proceedings, International Conference on Fluvial
as flow depth increases. Hydraulics (River Flow 2010), 8‒10 September 2010,
REFERENCES Braunschweig, Germany, Vol 1, pp 333‒341.
Albertson, M L, Barton, J R & Simons, D B 1960. Fluid Knight, D W, McGahey, C, Lamb, R & Samuels, P G
ACKNOWLEDGEMENTS Mechanics for Engineers. Englewood Cliffs, NJ: 2010. Practical Channel Hydraulics: Roughness,
The stage-discharge relationships in Prentice Hall. Conveyance and Afflux. London: CRC Press/
the 1.0 m wide flume were measured ASCE (American Society of Civil Engineers) 1963. Taylor & Francis.
by Darlington Ikegwuoha. The work is Friction factors in open channels: Progress Report Kothyari, U C, Hayashi, K & Hashimoto, H 2009.
based on research funded by the National of the Task Force on Friction Factors in Open Drag coefficients of unsubmerged rigid vegetation
Research Foundation, South Africa. Channels. Journal of the Hydraulics Division, ASCE, stems in open channel flows. Journal of Hydraulic
89(2): 97‒143. Research, 47(6): 691‒699.
Chow, V T 1959. Open Channel Hydraulics. New York: Madhi, Y & Bismilla, B A 2014. Drag force on emergent
NOTATION McGraw-Hill. broad circular cylinders. Final Year Investigational
Ab = Bed area unoccupied by elements Cowan, W L 1956. Estimating hydraulic roughness Project Report. Johannesburg: School of Civil &
Abf = Bed area subjected to surface shear coefficients. Agricultural Engineering, 37(7): 473‒475. Environmental Engineering, University of the
Ap = Projected area of elements David, G C L, Wohl, E, Yochum, S E & Bledsoe, B P Witwatersrand.
a = Constant in friction factor 2011. Comparative analysis of bed resistance Millar, R G 1999. Grain and form roughness in
relationship partitioning in high-gradient streams. Water gravel-bed rivers. Journal of Hydraulic Research,
b = Constant in friction factor Resources Research, 47: W07507. 37(3): 303‒312.
relationship Einstein, H A & Barbarossa, N L 1952. River channel Nkosi, M M 2007. Flow resistance effect of large-scale
C = Per cent areal density of emergent roughness. Transactions, ASCE, 117: 1121‒1143. roughness elements. Final Year Investigational
elements Ferguson, R 2007. Flow resistance equations for gravel- Project Report. Johannesburg: School of Civil &
CD = Drag coefficient and boulder-bed streams. Water Resources Research, Environmental Engineering, University of the
c = Constant in friction factor 43: W05427. Witwatersrand.
relationship Ferguson, R 2010. Time to abandon the Manning Pavlovski, N N 1931. On a design formula for uniform
D = Flow depth equation? Earth Surface Processes and Landforms, flow movement in channels with non-homogeneous
d = Cylinder diameter 35: 1873‒1876. walls (in Russian). Transactions of All-Union
F = Downslope weight component of Guillén-Ludeňa, S, Lopez, D, Mignot, E & Riviere, Scientific Research of Hydraulic Engineering,
water N 2020. Flow resistance for a varying density of Leningrad, 3: 157‒164.
F/ = Surface shear resisting force obstacles on smooth and rough beds. Journal of Petryk, S & Bosmajian, G 1975. Analysis of flow
F// = Form drag Hydraulic Engineering, 146(2): 04019059. through vegetation. Journal of the Hydraulics
f = Friction factor Hicks, D M & Mason, P D 1991. Roughness Division, ASCE, 101(HY7): 871‒884.
f/ = Friction factor associated with bed characteristics of New Zealand rivers. Wellington, Rouse, H 1965. Critical analysis of open-channel
shear New Zealand: Water Resources Survey. resistance. Journal of the Hydraulics Division, ASCE,
f // = Friction factor associated with form Horton, R E 1933. Separate roughness coefficients for 91(HY4): 1‒25.
drag channel bottom and sides. Engineering News Record, Schoneboom, T, Aberle, J & Dittrich, A 2010. Hydraulic
fb = Friction factor for bed 3(22): 652‒653. resistance of vegetated flows: Contribution of bed
g = Gravitational acceleration Jackson, K S 2016. The determination of form drag shear stress and vegetative drag to total hydraulic
ks = Nikuradse grain roughness coefficient for rigid, emergent objects in open channel resistance. In Dittrich, A, Koll, K, Aberle, J &
n = Manning resistance coefficient flow. MSc (Eng) Project Report. Johannesburg: Geisenhainer, P (Eds), Proceedings, International
n/ = Manning resistance coefficient asso- School of Civil & Environmental Engineering, Conference on Fluvial Hydraulics (River Flow 2010),
ciated with bed shear University of the Witwatersrand. 8‒10 September 2010, Braunschweig, Germany,
n// = Manning resistance coefficient asso- James, C S 1990. Prediction of entrainment conditions Vol 1, pp 269‒276.
ciated with form drag for nonuniform, noncohesive sediments. Journal of Shields, F D & Gippel, C J 1995. Prediction of effects of
N = Number of elements per unit area of Hydraulic Research, 28(1): 25‒41. woody debris removal on flow resistance. Journal of
bed James, C S 2010. Flow resistance in rivers, Chapter 7. Hydraulic Engineering, 121(4): 341‒354.
R = Hydraulic radius In James, C S & King, J M (Eds). Ecohydraulics for Stone, B M & Shen, H T 2002. Hydraulic resistance of
Re = Flow Reynolds number South African Rivers – A Review and Guide. WRC flow in channels with cylindrical roughness. Journal
Red = Element Reynolds number Report No. TT 453/10. Pretoria: Water Research of Hydraulic Engineering, 128(5): 500‒506.
Re* = Shear Reynolds number Commission. Tanino, Y & Nepf, H M 2008. Laboratory investigation
S = Channel slope James, C S, Birkhead, A L, Jordanova, A A & O’Sullivan, of mean drag in a random array of rigid, emergent
u* = Shear velocity J 2004. Flow resistance of emergent vegetation. cylinders. Journal of Hydraulic Engineering,
V = Average velocity Journal of Hydraulic Research, 42(4): 390‒398. 134(1): 34‒41.
Vol = Volume of water above bed James, C S, Goldbeck, U K, Patini, A & Jordanova, A A Thompson, G T & Roberson, J A 1976. A theory of
W = Channel width 2008. Influence of foliage on flow resistance of flow resistance for vegetated channels. Transactions
α = Factor to account for area of separa- emergent vegetation. Journal of Hydraulic Research, of the American Society of Agricultural Engineers,
tion zone 46(4): 536‒542. 19(2): 288‒293.

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 9
Turcotte, B, Millar, R G & Hassan, M A 2015. Drag Vanoni, V A & Brooks, N H 1957. Laboratory studies Yen, B C 2002. Open channel flow resistance. Journal
forces on large cylinders. River Research and of the roughness and suspended load of alluvial of Hydraulic Engineering, 128(1): 20‒39.
Applications, 32(3): 411‒417. streams. Sedimentation Laboratory Report No. E68,
Pasadena, CA: California Institute of Technology.

10 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Understanding the barriers TECHNICAL PAPER
and drivers of sustainable Journal of the South African
Institution of Civil Engineering

construction adoption and ISSN 1021-2019

Vol 63 No 4, December 2021, Pages 11–23, Paper 1202

implementation in South RALMAR MARSH (CEng MIEI) is a registered

Africa: A quantitative study

Chartered Engineer with Engineers Ireland.
She completed her BEng in Civil Engineering
in 2013 and her MEng in Engineering

using the Theoretical Domains

Management (Sustainable Systems) in 2020
at Stellenbosch University, and is currently
employed as a Structural Engineer at one of

Framework and COM-B model

Ireland’s leading engineering consultancies.
Her main fields of interest include sustainability in the construction
industry and contributing to the digital transformation of the design,
construction and management processes of the built environment.

Contact details:
R J Marsh, A C Brent, I H De Kock Department of Industrial Engineering
Faculty of Engineering
Stellenbosch University
Stellenbosch 7600
The construction industry is one of the largest consumers of natural resources. Improving the South Africa
sustainability of construction industry activities is therefore key to mitigating the negative T: +27 21 808 4234 / +353 89 486 2420
E: ralmar119@gmail.com
impact of the industry on the environment. Given the extent of the environmental challenges
faced by many countries, the transition towards the adoption of sustainable alternatives in the
PROF ALAN BRENT (Pr Eng) is registered
construction industry must include dimensions of changing human behaviour. These dimensions with the Engineering Council of South
include influencing the capability, opportunity, and motivation to adopt the desired change in Africa, and is also a Chartered Member of
behaviour. In order to improve the adoption and implementation of sustainable practices within Engineering New Zealand. He is a professor
of, and the inaugural holder of the chair in,
the construction industry, the behaviour change processes of stakeholders need to be considered.
Sustainable Energy Systems in the Faculty of
This study describes how the Capability, Opportunity, Motivation-Behaviour (COM-B) model Engineering at Victoria University of
and Theoretical Domains Framework (TDF) were used to identify the barriers to and drivers of Wellington, and an extraordinary professor
sustainable construction practices by construction industry stakeholders. The study included of Engineering Management and Sustainable Systems in the Department
of Industrial Engineering at Stellenbosch University.
a structured questionnaire survey completed by 108 construction industry stakeholders and
indicated a need to improve the capability, opportunity, and motivation amongst construction Contact details:
Department of Industrial Engineering
industry stakeholders to facilitate the adoption of sustainable construction practices. The Faculty of Engineering
questionnaire identified that an increase in the awareness, knowledge, interest, and demand for Stellenbosch University
sustainable construction will facilitate the adoption thereof. Additionally, providing training and Stellenbosch 7600
South Africa
access to education on best practices for sustainability can positively influence the behaviour of
T: +27 21 808 4234 / +64 4 463 5960
stakeholders and improve their confidence in implementing sustainable construction practices. E: acb@sun.ac.za
Economic factors such as the cost of implementing sustainable solutions and the perception
of the economic and social benefits of sustainable construction were identified as the critical DR IMKE DE KOCK (Pr Eng, MSAIIE) is a
barriers. These barriers and drivers are mapped to five TDF domains (knowledge, skills, social professionally registered industrial engineer
with international experience in the
influences, beliefs about capabilities, and beliefs about consequences), which can be targeted for
execution and management of industrial
behaviour change amongst construction industry stakeholders in future interventions. engineering and engineering management
projects. In 2015 she joined the academic
staff at Stellenbosch University, where she is
a senior lecturer in the Department of
Introduction aims to simultaneously pursue environ-
Industrial Engineering and part-time at the University of Stellenbosch
Over the past few decades, international mental, economic, and social dimensions Business School.
and local governments have recognised of development. However, various authors
Contact details:
that the construction industry is a key argue that sustainable development can be Department of Industrial Engineering
role-player in the adoption and implemen- interpreted and pursued within a specific Stellenbosch University
tation of sustainable development. The context, and therefore the objectives and Stellenbosch 7600
South Africa
construction industry, which includes the development solutions vary (Amui et al T: +27 21 883 9533 / +27 21 808 4234
entire life cycle of a building project, from 2017; Bebbington 2001; Du Plessis et al E: imkedk@sun.ac.za
pre-feasibility to decommissioning and 2002; Elliot 2006; Hjorth & Bagheri 2006).
disposal, has a significant impact on the The construction industry has a unique
environment, economy and society (Kibert position to create value by improving
2007). Since the initial development of the the quality of life of society and thereby
sustainability agenda in the late 1980s, positively contributing to sustainable Keywords: sustainable construction, barriers, drivers,
the principle of sustainable development development. Although the conventional behaviour change, quantitative, South Africa

Marsh RJ, Brent AC, De Kock IH. Understanding the barriers and drivers of sustainable construction adoption and implementation in South Africa: A quantitative study
using the Theoretical Domains Framework and COM-B model. J. S. Afr. Inst. Civ. Eng. 2021:63(4), Art. #1202, 13 pages. http://dx.doi.org/10.17159/2309-8775/2021/v63n4a2 11
approach to the built environment is driven
by cost, quality and performance criteria,
sustainable design and construction aims
to additionally minimise resource deple-
tion, minimise environmental degradation
and create a healthy built environment
(Kibert 2016). Sustainable construction
(SC), although lacking a general defini-
tion and therefore lacking a universal
approach (Ahn et al 2013; Du Plessis et al
2002; Ofori 2015), has been highlighted
in documents such as the Agenda 21 for
Sustainable Construction published by the
International Council for Research and
Innovation in Building and Construction
(CIB). The purpose of this document was
to provide guidance for all national agendas
on a global scale to implement SC research
and development strategies to overcome
issues and challenges of sustainable devel-
opment within the construction industry
(Du Plessis et al 2002). Responses to the SC
agenda have, however, varied amongst the
different contexts of countries.
Although SC has been studied by
authors in both developed and develop-
ing countries who recognise the need to
develop and implement SC agendas (Ahn
et al 2013; AlSanad 2015; Brennan &
Cotgrave 2014; Chan et al 2017a; Darko
et al 2017a; Häkkinen & Belloni 2011;
Ismail et al 2012; Marchman & Clarke
2011; Saleh & Alalouch 2015), only a
few researchers (Aigbavboa et al 2017;
Du Plessis 2007) have conducted studies
which extend the concept of SC to South
Africa. In 2002, Du Plessis prepared a dis-
cussion document named the Agenda 21
for Sustainable Construction in Developing
Countries (A21 SCDC), which sought to
present a sector-and-developing-country
response to the challenge of sustainable
development. It also aimed to highlight
the implications of sustainability in the
built environment in a regional context
and to suggest the principles and strategy
for action for the developing world in
partnership with the developed world.
The objectives of this agenda aimed to
provide a framework that could be used
to guide research and development (R&D)
in developing countries and to promote
the exchange of learning of SC within the
developing world. Du Plessis (2007) argues
that developing countries will become the
platform for infrastructure and industrial
development, and therefore the construc-
tion sector in these countries has a critical
role to play in responding to the sustain-
able agenda.
12
Although numerous strategies were fur-
ther developed, as stated by Dalal-Clayton
and Bass (2012), there has been very little
impact of such strategies as a result of lack
of integration between frameworks and
key decision-makers, limited links between
policies and practical applications, a nar-
row base of participation from multiple
stakeholders, and many strategies lacking
clear and concise priorities and achievable
targets (Du Plessis 2007). Another study
conducted by Sebake (2008), emphasised
the challenges faced by professionals in
the implementation of sustainability prin-
ciples during the development of building
projects. The study recognised that any
strategies related to sustainable building
projects must be dealt with at the initial
stages of a project and must include both
architects and the rest of the professional
team (such as engineers and contractors).
In doing so, the client’s brief can be formu-
lated and developed by taking advantage
of all the participants’ competencies and
prerequisites, thereby ensuring a sustain-
able build and minimising silo design
and development and clashes later in the
project. The thematic analysis employed
by Windapo (2014) found that green build-
ing is at an early stage of development in
South Africa, and that rising energy costs
and the Green Building Council of South
Africa’s (GBCSA) Green Star rating system
were the primary drivers of green building.
Windapo (2014) noted that these drivers
have not changed since 2006, nearly a
decade since the publication of the article,
and highlighted that the drivers of green
building were mainly financial and not
necessarily to protect the environment.
Evidence of this is substantiated by the
case studies, which suggest that competi-
tive advantage and corporate image were
key drivers to achieving Green Star status.
Therefore, the author concludes that
without the economic benefit that green
buildings provide to clients and developers,
and with a lack of government regulations
to support green building construction, a
limited number of clients and developers
would drive green building projects for-
ward for the sole purpose of environmental
sustainability. An empirical study done by
Aigbavboa et al (2017) on the “lazy view”
of construction professionals in the South
African construction industry suggests that
the following reasons were the key chal-
lenges facing the construction industry in
a bid to achieve sustainable construction:
additional cost of building sustainable
Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
buildings, a limited understanding of the
benefits of sustainable construction, and
lastly, the lack of mobilisation of sustain-
able construction resources to support
technological changes in the industry.
Although there has been a shift
towards adopting a green approach
towards buildings since the establishment
of the Green Building Council of South
Africa in 2007, professional stakeholders
in the built environment have been reluc-
tant to pursue greening of buildings in
their design and construction methodolo-
gies. In a study conducted by Coetzee and
Brent (2015) it was found that the percep-
tion amongst key stakeholders in the built
environment was that the additional costs
associated with the design of sustainable
buildings are too high. The results of
the study, however, indicated that the
perceived cost of sustainable design and
construction was more than double what
the actual cost was. This highlights one of
the barriers towards effectively adopting
and implementing SC. In addition to this,
the lack of regulations and policies which
govern SC in many developing countries
has made it less likely for stakeholders
in the industry to adopt SC practices
(Aghimien et al 2018; Djokoto et al 2014;
Foong et al 2017; Suliman & Omran 2009).
Although there are regulations and frame-
works available in South Africa, such as
the Green Economy Accord, National
Greening Framework and the Green
Building Council, the latter which certifies
green buildings, Simpeh and Smallwood
(2015) suggest that the adoption of sus-
tainable practices in the South African
construction industry has been slow due
to conflicting regulations and stakehold-
ers who fear the liability and litigation
when it comes to the performance of new
sustainable products and systems. The
current legislation is generally voluntary
and prescriptive, which further reduces
the transition towards a more sustainable
approach to the multiple phases of the
building life cycle (Ametepey et al 2015;
Gan et al 2015; Mousa 2015). Assessing
the current awareness amongst built
environment stakeholders along with
their perceptions is critical to identify
and understand the barriers towards a
more sustainable built environment in
the South African context. By identifying
these barriers, it allows key decision-
makers to look at possible solutions to
overcome these barriers and work towards
driving the SC agenda forward.
Behaviour changes for sustainable
construction adoption
On a regular basis, individuals, organisa-
tions, key stakeholders, and policymakers
make decisions that have an impact on
the earth’s natural resources. Policies and
intervention strategies that focus on transi-
tioning these everyday behaviours towards
more sustainable outcomes are imperative
to achieving development and consumption
that are more sustainable (UNEP 2017).
However, changing human behaviour
can be challenging (Cane et al 2012), as
we do not always make rational decisions
or behave in predictable ways. (Cane et
al 2012) suggest that, although changing
behaviour is not easy, it can be more effec-
tive if interventions are implemented that
are based on evidence-based principles of
behaviour change.

Theoretical Domains Framework

To change professional practice, it has
been agreed that there is no “magic bullet”
(Oxman et al 1995) and that the effective-
ness of intervention strategies is sensitive
to context (Wensing et al 1998). Michie et
al (2005), however, suggest that the mixed
results and limited practical value of imple-
mentation research were due to a limited
theoretical underpinning for the develop-
ment of interventions. The authors argue
that a consensus of psychological theories
is necessary to provide clarity and simplify
the accessibility and usefulness of theories
in behavioural change research. A team
of behavioural scientists in collaboration
with implementation researchers there-
fore developed the Theoretical Domains
Framework (TDF) to provide access to
a theoretical basis for implementation
research. The TDF is an integrated frame-
work grounded in psychological theory
which synthesises 128 theoretical con-
structs from 33 theories relevant to imple-
mentation (Michie et al 2005). The original
TDF contained 12 domains, and after
validation through a consensus conducted
by Cane et al (2012), 14 domains covering
84 theoretical constructs were identified, as
illustrated in Table 1 (see page 14), which
is adopted in this research study. The TDF
domains can be condensed into three core
components of behaviour, namely capabil-
ity, opportunity and motivation (Michie
et al 2014) (Figure 1). The COM-B model
demonstrates that behaviour (B) is a result
of the interaction between psychological
and physical capabilities (C), to utilise
social and physical opportunities (O)
Soc Social influences
Env Environmental context and resources
Id Social/professional role and identity
Bel Cap Beliefs about capabilities
Opt Optimism
Sources of behaviour Int Intentions
Goals Goals
Bel Cons Beliefs about consequences
Reinf Reinforcement
Em Emotion
Know Knowledge
Cog Cognitive and interpersonal skills
TDF domains
Mem Memory, attention and decision processes
Beh Reg Behavioural regulation
Phys Physical skills
Figure 1 TDF domains linked to COM-B components (Michie et al 2014)
through reflective (self-conscious inten- application of utilising the TDF and COM-B
tions) and automatic (personal needs and model in research studies has mostly been
desires) motivators (M). adopted in literature to support research
Globally, government authorities, inter- in the medical and implementation science
governmental organisations, educational field (Atkins et al 2017) and has yet to be
institutions, businesses and organisations applied to understanding and targeting
have begun to adopt behavioural science change of behaviour amongst construction
theory and methodologies to design and industry stakeholders. However, it has been
implement effective behaviour change adopted in studies about behaviour change
policies and programmes (Klaniecki et al relating to sustainable consumption and
2018). Most behaviour change interventions sustainable development. An example of this
have primarily been applied in developed is a case study conducted by Gainforth et al
countries with high per-capita consumption (2016), which adopts the TDF and Behaviour
rates. However, developing countries present Change Wheel (Michie et al 2014) to
many opportunities to adopt behavioural develop interventions to change recycling
science to inform policy to align with sus- behaviours. Additionally, there have been
tainable development goals due to rapid examples of successful applications of
growth of consumption (UNEP 2017). The behavioural design to priority consumption

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 13
Table 1 Theoretical Domains Framework (Cane et al 2012)
Domain (definition1) Theoretical construct Domain (definition1) Theoretical construct
QQ Knowledge (including knowledge of QQ Goals (distal/proximal)
Knowledge
condition / scientific rationale) Goals QQ Goal priority
An awareness of the existence QQ Procedural knowledge Mental representations of QQ Goal/target setting
of something QQ Knowledge of task environment outcomes or end states that QQ Goals (autonomous/controlled)
an individual wants to achieve QQ Action planning
QQ Skills QQ Implementation intention
QQ Skills development
Skills QQ Competence Memory, attention and
An ability or proficiency QQ Ability decision processes QQ Memory
acquired through practice QQ Interpersonal skills The ability to retain QQ Attention
QQ Practice information, focus selectively QQ Attention control
QQ Skill assessment on aspects of the environment QQ Decision-making
QQ Professional identity and choose between two or QQ Cognitive overload/tiredness
QQ Professional role more alternatives
Social/professional role and QQ Social identity Environmental context and
identity QQ Identity resources
A coherent set of behaviours Environmental stressors
QQ Professional boundaries Any circumstance of a person’s
QQ

and displayed personal Resources / material resources

QQ Professional confidence situation or environment that
QQ

qualities of an individual in a Organisational culture/climate

QQ Group identity discourages or encourages
QQ

social or work setting Salient events / critical incidents

QQ Leadership the development of skills and
QQ

QQ Organisational commitment QQ Person X environment interaction

abilities, independence, social QQ Barriers and facilitators
QQ Self-confidence competence and adaptive
Beliefs about capabilities QQ Perceived competence behaviour
Acceptance of the truth, QQ Self-efficacy QQ Social pressure
reality, or validity about an QQ Perceived behavioural control QQ Social norms
ability, talent or facility that a QQ Beliefs QQ Group conformity
person can put to constructive QQ Self-esteem Social influences QQ Social comparisons
use QQ Empowerment Those interpersonal processes QQ Group norms
QQ Professional confidence that can cause individuals QQ Social support
Optimism QQ Optimism to change their thoughts, QQ Power
The confidence that things will QQ Pessimism feelings or behaviours QQ Intergroup conflict
happen for the best or that QQ Unrealistic optimism
QQ Alienation
desired goals will be attained QQ Identity
QQ Group identity
QQ Modelling
QQ Outcome expectancies
Beliefs about consequences Emotion
QQ Characteristics of outcome QQ Fear
Acceptance of the truth, reality A complex reaction pattern,
expectancies QQ Anxiety
or validity about outcomes involving experiential,
QQ Beliefs QQ Affect
of a behaviour in a given behavioural and physiological
QQ Anticipated regret QQ Stress
situation elements, by which the
QQ Consequents QQ Depression
individual attempts to deal QQ Positive / negative effect
QQ Rewards (proximal/distal, valued / not with a personally significant
Reinforcement QQ Burn-out
valued, probable/improbable) matter or event
Increasing the probability of QQ Incentives
a response by arranging a QQ Punishment
dependent relationship, or Behavioural regulation
QQ Consequents Self-monitoring
contingency, between the Anything aimed at managing
QQ
QQ Reinforcement Breaking habit
response and a given stimulus or changing objectively
QQ
QQ Contingencies and sanctions QQ Action planning
observed or measured actions
Intentions QQ Stability of intentions
A conscious decision to QQ Stages of change model
perform a behaviour or a 1 All definitions are based on definitions from the American
QQ Transition model/stages of change
resolve to act in a certain way Psychological Associations’ Dictionary of Psychology (VandenBos 2007).

areas: energy, water, transportation and
mobility, food and diet, and waste and dis-
posal (UNEP 2017).
Objective of this paper
There is a need to change the current
behaviour of stakeholders in the construc-
tion industry to ensure the sustainability
of the industry by adopting sustainable
alternatives throughout the life cycle
of construction projects. Moreover, the
impact of the construction industry on the
environment needs to be mitigated. The
aim of this study was to use the COM-B
model and TDF to identify barriers and
drivers of SC to inform the design of future
interventions to promote SC adoption and
implementation through SC methods,
processes and materials.
Research methods
Given the exploratory nature of this study,
a survey research approach was adopted.
A structured questionnaire survey was
conducted through recruiting various
stakeholders who are currently working in
the construction industry in South Africa.
The purpose of the survey was to under-
stand the perception and attitude towards
sustainable construction by stakeholders in
the construction industry.
Study population and sampling
The population of this study comprised
construction industry professionals working
in the built environment in South Africa on
building projects. Due to the large number
of professionals in the construction industry,
no available list of all construction industry
professionals, and potential challenges with
access and communication, the sampling

14 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
frame adopted was a nonprobability sample.
Nonprobability sampling is a sampling
technique in which the researcher chooses
participants based on their willingness to
participate in the research study (Etikan et
al 2015). A purposive snowball sampling
technique was therefore used in this
research to obtain a valid and effective
overall sample size. This method has been
used in previous construction, engineering
and management studies and allows for
data to be gathered from participants who
share the research study or make referrals
of prospective participants (Gan et al 2015;
Jiang & Wong 2016; Darko et al 2017b; Chan
et al 2018). The online questionnaire survey
was distributed via LinkedIn (a professional
social media platform) to professionals who
met the inclusion criteria (i.e. working in
the built environment in South Africa on
building projects). Using search criteria to
filter each of the discipline services provided
within the design and development phase
of building projects (e.g. quantity surveyors,
architects, landscape architects, structural
engineers, civil engineers, building services
engineers which include mechanical, elec-
trical and fire), as well as location (South
Africa), individuals were sent messages
inviting them to participate in the research
study. Initial participants were asked to
share the survey with other professionals in
the built environment who they knew also
worked on building projects.
Research instrument
The purpose of the questionnaire survey in
this study was to understand which barri-
ers and drivers of sustainable construction
need to be prioritised for an intervention
strategy. A self-administered online
questionnaire is the preferred type of data
collection procedure for this phase of the
research for the following reasons:
i. Cost-effective to administer.
ii. Ease of gathering data, as surveys can be
distributed and completed rapidly.
iii. Data inputs are readily available from
an online survey database.
iv. Due to automated data capturing, there
is a reduced possibility of errors whilst
handling the data.
v. Online surveys are convenient for
potential participants as they can com-
plete the survey at a time and a place
that are convenient for them.
Design of the questionnaire survey
The questionnaire survey was adapted
from Huijg et al (2014) and informed by
Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021
the TDF and COM-B model to guide the
behavioural analysis and understand the
perceived barriers and drivers of sustain-
able construction identified by participants.
The results of the questionnaire survey
allowed the researcher to identify key per-
ceptions from the various domains of the
TDF that can be targeted in an interven-
tion strategy to facilitate the adoption and
implementation of SC. The questionnaire
survey consisted of seven sections covering:
i. Background to the research study and
informed consent
ii. Key definitions pertaining to the study
iii. Demographics of participants
iv. Knowledge, skills and social and profes-
sional role and identity
v. Beliefs about capabilities, optimism, and
beliefs about consequences
vi. Reinforcement, intentions, goals, social
influences, and behavioural regulation
vii. Environmental context and resources.
The scales used to measure the items
include categorical nominal (e.g. type of
profession, type of service disciplines at
organisations, yes/no questions) and ordi-
nal scales (e.g. years of experience, number
of employees at an organisation, Likert
scale). The Likert scale which makes up the
majority of the questionnaire items was
a five-point scale measuring: 1 ‒ strongly
disagree, 2 – disagree, 3 – neither disagree
or agree, 4 – agree, and 5 – strongly agree.
The Likert scale provides an overview
of the intensity of an attitude towards a
statement and was therefore deemed the
appropriate scale used to measure the TDF
domain items (Bryman 2010).
Data analysis
The data collected was manipulated and
analysed using the statistical software pack-
age SPSS. The data analysis was conducted
Table 2 Profile of respondents
Professions
Contractor
Quantity surveyor
Architect
Structural engineer
Civil engineer
Mechanical engineer
Electrical engineer
Environmental engineer
Façade engineer
Total
by performing various statistical analysis
approaches. Firstly, the Cronbach’s Alpha
test was used to measure the reliability
of the data, frequency tables were used to
report the results from the Likert scale
items, followed by determining if there was
a significant difference between two or more
groups of participants using the Mann-
Whitney U test and Kruskal Wallis test.
The Mann-Whitney test is a nonparametric
test to compare two categorical unpaired
groups of data (Gaddis & Gaddis 1990). In
this study, the two groups that were studied
were structural and civil engineers, as these
groups represented a significant sample
of at least 30 data sets each. The Kruskal
Wallis test is another nonparametric used
when more than two categorical groups are
compared (Gaddis & Gaddis 1990), which in
the case of this research study was identified
as the years of experience, which had four
groups (0 to 4 years, 5 to 9 years, 10 to 19
years, more than or equal to 20 years).
Analysis of Results
Demographics and
background information
Out of the 290 questionnaire surveys that
were sent out, 108 responses were received,
which represents a 37% response rate. It
has been argued that studies within the
construction industry generally have a
lower response rate between 20% and 30%
(Abidin & Powmya 2014); Gan et al 2015).
However, the sample size can be considered
representative of the population, as the
central limit theorem holds true with a
sample size greater than 30, which means
statistical analysis can be conducted. The
descriptive statistics of the sample popula-
tion are presented in Tables 2 and 3.
Frequency Percentage (%)
1 0.9
7 6.5
7 6.5
37 34.3
36 33.3
14 13.0
2 1.9
3 2.8
1 0.9
108 100.0
15
Table 3 Experience of respondents implementation of SC. Table 4 displays
Years of experience Sustainable /
the means and standard deviations for
(construction industry) Green Building experience the knowledge, skills and behavioural
Professions
regulation domains linked to psychological
0–4 5–9 10–19 ≥20 Yes No
capability and provides the frequency of
Contractor 0 1 0 0 1 0 responses for each questionnaire statement.
Quantity surveyor 4 3 0 0 0 7 Construction industry stakeholders’
assessment of their capabilities regarding
Architect 0 4 3 0 3 4
their knowledge of SC was generally posi-
Structural engineer 17 12 6 2 12 25 tive, as the most frequent responses to all
Civil engineer 22 10 3 1 8 28 five questions was “agree”. Responses in the
skills domain indicated that the majority
Mechanical engineer 4 4 4 2 12 2
of stakeholders understand SC (88.9%) and
Electrical engineer 0 2 0 0 1 1 participants agree that education on SC is
available (61.1%) through external service
Environmental engineer 0 2 1 0 1 2
providers (56.5%). However, most partici-
Façade engineer 0 0 1 0 1 0 pants do not have access to training within
Subtotal 47 38 18 5 39 69 their organisations (38.9%). Participants
perceived that there is a requirement for
Percentage (%) 43.5 35.2 16.7 4.6 36.1 63.9
behaviour change in the construction
industry (92.6%), as current construction
TDF and COM-B variable analysis provided. Questionnaire statements have practices are not sustainable and 71.3%
The questionnaire provided five response been simplified for ease of reference. The agreed that there is resistance to adopt
categories ranging from “strongly following section describes the data within sustainable alternatives to construction
disagree” to “strongly agree”. During the the TDF and COM-B model domains and industry processes and methods.
frequency analysis, the responses “agree” categories.
and “strongly agree” were combined into Opportunity
one category, agree, which represents a Capability Opportunity represents the social and
participant who agrees with a statement, Capability represents an individual’s ability physical environment that shapes the
and the responses “disagree” and “strongly to engage in a target behaviour and hav- opportunity to engage in a target behaviour
disagree” were combined into one category, ing the necessary knowledge and skills to (Michie et al 2011). Social opportunity
disagree, which represents a participant do so (Michie et al 2011). Construction refers to the social factors that influence
who disagrees with a statement. This is industry stakeholders’ psychological the way we think about things (i.e. cul-
referred to as categorisation in the tables capabilities influence their adoption and tural norms and social cues). Physical

Table 4 Capability questionnaire statement responses

Capability Questionnaire response (%)

Questionnaire statement Mean (SD) 1 2 3 4 5 Categorisation

Knowledge domain

CF1 – Awareness of SC 3.83 (0.78) 1.9 4.6 14.8 65.7 13.0 Agree

CF2 – Knowledge of SC 3.46 (0.88) 1.9 11.1 35.2 42.6 9.3 Agree

CF3 – Familiarity with SC 3.72 (0.81) 2.8 3.7 22.2 61.1 10.2 Agree

CF4 – Interest in SC 3.56 (0.98) 3.7 12.0 21.3 50.9 12.0 Agree

CF5 – Demand for SC 3.47 (0.94) 0.9 14.8 34.3 36.1 13.9 Agree

Skills domain

CF6 – Understanding of SC 4.13 (0.76) 1.9 1.9 7.4 59.3 29.6 Agree

CF7 – S C skills are within the scope of construction industry professionals 3.87 (0.86) 2.8 2.8 18.5 56.5 19.4 Agree

CF8 – SC education is available 3.66 (0.90) 1.9 7.4 29.6 45.4 15.7 Agree

CF9 – External training on SC is available 3.52 (0.95) 1.9 13.9 27.8 43.5 13.0 Agree

CF10 – Internal training on SC is available 2.86 (0.99) 8.3 30.6 28.7 31.5 0.9 Disagree

Behavioural regulation domain

CF51 – Resistance to change behaviour 3.81 (0.80) 0.9 4.6 23.1 54.6 16.7 Agree

CF52 – Commitment to change behaviour 4.26 (0.62) 0 0.9 6.5 58.3 34.3 Agree

16 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Table 5 Opportunity questionnaire statement responses
Opportunity Questionnaire response (%)

Questionnaire statement Mean (SD) 1 2 3 4 5 Categorisation

Social influences domain

CF48 – Industry peers support and encourage SC 3.83 (0.79) 0.9 5.6 18.5 59.3 15.7 Agree

CF49 – Industry superiors encourage and support SC skills development 3.37 (0.86) 0.9 13.9 40.7 36.1 8.3 Agree

CF50 – Industry peers demonstrate an interest in SC 3.40 (0.93) 2.8 14.8 29.6 45.4 7.4 Agree

Environmental context and resources domain

CF53 – Lack of interest in SC 3.30 (0.85) 0.9 19.4 31.5 45.4 2.8 Agree

CF54 – L ack of communication and integrated work environment 3.58 (0.83) 0.0 12.0 27.8 50.0 10.2 Agree

CF55 – Lack of sustainable technology specifications 3.56 (0.82) 0.9 11.1 26.9 53.7 7.4 Agree

CF56 – L imited availability of sustainable product suppliers 3.41 (0.90) 1.9 13.9 34.3 41.7 8.3 Agree

CF57 – L ack of databases and information for SC technologies 3.55 (0.77) 0.0 10.2 31.5 51.9 6.5 Agree

CF58 – SC promotes environmental protection 4.31 (0.69) 0.0 0.9 10.2 45.4 43.5 Agree

CF59 – SC promotes energy conservation 4.34 (0.67) 0.0 0.9 8.3 46.3 44.4 Agree

CF60 – SC promotes water conservation 4.37 (0.66) 0.0 0.9 7.4 45.4 46.3 Agree

CF61 – SC promotes resource conservation 4.38 (0.64) 0.0 0.9 5.6 48.1 45.4 Agree

CF62 – SC promotes waste reduction 4.19 (0.80) 0.0 3.7 13.0 44.4 38.9 Agree

CF63 – S C requires support from executive management 4.18 (0.64) 0.0 0.0 13.0 56.5 30.6 Agree

CF64 – Organisations should promote a culture and awareness about SC 4.25 (0.60) 0.0 0.0 8.3 58.3 33.3 Agree

CF65 – O
 rganisations should facilitate best practice sharing of SC 4.28 (0.61) 0.0 0.9 5.6 58.3 35.2 Agree

CF66 – M
 utual understanding and commitment amongst SC
4.20 (0.65) 0.0 0.9 10.2 56.5 32.4 Agree
professionals is required

CF67 – P
 erformance-based measurements will promote SC 4.19 (0.73) 0.0 1.9 13.0 49.1 36.1 Agree

CF68 – Mandatory SC building policies and regulations will promote SC 4.17 (0.74)) 0.0 2.8 12.0 50.9 34.3 Agree

CF69 – B
 etter enforcement of SC building policies after development
4.11 (0.74) 0.0 2.8 13.9 52.8 30.6 Agree
will promote SC

CF70 – D
 eveloping and strengthening regulatory mechanisms will
4.14 (0.76) 0.0 3.7 11.1 52.8 32.4 Agree
promote SC

CF71 – Institutional frameworks to guide SC is required 4.24 (0.65) 0.0 0.9 9.3 54.6 35.2 Agree

CF72 – P
 roduct and material innovation and/or certification is required 4.22 (0.62) 0.0 0.9 7.4 60.2 31.5 Agree

CF73 – M
 aterials manufacturers should be proactive in development of
4.31 (0.65) 0.0 1.9 4.6 53.7 39.8 Agree
SC technologies

CF74 – C
 ollaborative and strengthened R&D within the construction
4.24 (0.65) 0.0 1.9 6.5 57.4 34.3 Agree
industry

CF75 – A
 vailability of better information on cost and benefits of SC
4.32 (0.68) 0.0 1.9 6.5 49.1 42.6 Agree
technologies from a reliable database

opportunity is represented by the envi-
ronment which includes time, resources
and location. The results from the social
influences and environmental context and
resources domains linked to social and
physical opportunity are presented in
Table 5.
Social structures within organisations,
as well as in the construction industry,
influence stakeholders’ behaviour towards
adopting and implementing SC. 75% of
participants agreed that their peers in
the construction industry support and
encourage SC, and more than half of the
participants (52.8%) agreed that their peers
in the construction industry demonstrate
an interest in SC. Most participants agreed
that there is a lack of communication and
an integrated work environment amongst
construction industry stakeholders (60.2%)
and a lack of sustainable technology speci-
fications (61.1%). Participants indicated that
organisations are responsible for facilitat-
ing a culture of best practice sharing in
relation to SC (93.5%) and should promote
a culture and awareness about SC (91.6%).
The results further indicated that, for the
successful implementation of SC, materials
manufacturers should be proactive in the
development of SC technologies (93.5%),
and availability of better information on
cost and benefits of SC technologies from a
reliable database is necessary (91.7%).
Motivation
Motivation to engage in a target behaviour
can either be reflective or automatic.
Reflective motivation processes involve
plans (self-conscious intentions) and

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 17
Table 6 Motivation questionnaire statement responses
Motivation Questionnaire response (%)
Questionnaire statement Mean (SD) 1 2 3 4 5 Categorisation
Social/professional role and identity domain
CF11 – Construction industry professionals have a responsibility towards SC 4.10 (0.81) 0.0 5.6 11.1 50.9 32.4 Agree
CF12 – S C forms part of construction industry professionals’ work 4.28 (0.78) 0.9 1.9 9.3 44.4 43.5 Agree
CF13 – C
 onstruction industry professionals play a role in the implementation of SC 4.30 (0.78) 0.9 2.8 5.6 47.2 43.5 Agree
CF14 – S C provides a good corporate image 4.29 (0.81) 0.9 0.9 13.9 37.0 47.2 Agree
CF15 – S C provides a competitive advantage 4.12 (0.92) 0.9 7.4 9.3 43.5 38.9 Agree
CF16 – SC shows an organisation’s commitment to social responsibility 4.28 (0.81) 0.9 3.7 5.6 46.3 43.5 Agree
CF17 – S C provides collaboration amongst construction industry professionals 4.15 (0.76) 0.0 2.8 13.9 49.1 34.3 Agree
CF18 – SC provides an integrated and whole-building design approach 4.10 (0.72) 0.0 0.9 18.5 50.0 30.6 Agree
CF19 – S C sets a standard for future development in the built environment 4.20 (0.77) 0.9 1.9 10.2 50.0 37.0 Agree
Beliefs about capabilities domain
CF20 – C
 I professionals would adopt SC if they had the skills 4.19 (0.71) 0.0 2.8 9.3 54.6 33.3 Agree
CF21 – CI professionals would adopt SC even if there are time constraints on
3.77 (0.82) 0.0 9.3 19.4 56.5 14.8 Agree
building projects
CF22 – CI professionals would adopt SC even if there is no incentive to do so on
3.94 (0.71) 0.0 3.7 16.7 61.1 18.5 Agree
building projects
Optimism domain
CF23 – A
 dopting and implementing SC will mean positive outcomes for the
4.27 (0.68) 0.0 0.9 10.2 50.0 38.9 Agree
economy, society and the environment
CF24 – SC represents a positive change for the construction industry in South Africa 4.37 (0.69) 0.0 1.9 6.5 44.4 47.2 Agree
Beliefs about consequences domain
CF25 – Sustainable buildings improve end-users’ productivity 3.72 (0.85) 0.9 4.6 34.3 41.7 18.5 Agree
CF26 – S ustainable buildings improve indoor environmental quality, enhanced
4.07 (0.72) 0.0 0.9 19.4 50.9 28.7 Agree
occupants’ health, comfort and wellbeing
CF27 – S C ensures green market growth and provides job opportunities 3.95 (0.78) 0.0 1.9 26.9 45.4 25.9 Agree
CF28 – Necessity to adopt and implement SC 4.08 (0.81) 0.0 4.6 14.8 48.1 32.4 Agree
CF29 – S C is time-consuming which could cause project delays 3.24 (0.98) 4.6 15.7 38.9 32.4 8.3 Agree
CF30 – SC implementation will increase the capital cost of construction of building
3.61 (0.86) 0.9 9.3 30.6 46.3 13.0 Agree
projects
CF31 – S ustainable products and materials will increase the capital cost of building
3.54 (0.89) 0.9 13.0 28.7 46.3 11.1 Agree
projects
CF32 – R
 isk of investment with implementing new sustainable materials and
3.13 (0.87) 1.9 24.1 35.2 37.0 1.9 Agree
construction methods
CF33 – S cepticism about the necessity to implement SC principles on building
2.21 (0.88) 18.5 51.9 20.4 8.3 0.9 Disagree
projects
CF34 – SC reduces the whole life-cycle cost of a building 3.29 (1.07) 6.5 14.8 34.3 32.4 12.0 Agree
CF35 – SC has long pay-back periods 3.25 (0.87) 2.8 12.0 50.0 27.8 7.4 Neutral
CF36 – S C provides high return on investment of buildings 3.46 (0.75) 0.0 7.4 47.2 37.0 8.3 Neutral
CF37 – SC enhances property value 3.52 (0.78) 0.0 8.3 40.7 41.7 9.3 Agree
CF38 – S C improves performance of the economy 3.74 (0.81) 0.0 4.6 35.2 41.7 18.5 Agree
Intentions and goals domain
CF43 – Intention to develop SC knowledge and skills 4.09 (0.62) 0.0 0.0 14.8 61.1 24.1 Agree
CF44 – Intention to consider adopting SC 4.20 (0.54) 0.0 0.0 6.5 66.7 26.9 Agree
CF45 – Intention to promote the education of SC internally 3.85 (0.72) 0.0 1.9 28.7 51.9 17.6 Agree
CF46 – Intention to promote internal training of SC 3.74 (0.79) 0.0 6.5 27.8 50.9 14.8 Agree
CF47 – Increasing SC awareness and knowledge is an important career goal 4.06 (0.72) 0.0 0.9 20.4 50.9 27.8 Agree
Reinforcement domain
CF39 – Access to funding 4.13 (0.70) 0.0 2.8 10.2 58.3 28.7 Agree
CF40 – Financial and market-based incentives 4.33 (0.56) 0.0 0.0 4.6 57.4 38.0 Agree
CF41 – T ax relief 4.33 (0.76) 0.9 1.9 6.5 44.4 46.3 Agree
CF42 – S ubsidies for R&D 4.33 (0.77) 0.9 1.9 7.4 42.6 47.2 Agree

18 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
evaluations (beliefs about what is good
and bad) (Michie et al 2014). On the other
hand, automatic motivation processes
involve the wants and needs, desires,
impulses and reflex responses of individu-
als. The results from the social/professional
role and identity, beliefs about capabilities,
optimism, beliefs about consequences,
intentions and goals, and reinforcement
domains linked to reflective and automatic
motivation are presented in Table 6.
All the questions related to participants’
social and professional role and identity
had highly positive responses with small
variance. Participants agree that construc-
tion industry professionals play a role in
the implementation of SC (90.7%) and that
SC adoption shows an organisation’s com-
mitment towards its social and environ-
mental responsibility (89.8%). Construction
industry professionals are confident that
they would adopt SC if they had the skills
(87.9%) even if they had no incentive to
do so (79.6%). Participants are highly
optimistic about SC as it represents a posi-
tive change for the construction industry
(91.6%) and the adoption and implementa-
tion thereof will have a positive impact
on the economy, society and environment
(88.9%).
Participants agree that SC benefits the
end-user through sustainable buildings,
which improve indoor air quality and
comfort (79.6%) and end-user’s productivity
(60.2%). Participants were probed about the
economic benefits of SC, and the results
suggest that SC will ensure green market
growth and provide job opportunities
(71.3%) and improve the performance of
the economy (60.2%). On the other hand,
participants believe that the key barriers
to adopting SC are that the implementa-
tion of SC will increase the capital cost of
construction (59.3%) through the adoption
of sustainable products and materials
(57.4%). Despite this, most participants
indicated that there is a need to adopt SC
(80.5%). The highly positive responses to
questions about participants’ intentions
and goals indicated that they do intend
to consider adopting SC (93.6%) and
further develop their SC knowledge and
skills (85.2%). Participants indicated that
government support and encouragement
through financial and market-based incen-
tives (95.4%), tax relief on projects which
adopt SC (90.7%), subsidies for research
and development linked to SC (89.8%), and
access to funding through financial institu-
tions (87%), would increase SC adoption
and implementation.
TDF domains analysis
The Cronbach’s alpha values for the sub-
scales of the questionnaire (i.e. the TDF
domains) were shown to have very good
reliability (above 0.80 for five scales and
above 0.70 for four scales) (Table 7). The
items in the questionnaire were therefore
combined into the TDF domains for fur-
ther analysis. To identify which domains
need to be targeted in the behaviour
change intervention, the TDF domains
were analysed as scales. Table 7 displays
the minimum and maximum value for the
responses, means and standard deviations
averaged across all participants. The means
for all scales indicate that, on average, par-
ticipants were positively inclined towards
the adoption and implementation of sus-
tainable construction. The optimism scale
had the highest mean (4.32{0.60}) followed
by the reinforcement scale (4.28{0.57}) and
social/professional role and identity scale
(4.20{0.60}). All these scales fall under the
category of Motivation within the COM-B
model which indicates that there is a strong
motivation to adopt and implement SC
amongst construction industry profession-
als. The beliefs about consequences scale
had the lowest mean (3.49{0.43}) followed
by the social influences scale (3.53{0.74}),
the knowledge scale (3.61{0.59}) and the
skills scale (3.61{0.65}). These scales also
indicate the largest variability in responses,
which indicates that there could be
improvement in all the COM-B categories
to improve the adoption and implementa-
tion of SC. However, capability (knowledge
and skills) and opportunity (social influ-
ences) will often influence the motivation
(beliefs about consequences) to enact a
behaviour (Michie et al 2014). Therefore,
the more capable stakeholders are by devel-
oping their SC knowledge and skills, and by
being in an environment which encourages
their SC development, the more they are
likely to adopt and implement SC practices.

Table 7 Descriptive statistics for TDF domains and COM-B components

Number Scale
Scale Minimum Maximum Mean (*) SD {*}
of items reliability (α)

Capability

Knowledge 5 0.683 1 5 3.61 0.59

Skills 5 0.770 1 5 3.61 0.65

Behavioural regulation 2 0.521 2 5 4.04 0.59

Opportunity

Social influences 3 0.824 1 5 3.53 0.74

Environmental context and resources 23 0.912 3 5 4.08 0.42

Motivation

Social / professional role identity 9 0.905 2 5 4.20 0.60

Beliefs about capabilities 3 0.761 2 5 3.97 0.62

Optimism 2 0.704 3 5 4.32 0.60

Beliefs about consequences 14 0.768 3 5 3.49 0.43

Intentions and goals 5 0.885 3 5 3.99 0.57

Reinforcement 4 0.83 3 5 4.28 0.57

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 19
Table 8 Barriers and drivers mapped to COM-B and TDF domains
COM-B
TDF
Awareness, knowledge and
information of SC
Interest in SC and demand
for SC
Training availability of SC
Access to education on SC
Barriers and drivers
Behavioural change towards
SC
Industry peer influences
Confidence in SC
implementation
Economic factors
Perception of SC
Social benefits
The analysis of the TDF domains indicate
that there are five key behavioural domains
which should be targeted, including
construction industry stakeholders’ beliefs
about consequences, knowledge, social
influences, skills, and beliefs about capabili-
ties. Ten key barriers and drivers identified
in each of these domains are summarised
in Table 8.
Discussion
Although there is a high level of aware-
ness and familiarity with SC, there is a
variation in the results when it comes to
knowledge about SC and the demand for
SC in the construction industry. The high
level of awareness and familiarity amongst
construction industry stakeholders is
contrary to previous studies conducted
in developing countries by authors such
as Ismail et al (2012), Shari and Soebarto
(2012), and Djokoto et al (2014). Their stud-
ies suggest that there is a lack of awareness
amongst construction industry stakehold-
ers and the public. Firstly, the difference
in results could be due to the nature of
the research designs adopted ‒ Shari and
Soebarto (2012) conducted a qualitative
study using semi-structured interviews
to identify the barriers to the adoption of
20
Capability Opportunity
Psychological Social
Cognitive and
Behavioural
Knowledge interpersonal
regulation
skills
SC, whilst the current research study used
a survey questionnaire. Secondly, previ-
ous research studies focused on specific
stakeholder groups, such as developers
only, architects only, or contractors only.
An example of this is the research study
conducted by Ismail et al (2012), which
only included developers. This is a different
target population to the current research
study which included architects, engineers,
contractors and quantity surveyors. Lastly,
the research instrument is structured dif-
ferently, as shown in the study conducted
by Djokoto et al (2014), where the authors
prompted participants to rate the extent
to which each of the identified barriers
affects SC using a five-point Likert scale,
whereas in the current study participants
were prompted to rate the extent to which
they agreed or disagreed with various belief
statements about SC.
In a study conducted by Lim et al
(2019), the authors stated that, although
quantity surveyors in Australia had a
reasonable level of awareness of SC, there
was still a lack of implementation, which
could be attributed to individuals’ attitudes
towards SC, cultural and institutional chal-
lenges. This provides insight into the varia-
tion of results regarding the knowledge and
demand for SC, which could be explained
Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Motivation
Reflective
Social Beliefs about Beliefs about
influences capabilities consequences
by a lack of “actionable” knowledge that
presents itself in the form of limited
information about best practices, access
to existing relevant knowledge and the
perception of information overload (Wilson
& Rezgui 2013). In other words, although
participants have the knowledge of what
the contents and objectives of SC are, there
is no knowledge sharing amongst profes-
sionals, or information guides to inform
SC implementation. Furthermore, the
variation in the results for the demand for
SC suggests that there might be a lack of
demand from clients for sustainable proj-
ects (Serpell et al 2013), a lack of evidence
of the benefits and opportunities of SC, and
the perception that SC will increase project
costs (Shari & Soebarto 2012). Participants
indicated that SC is within the scope of
construction industry professionals and
that they have access to education and
training related to SC. This, however, high-
lights the gap between access to knowledge
and training related to SC and the relatively
low percentage of participants who have
worked on a sustainable building project
and successfully implemented SC practices.
Additionally, the resistance to the change
in current construction industry practices
could further hinder the adoption and
implementation of SC. Participants felt that
it is critical for construction industry stake-
holders to change their current behaviour
and adopt sustainable alternatives to design
and construction methods and processes.
Participants described the social
opportunity to adopt and implement SC
as positive in relation to the support and
encouragement from peers in the construc-
tion industry. However, there could be
limitations to career development within
the scope of sustainability, as superiors
in the construction industry might not
support the development of skills related
to SC. As Opoku and Ahmed (2014) state,
there is a shortage of skills and capacity in
terms of numbers of construction industry
professionals who can support the imple-
mentation of SC. Construction industry
stakeholders should therefore be encour-
aged by their industry peers and superiors
and given the opportunity to develop their
skills to improve the sustainability of the
construction industry. Furthermore, with-
out an integrated work environment where
stakeholders are working in collaboration
to achieve the objectives of SC, the success-
ful adoption and implementation of SC will
be deterred (Häkkinen & Belloni 2011). The
physical opportunity to adopt and imple-
ment SC highlights various barriers, such
as the lack of availability of sustainable
technologies, lack of sustainable product
and material suppliers, and lack of data-
bases with information about sustainable
products. This is well-documented in the
literature as common technological barri-
ers to the adoption and implementation of
SC (Aigbavboa et al 2017; Chan et al 2017b;
Häkkinen & Belloni 2011; Khalfan et al
2015; Pham et al 2019; Pitt et al 2009; Shi
et al 2013). Participants appreciate that the
principles of SC encourage environmental
protection through the conservation of
resources, energy and water, and waste
reduction. Providing opportunities
within organisations to develop a culture
of sustainability, best practice sharing and
commitment to increasing the awareness
of SC will improve the adoption of SC.
Furthermore, regulations and frameworks
to guide the adoption and implementation
of SC, as well as reliable databases with
sustainable product information, will fur-
ther aid the adoption and implementation
of SC. The findings suggest that there are
opportunities to transition towards a more
sustainable construction industry.
The survey results indicate that
motivation plays a role in determining
the likelihood of participants engaging
Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021
with SC. In terms of reflective motiva-
tion, all participants indicated that SC
was beneficial to the development of their
careers as construction industry profes-
sionals. Participants agreed that if they had
the skills, they would adopt SC and were
highly optimistic about the benefits of SC
for the construction industry in the long
term. Participants had strong intentions
to develop their SC knowledge and skills
and encourage the education and training
thereof. In terms of automatic motiva-
tion, participants indicated government
should support and encourage SC through
incentives, tax relief, subsidies and access
to funding for projects that incorporate
sustainable principles and practices.
These findings are supported by previous
literature, which suggests that motivation
and support from key role players in the
construction industry, such as government,
developers and clients, are essential for the
successful and widespread adoption of SC,
specifically in the early stages of adoption.
The results from the survey suggest that
addressing the barriers to the implemen-
tation of SC linked to the capabilities,
opportunities and motivation of construc-
tion industry stakeholders may facilitate
the transition towards a more sustainable
construction industry.
Limitations
The findings from this study should be
interpreted and understood with the fol-
lowing limitations in mind: Although the
study aimed to seek the perceptions of all
construction industry stakeholders, struc-
tural and civil engineers represented 67.6%
of the sample. The findings of the research
related to the targeted TDF domains
should therefore be interpreted as such;
however, it is deemed to be transferrable to
other stakeholder groups. There were limi-
tations and challenges with recruitment of
participants. Nonetheless, the participants
who responded provided sufficient con-
textual data, which helped to achieve the
objectives of this research study.
Conclusions and
recommendations
Conclusions
The overall positive response to the study
suggests that there is a need to adopt and
implement SC, and construction industry
stakeholders are optimistic about the posi-
tive impact SC could have on the economy,
environment and society. The findings
of the research study highlight a range of
factors related to construction industry
stakeholders’ capability, opportunity and
motivation that require improvement to
increase the adoption and implementa-
tion of SC practices. Using a theory-based
approach, ten barriers and drivers of SC
adoption and implementation, related to
construction industry stakeholders’ capa-
bility, opportunity and motivation, were
identified. The ten barriers and drivers
included: awareness, knowledge and infor-
mation of SC, interest in SC and demand
for SC adoption, training availability of
SC, access to education on SC, behavioural
change towards SC, industry peer influ-
ences, confidence in SC implementation,
economic factors, perception of SC, and
social benefits of adopting SC. Five key tar-
get domains were identified: beliefs about
consequences, knowledge, social influ-
ences, skills and beliefs about capabilities,
which can be used to develop behaviour
change interventions, such as interven-
tions aimed at adopting circular economy
principles or agile and lean principles to
improve the adoption and implementa-
tion of SC amongst construction industry
stakeholders.
Recommendations
Recommendation 1
By adopting a qualitative theory-based
approach, a future research study may
provide greater in-depth detail about which
domains should be targeted using the
TDF domains as a guide. Specific domains
may also be identified that were not high-
lighted by the quantitative approach, which
restricts the participants’ responses to a
structured set of questions.
Recommendation 2
Future research should include additional
disciplines from other stages of the project
life cycle, such as clients and project
managers, to understand the context of the
research from varying perspectives.
Recommendation 3
Universities are key contributors to educa-
tion, research and innovation which are
essential to achieve the goals of sustainable
development within the construction
industry. Integrating sustainability as part
of the undergraduate curriculum and post-
graduate studies in higher education is nec-
essary to promote sustainable development
21
in the built environment. This ensures that
graduates are aware of their ethical and
social responsibility towards the environ-
ment and the welfare of society.
Recommendation 4
Organisations should provide stakehold-
ers with access to education, training and
upskilling through continuing profes-
sional development courses to improve the
knowledge and skills available to facilitate
the adoption and implementation of SC.
Sustainable development strategies and
internal policies should be developed to
further enhance the organisation’s offering
within the industry.
Recommendation 5
Policies and legislation which govern the
adoption and implementation of SC should
be developed and regulated by government
to emphasise the need to change the cur-
rent trajectory of the construction industry.
Financial and market-based incentives
would further encourage organisations to
adopt best practices within the industry in
relation to sustainable construction.
References
Abidin, N Z & Powmya, A 2014. Perceptions on
motivating factors and future prospects of green
construction in Oman. Journal of Sustainable
Development, 7(5): 231–239.
Aghimien, D, Aigbavboa, C, Oke, A & Musenga, C
2018. Barriers to sustainable construction practices
in the Zambian construction industry. Proceedings,
8th International Conference on Industrial
Engineering and Operations Management, 6–8
March 2018, Bandung, Indonesia, pp 2383–2392.
Ahn, Y H, Pearce, A R, Wang, Y & Wang, G 2013.
Drivers and barriers of sustainable design and
construction: The perception of green building
experience. International Journal of Sustainable
Building Technology and Urban Development,
4(1): 35–45.
Aigbavboa, C, Ohiomah, I & Zwane, T 2017.
Sustainable construction practices: A “lazy
view” of construction professionals in the South
African construction industry. Energy Procedia,
105: 3003–3010.
AlSanad, S 2015. Awareness, drivers, actions, and
barriers of sustainable construction in Kuwait.
Procedia Engineering, 118: 969–983.
Ametepey, O, Aigbavboa, C & Ansah, K 2015. Barriers
to successful implementation of sustainable
construction in the Ghanaian construction industry.
Procedia Manufacturing, 3: 1682–1689.
Amui, L B L, Jabbour, C J C, Jabbour, A B L & Kannan,
D 2017. Sustainability as a dynamic organizational
22
capability: A systematic review and a future agenda
toward a sustainable transition. Journal of Cleaner
Production, 142: 308–322.
Atkins, L, Francis, J, Islam, R, O’Connor, D, Patey,
A, Ivers, N, et al 2017. A guide to using the
Theoretical Domains Framework of behaviour
change to investigate implementation problems.
Implementation Science, 12(1): 1–18.
Bebbington, J 2001. Sustainable development: A
review of the international development, business
and accounting literature. Accounting Forum,
25(2): 128–157.
Brennan, M C & Cotgrave, A J 2014. Sustainable
development: A qualitative inquiry into the current
state of the UK construction industry. Structural
Survey, 32(4): 315–330.
Bryman, A 2010. Social Research Methods, 4th ed. New
York: Oxford University Press.
Cane, J, Connor, D O & Michie, S 2012. Validation
of the Theoretical Domains Framework for use in
behaviour change and implementation research.
Implementation Science, 7(37): 1–17.
Chan, A P C, Darko, A & Ameyaw, E E 2017a.
Strategies for promoting green building
technologies adoption in the construction industry:
An international study. Sustainability, 9(6): 969.
Chan, A P C, Darko, A, Ameyaw, E E & Owusu-Manu,
D-G. 2017b. Barriers affecting the adoption of green
building technologies. Journal of Management in
Engineering, 33(3): 04016057.
Chan, A P C, Darko, A, Olanipekun, A O & Ameyaw, E E
2018. Critical barriers to green building technologies
adoption in developing countries: The case of Ghana.
Journal of Cleaner Production, 172: 1067–1079.
Coetzee, D A & Brent, A C 2015. Perceptions of
professional practitioners and property developers
relating to the costs of green buildings in South
Africa. Journal of the South African Institution of
Civil Engineering, 57(4): 12–19.
Dalal-Clayton, B & Bass, S 2012. Sustainable
Development Strategies, 1st ed.: London: Earthscan.
Darko, A, Zhang, C & Chan, A P C 2017a. Drivers
for green building: A review of empirical studies.
Habitat International, 60: 34–49.
Darko, A, Chan, A P C, Gyamfi, S, Olanipekun, A O,
He, B J & Yu, Y 2017b. Driving forces for green
building technologies adoption in the construction
industry: Ghanaian perspective. Building and
Environment, 125: 206–215.
Djokoto, S D, Dadzie, J & Ohemeng-Ababio, E
2014. Barriers to sustainable construction in the
Ghanaian construction industry : Consultants’
perspectives. Journal of Sustainable Development,
7(1): 134–143.
Du Plessis, C 2007. A strategic framework for
sustainable construction in developing countries.
Construction Management and Economics,
25(1): 67–76.
Du Plessis, C, Adebayo, A, Ebohan, J, Irurah, D,
Rwelamila, P D, Hassan et al. 2002. Sustainable
Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
construction in developing countries. Pretoria: CSIR
Building and Construction Technology.
Elliott, J A. 2006. An Introduction to Sustainable
Development. 3rd ed. London: Routledge.
Etikan, I, Alkassim, R & Abubakar, S 2015.
Comparision of snowball sampling and sequential
sampling technique. Biometrics & Biostatistics
International Journal, 3(1): 00055.
Foong, D, Mitchell, P, Wagstaff, N, Duncan, E
& McManus, P 2017. Transitioning to a more
sustainable residential built environment in Sydney.
Geo: Geography and Environment, 4(1): e00033.
Gaddis, G M & Gaddis, M L 1990. Introduction to
biostatistics. Part 5. Statistical inference techniques
for hypothesis testing with nonparametric data.
Annals of Emergency Medicine, 19(9): 1054–1059.
Gainforth, H L, Sheals, K, Atkins, L, Jackson, R &
Michie, S 2016. Developing interventions to change
recycling behaviors: A case study of applying
behavioral science. Applied Environmental
Education and Communication, 15(4): 325–339.
Gan, X, Zuo, J, Ye, K, Skitmore, M & Xiong, B
2015. Why sustainable construction? Why not?
An owner’s perspective. Habitat International,
47: 61–68.
Häkkinen, T & Belloni, K 2011. Barriers and drivers
for sustainable building. Building Research &
Information, 39(3): 239–255.
Hjorth, P & Bagheri, A 2006. Navigating towards
sustainable development: A system dynamics
approach. Futures, 38: 74–92.
Huijg, J M, Gebhardt, W A, Crone, M R, Dusseldorp,
E & Presseau, J 2014. Discriminant content validity
of a theoretical domains framework questionnaire
for use in implementation research. Implementation
Science, 9(11): 1–16.
Ismail, Z, Idris, N H, Nasir, N M & Alam, S 2012.
Sustainable initiative and impediments towards
promoting sustainable construction in Malaysia.
Proceedings, IEEE Colloquium on Humanities,
Science & Engineering Research (CHUSER 2012), 3–4
December 2012, Kota Kinabalu, Malaysia, pp 32–37.
Jiang, W & Wong, J K W 2016. Key activity areas
of corporate social responsibility (CSR) in the
construction industry: A study of China. Journal of
Cleaner Production, 113: 850–860.
Khalfan, M, Noor, M A, Maqsood, T, Alshanbri,
N & Sagoo, A 2015. Perceptions towards
sustainable construction amongst construction
contractors in the State of Victoria, Australia.
Journal of Economics, Business and Management,
3(10): 940–947.
Kibert, C J 2007. The next generation of sustainable
construction. Building Research & Information,
35(6): 595–601.
Kibert, C J 2016. Sustainable Construction: Green
Building Design and Delivery, 4th ed. New York:
Wiley.
Klaniecki, K, Wuropulos, K & Hager, C P 2018.
Behaviour change for sustainable development. In
 Leal Filho, W (Ed). Encyclopedia of Sustainability
in Higher Education. Cham, Switzerland: Springer
International, pp 1–10.
Lim, B TH, Liu, A Z J & Oo, B L 2019. Awareness and
practices of sustainable construction in Australia:
Consultant quantity surveyors’ perception. AIP
Conference Proceedings, 2124: 020060.
Marchman, M & Clarke, S N 2011. Overcoming the
barriers to sustainable construction and design
through a cross-reference of West Coast Practices.
Proceedings, 47th ASC Annual International
Conference, 6–9 April 2011, Omaha, Nebraska.
Michie, S, Johnston, M, Abraham, C, Lawton, R &
Parker, D 2005. Making psychological theory
useful for implementing evidence based practice: a
consensus approach. Quality and Safety in Health
Care, 14: 26–33.
Michie, S, Van Stralen, M M & West, R 2011. The
Behaviour Change Wheel: A new method for
characterising and designing behaviour change
interventions. Implementation Science, 6(1): 42.
Michie, S, Atkins, L & West, R 2014. The Behaviour
Change Wheel: A Guide to Designing Interventions,
1st ed. London: Silverback. Available at: www.
behaviourchangewheel.com.
Mousa, A 2015. A business approach for
transformation to sustainable construction: An
implementation on a developing country. Resources,
Conservation and Recycling, 101: 9–19.
Ofori, G 2015. Nature of the construction industry, its
needs and its development: A review of four decades
of research. Journal of Construction in Developing
Countries, 20(2): 115–135.
Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021
Opoku, A & Ahmed, V 2014. Embracing sustainability
practices in UK construction organizations:
Challenges facing intra-organizational leadership.
Built Environment Project and Asset Management,
4(1): 90–107.
Oxman, A D, Thomson, M A, Davis, D A & Haynes,
R B 1995. No magic bullets: A systematic review of
102 trials of interventions to improve professional
practice. Canadian Medical Association Journal,
153(10): 1423–1431.
Pham, H, Kim, S Y & Luu, T Van 2019. Managerial
perceptions on barriers to sustainable construction
in developing countries: Vietnam case. Environment,
Development and Sustainability, 22: 2979–3003.
Pitt, M, Tucker, M, Riley, M & Longden, J 2009.
Towards sustainable construction: Promotion
and best practices. Construction Innovation,
9(2): 201–224.
Saleh, M S & Alalouch, C 2015. Towards sustainable
construction in Oman: Challenges and
opportunities. Procedia Engineering, 118: 177–184.
Sebake, T N 2008. Limitations of implementing
sustainable construction principles in the
conventional South African design approach.
Proceedings, 3rd Built Environment Conference,
6–8 July 2008, Cape Town, pp 479–491.
Serpell, A, Kort, J & Vera, S 2013. Awareness, actions,
drivers and barriers of sustainable construction in
Chile. Technological and Economic Development of
Economy, 19(2): 272–288.
Shari, Z & Soebarto, V I 2012. Delivering sustainable
building strategies in Malaysia : Stakeholders’
barriers and aspirations. Alam Cipta, 5(2): 3–12.
Shi, Q, Zuo, J, Huang, R, Huang, J & Pullen, S
2013. Identifying the critical factors for green
construction: An empirical study in China. Habitat
International, 40: 1–8.
Simpeh, E K & Smallwood, J J 2015. Factors influencing
the growth of green building in the South African
construction industry. Proceedings, Smart and
Sustainable Built Environment Conference, 9–11
December 2015, Pretoria.
Suliman, L K M & Omran, A 2009. Sustainable
development and construction industry in Malaysia.
Manager Journal, 10(10): 76–85.
UNEP (UN Environment Programme) 2017.
Consuming differently, consuming sustainably:
Behavioural insights for policymaking. Nairobi,
Kenya: UNEP.
VandenBos, G R 2007. APA Dictionary of Psychology.
Washington, DC: American Psychological
Association.
Wensing, M, Van Der Weijden, T & Grol, R 1998.
Implementing guidelines and innovations in general
practice: Which interventions are effective? British
Journal of General Practice, 48: 991–997.
Wilson, I E & Rezgui, Y 2013. Barriers to
construction industry stakeholders’ engagement
with sustainability: Toward a shared knowledge
experience. Technological and Economic
Development of Economy, 19(2): 289–309.
Windapo, A O 2014. Examination of green building
drivers in the South African construction
industry: Economics versus ecology. Sustainability,
6: 6088–6106.
23
TECHNICAL PAPER
An investigation of South
African low-income housing
Journal of the South African
Institution of Civil Engineering

roof anchor systems

ISSN 1021-2019
Vol 63 No 4, December 2021, Pages 24–34, Paper 1280

RUBEN VAN DER MERWE is a Candidate Civil

Engineer who specialises in the field of R Van der Merwe, J Mahachi
structural engineering. He obtained his BIng
Civil and part-time MEng Structural (with
distinction) at the University of
Johannesburg in 2015 and 2020, The current roof anchoring methods for houses in South Africa are described in the standard
respectively. His main research interests are
SANS 10400. The adequacy of these prescribed roof anchoring methods for low-income houses
in the field of steel construction. He is
currently employed as Design Engineer at Structa Konsult, situated in the (LIHs) is questionable, due to several recently reported failures caused by strong winds. This study
province of Gauteng. quantitively investigates the performance of the prescribed roof anchoring methods through
Contact details:
numerical methods, and focuses on LIHs with light-weight roofs supported on single-leaf masonry
Department of Civil Engineering walls. The masonry walls comprise either solid bricks or hollow cement blocks. The peak wind
University of Johannesburg reaction forces that are expected to occur at the roof anchor systems were determined through a
PO Box 524
series of static analyses. Finite element analysis techniques were performed to predict the capacity
Auckland Park 2006
South Africa of the roof anchor systems. The predicted resistance of the roof anchor systems was compared
T: +27 84 249 9538 to the calculated peak wind reaction forces at the roof anchors to determine the adequacy of
E: rubenvandermerwe24@gmail.com the roof anchor systems. The results of the research suggest that, for LIHs constructed from solid
bricks, the prescribed roof anchor systems perform poorly under the expected South African
DR JEFFREY MAHACHI is the Head of School
strong wind climate. The results indicated that the roof anchors will pull out at peak basic wind
of Civil Engineering and the Built
Environment at the University of speeds of between 27 m/s and 32 m/s, and cracks will develop in the masonry prior to anchor
Johannesburg, and also a Director of the pull-out. This study suggests that further research is required to develop adequate roof anchoring
Sustainable Materials and Construction methods for LIHs with light-weight roofs, supported on solid brick walls, and that the relevant
Technologies (SMaCT) Research Centre. He
codes should be amended accordingly. Furthermore, the results showed that the prescribed
obtained his PhD in Civil Engineering from
the University of the Witwatersrand, and his roof anchor systems for LIHs constructed from hollow blocks performed well and were able to
MSc in Structural Engineering from the University of Surrey, United withstand the expected wind loads under the South African strong wind climate.
Kingdom. He has previously held several executive positions at the
National Home Builders Registration Council and worked as a Research
Engineer at the CSIR. His research interests are in disruptive construction
technologies and structural risks. introduction Blown-off roofs of LIHs during strong
Contact details:
South Africa has a huge housing backlog winds have been recognised in a study by
School of Civil Engineering and the Built Environment which is escalating continually. In 2018 the Ogunfiditimi (2008). The roofs in such
University of Johannesburg backlog was more than 2.3 million (Mahachi an event become flying objects with the
Johannesburg
2021). To eradicate the housing backlog, potential of causing injury to (or death of)
South Africa
T: +27 82 904 9569 the South African government provides a humans and damage to surrounding build-
E: jmahachi@uj.ac.za housing subsidy to households earning less ings and infrastructure. Also, the roofs and
than R3 500 per month. The subsidies are support walls rely on each other for lateral
used to construct houses commonly referred support. Failure of the roof, therefore, also
to as low-income houses (LIHs). Typically, compromises the structural integrity of the
such houses have a footprint of 36‒60 square support walls. Figure 1 (Goliger 2002) shows
metres, and the roofs are constructed with an illustrative example of roof and wall
lightweight corrugated sheets (0.5 mm damage of a LIH caused by strong winds.
thick) anchored to the walling system. The Due to the likelihood of roofs blowing off
structural integrity of South African LIHs during strong winds, many informal shelters
is often reported as substandard, mainly as and LIHs use objects such as tyres, bricks,
a result of poor workmanship and the use of or other heavy objects to assist in securing
poor-quality materials, as reported by many the roof. Figure 2 (Mabuya & Scholes 2020)
inhabitants (Zunguzane et al 2012). shows an example of such a case.
About five devastating wind events The current methods for anchoring
occur annually in South Africa, causing roofs are prescribed in the South African
damage to structures and leaving thousands Standard SANS 10400-K (SANS 2015). A
of people homeless, as cited in Goliger schematic illustration of the prescribed
(2002). Mahachi et al (2018) suggest that roof anchor methods for single-leaf mason-
wind-induced damage to non-engineered ry walls is provided in Figure 3.
Keywords: finite element analysis, low-income housing, masonry, structures, such as LIHs, in most cases is This research aims to determine the
roof anchors, wind damage caused by wind speeds of about 30 m/s. adequacy of the currently prescribed roof

Van der Merwe R, Mahachi J. An investigation of South African low-income housing roof anchor systems.
24 J. S. Afr. Inst. Civ. Eng. 2021:63(4), Art. #1280, 11 pages. http://dx.doi.org/10.17159/2309-8775/2021/v63n4a3
Figure 1 D
 amage of gable wall and failure of
a roof caused by inadequate roof
anchorage (Goliger 2002 – permission Figure 2 I llustration of heavy objects used to secure a roof of a LIH against wind uplift (Mabuya &
for re-use of photo granted) Scholes 2020 – photo reproduced here under the Creative Commons Attribution Licence)

anchors for South African LIHs under a

strong wind climate and provides recom- Anchor strap
(30 mm × 16 mm) Anchor strap
mendations for further research on the (30 mm × 16 mm)
topic. The aim was achieved through the
following objectives:
QQ To determine the expected wind uplift
Embedment height = 600 mm

Roof support
reaction forces at the roof anchor sys- Roof support

Embedment length =
tem through static analyses.

minimum 600 mm
QQ To predict the roof anchor system’s per-
formance against wind uplift through
finite element analysis (FEA).
QQ To determine the roof anchor system’s
adequacy by comparing the peak reac-
tion loads expected at the roof anchors
to the predicted resistance offered by
the roof anchor systems.
Anchor embedded at least 70 mm Cores wit concrete infill
into masonry joint
(a) (b)
Methodology
The study was conducted through a quanti- Figure 3 P
 rescribed roof anchoring methods for buildings with light roofs supported on single-
tative research approach. The methodology leaf masonry: (a) solid-brick walls, (b) hollow-block walls (redrawn from SANS 2015)
of the research is divided into two main
parts. The first part concerns the predic-
tion of the peak wind reaction forces that Numerical investigation of the roof anchor system's performance
are expected to occur at the roof anchor
systems. The second part concerns the pre- Determine wind uplift reaction Predict the roof anchor Define parts
diction of the roof anchor system’s response. forces at roof anchors system's capacity
Define materials
Figure 4 shows a flow diagram that outlines
Model typical roofs of LIH: FEA of roof anchor system:
the methodology of the research. A detailed QQ 1.0 m roof support spacing (Abaqus) Define assembly
description of the steps is provided under
QQ 1.2 m roof support spacing QQ Solid brick walls
Define contact
QQ 1.5 m roof support spacing QQ Hollow-block walls

the sub-headings that follow. Define mesh

Detemine wind pressure on the roof:
Define constraints
(SANS 10160-3: 2018) Observe state Extract load-
Determination of the wind QQ Buildings based on typical LIH
of masonry displacement
uplift reaction forces at the dimensions components diagrams
QQ “Terrain Category C”
roof anchor system QQ 0 m altitude above mean sea level

The reaction forces at the roof anchor sys- QQ Disregard topography effects

QQ Peak basic wind speeds: 25, 32, 36, Determine ultimate capacity
tems were determined through a static FEA. 40, 44 m/s of roof anchor system
Four distinct models representing typical
LIHs with different roof configurations were Static linear analyses:
(Autodesk Robot Structural Analysis)
defined. The models were designated as
Determine peak reaction forces for
Model 1 to Model 4. The dimensions of the various roof support spacings and range
houses were kept constant, while the posi- of peak basic wind speeds considered
tions of the external openings, representing
the front door and windows, were altered. Determine the roof anchor systems’ adequacy

The geometry of each model was based on

typical LIHs in South Africa, as illustrated Figure 4 Outline of research methodology

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 25
in Figure 5. The regulations from SANS
Ro g
10400-K (SANS 2015) were utilised as a of c in
sup spa
guideline to determine the dimensions of po
rt s p or t
pa sup
the LIH models. It was assumed that the c in of
g Ro
front wall (for a single-storey house) will
have openings of more than 15% of the total

2.7 m
2.7 m
wall area and is therefore limited to a 6 m
long by 2.7 m high wall panel for a 140 mm
thick masonry wall. The sizes of the exter- 6m 6m
6m 6m
nal walls were kept constant and hence
resulted in a 6 m long by 6 m wide house (a) (b)
with a height to the eaves of 2.7 m. Truss or Ro
of Wind 180°
rafter spacings of 1.0 m, 1.2 m, and 1.5 m sup c in g
po pa
rt s r ts
were considered for each model. These spac- pa po
c in p
ings were chosen so that the length of the g su
of
Ro
support walls is divisible by the roof spacing
2.7 m

2.7 m
and, except for 1.5 m spacing, are within the
range of the regulations of SANS 10400-L
(SANS 2011), since the regulations limit the
roof spacing to 1.4 m. The roof slopes were 6m 6m
6m 6m
taken as 12.5⁰ and 7⁰ for the duo-pitch and
mono-pitch roofs, respectively. The roofs of (c) (d)
the LIH models were modelled and analysed Wind 0° Wind 90°
in Autodesk Robot Structural Analysis soft-
ware due to its efficiency (in terms of time Figure 5 W
 ind loading directions and geometry of LIH models: (a) Model 1, (b) Model 2,
and effort) in simulating pressures on the (c) Model 3, and (d) Model 4
roofs and determining the reaction forces.
The roof anchor systems were resembled by
defined support constraints. Resultant peak +6.00
+6.00
reaction forces were obtained for combined
+4.50 +4.50
uplift wind loading and the self-weight of +3.00 +3.00
the roof structure and cladding. +1.50 +1.50
The wind loading that acts on the ±0.00 ±0.00
roofs was calculated per the South African ±0.00 +0.67
Standard SANS 10160-3 (SANS 2018). The +0.34 +0.34
wind direction and wind pressure calcula- ±0.00 ±0.00
+6.00 1
+6.00
tions were based on the following parameters: +4.50 +4.50
QQ The peak wind pressures were calcu- +3.00
+1.50
lated for a range of 1 in 50-year basic
±0.00 ±0.00
wind speed values provided on the wind
map in SANS 10160-3 (SANS 2018),
i.e. 32 m/s, 36 m/s, 40 m/s and 44 m/s. Figure 6 Illustrative example of net wind pressures applied to a roof structure
Also, a basic wind speed of 25 m/s was
introduced to broaden the range of wind Predicting the response The modelling approach herein accounted
speeds of the investigation. and resistance of the for unit-mortar bond failure through defined
QQ The terrain category defined as roof anchor systems surface contact interaction. Crushing and
“Category C” per SANS 10160-3 (SANS The resistances of the roof anchor systems cracking of the masonry units and the
2018) was adopted to determine the were predicted through FEAs conducted mortar were accounted for by defined non-
peak wind pressures. in Abaqus software. Abaqus software was linear behaviour of the Finite Elements (FEs)
QQ Topography effects were ignored. used due to its capability to model and through the Concrete Damaged Plasticity
QQ An accidental load case was consid- solve complex engineering structures such (CDP) model. In this approach, a single part,
ered by assuming a dominant wall in as micro-modelling of masonry structures referred to as a Representative Element (RE),
the case where the openings on the with contact simulation, as required for was defined to represent a masonry unit, as
windward wall of the house are open this study. The roof anchor systems were well as one bed joint below and half the thick-
through the failure of the door and modelled and analysed to capture the ness of the head joints on either side of the
windows. capacity thereof due to simulations of the masonry unit. The FE modelling approach is
Figure 6 shows an illustrative example force applied, in turn due to uplift wind schematically presented in Figure 7. Through
of the roof structure and wind pressure pressure on the roof. The pull-out capacities this modelling approach, the amount of
symbols, for Model 1, as modelled in Robot of the roof anchor systems were determined contact is significantly reduced without com-
Structural Analysis software for wind at 0°. through load-displacement diagrams. promising the accuracy of the model.

26 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
 Modelling of roof anchor systems
Unit-mortar nonlinear The roof anchor systems concern the roof
interface defined through
Masonry unit 3D anchoring methods per SANS 10400-K
surface-surface contact
FE with nonlinear
behaviour (SANS 2015) for single-leaf masonry walls.
The dimension of the wall assembly was
chosen to accommodate a result compari-
son to potential future physical testing.
The assembly of the FEA model for a
roof anchor system consisting of a solid
brick masonry wall is illustrated in Figure 8.
A wall with dimensions of 1 500 mm long,
700 mm high, and 140 mm thick was mod-
elled. The parts were modelled using explicit
Representative Element Mortar 3D FE with
3D hexahedral-shaped, eight-node, linear
nonlinear behaviour brick elements with reduced integration and
hourglass control (type C3D8R elements).
Figure 7 Approach adopted for FE modelling of masonry components The anchor was embedded between
two brick courses at the prescribed depth
of 600 mm from the top of the wall. The
5 mm Head joint (mortar) 300 mm brick situated on top of the anchor was
both sides of masonry unit modified by subtracting a region to suit the
anchor embedment and is referred to as the

100 mm
Masonry unit
modified RE. The bonding of the anchor to
the mortar was simulated through defined
10 mm Bed joint (mortar)
surface-surface contact.
m
140 m The bottom surface at the base of the
Brick part
(Representative Element) wall was fully constrained against transla-
600 mm

Defined velocity constraint

tion and rotation. A velocity constraint was
applied to a reference point (RP), defined
30 m Bed joints at the top of the anchor, to simulate an
m
upwards ‘’pulling’’ force. The imposed
m Head joints velocity produced a reaction force at the
70 m
RP. Also, the RP was used to capture the
1.6 mm thick anchor strap Encastre constraint
magnitude of the vertical (uplift) reaction
Anchor embedment location
force throughout the loading period.
Assembly The mesh defined for the FEA model is
presented in Figure 9. The brick elements in
the vicinity surrounding the anchor embed-
ment are expected to experience high-stress
Section subtracted from brick to suit gradients due to the load being transferred
anchor embedment (70 mm long × from the anchor to the wall through the
30 mm wide × 1.6 mm high) localised region. A refined mesh was
Modified brick part
(Representative Element) introduced in the areas where high-stress
gradients were expected. The mesh size was
Figure 8 Assembly of FEA model representing the roof anchor system consisting of solid bricks gradually increased towards the outer edges
of the wall, where less focus was required, to
effectively reduce the computational time.
0.8 mm Mesh size defined for anchor strap Figure 10 shows a section through the
wall indicating the modelling detail and
meshing of the anchor embedment. A
refined mesh was applied to the surfaces of
45 mm Mesh size the anchor strap involved in contact pair
25 mm Mesh size simulation with surfaces of the masonry
components and was defined as the slave
10 mm Mesh size
surface in the contact pair interactions.
5 mm Mesh size A similar modelling approach was fol-
lowed for the hollow-block masonry walls.
A wall of 1 650 mm long, 800 mm high and
140 mm thick was modelled as illustrated in
Figure 9 Mesh definition of the FEA model for the solid-brick wall Figure 11. A straight anchor was embedded

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 27
 600 mm into a concrete infill block with a
70 mm Embedment depth hole subtracted through the block to suit the
Anchor strap anchor embedment. The anchor bonding was
simulated through a defined surface-surface
contact interaction. The concrete infill block
was placed continuously into the cores of the
Modified brick
hollow blocks with cores that are vertically
aligned near the roof support and connected
Relatively coarser mesh for through surface-surface contact interaction.
master surface of brick part Figure 12 shows a section through the wall
Refined mesh for slave indicating the anchor embedment detail of
surface of anchor strap the FEA model. The mesh definition of the
Y
assembly model is illustrated in Figure 13.
Z X
Coarser mesh for master For the FE modelling of the roof anchor
surface of brick part systems, the brick-mortar interfacial bond
properties are required, as well as the
Figure 10 S ection through a modified brick part showing the modelling detail for anchor mechanical properties of the masonry
embedment of a roof anchor system consisting of solid bricks units and mortar. For this study the
interfacial bond properties are particularly
important, since the loading of the roof
mm 90 anchors are expected to induce mainly
300 m mm
.5 m tensile and shear stresses on the masonry
107 5 mm mortar
Region subtracted through head joint wall. The bond strength of masonry is
concrete infill “block”
affected by various factors pertaining to
200 mm

25 mm flange masonry units, mortars and construction

practice. The combined engineering prop-
140 10 mm thick × 25 mm erties of masonry, therefore, are scattered
mm wide mortar joint
Anchor strap with over a broad spectrum. Approximately
velocity constraint
Hollow-block unit realistic values for material properties
(Representative Element) should, however, be selected to ensure that
15 MPa concrete infill “block” the self-weight load effects and the overall
stiffness are of the correct order. Masonry
subjected to tensile and shear most often
fails through the brick-mortar interface,
especially in low bond strength masonry
(Sarhosis et al 2015), which may be
Encastre constraint expected for this study due to the quality
Y concerns of South African LIHs. The joint
X
Z Assembly interface properties are, therefore, regarded
to be of greater importance than the mate-
Figure 11 A
 ssembly of FEA model that represents the roof anchor system consisting of hollow blocks rial properties of the units and mortar.

Anchor strap 5 mm Mesh on anchor straps 15 mm Mesh on concrete block

600 mm Embedment depth

Cores of hollow 25 mm Mesh on

blocks filled hollow-block units
with concrete
for fixing of
anchot strap

Z X

Y
Figure 12 S ection presenting the nature of X
the anchor embedment for the roof Z
anchor system consisting of hollow
blocks Figure 13 M
 esh defined for FEA model of roof anchor system consisting of hollow blocks

28 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Table 1 Density, Young’s Modulus and Poisson’s Ratio of masonry constituents
Density (kg/m3)
Material
group Masonry
Mortar
units
G1 2 090 2 200
G2 2 200 2 200
G3 1 430 2 400
G4 1 510 1 860
References: 
In this study, four groups of materials
with distinctive mechanical and interfacial
bond properties for brick units and mortar
were defined in the FEAs to consider the
effect of the material properties on the
performance of the roof anchor systems.
The four material groups designated as
“G1” to “G4” were obtained from previous
studies. Where some required parameters
were not reported in the relevant sources,
comparable values were assumed to values
obtained from alternative sources.
The first material group (G1) consisted
of solid concrete masonry units (com-
monly used to construct South African
LIHs) and the parameters of the units were
obtained from a study by Fourie (2017).
The Poisson’s Ratio of the masonry units
obtained from Fourie (2017) was unreli-
able, therefore a more realistic value was
assumed. Fourie (2017) only focused on the
material parameters of the masonry units,
therefore reasonable values for the Young’s
Modulus, Poisson’s Ratio and density of the
mortar were assumed for material group
“G1”. The second material group (G2) con-
sisted of hollow concrete blocks obtained
from a study by Santos et al (2017). The
density and the Poisson’s Ratio of the
Young’s Modulus (MPa) Poisson’s Ratio
Masonry Masonry
Mortar
units units
17 900 4 500 a 0.150 b
35 370 9 290 0.150 b
2 170 7 760 0.155
6 540 1 550 0.160
a: Oliveira et al (2009) b: Angelillo (2014)
masonry units and mortar were not report-
ed by Santos et al (2017) and were therefore
assumed. The parameters of the third and
fourth material groups, designated as “G3”
and “G4”, respectively, were obtained from
Agüera et al (2016) and the masonry units
consist of soil-cement blocks.
The values for the densities, Young’s
Modulus and Poisson’s Ratio of the mason-
ry units and mortar for the material data
sets are presented in Table 1. References are
provided for the values not provided by the
relevant sources. The unknown material
densities of masonry units and mortar were
assumed a reasonable value of 2 200 kg/m3.
For the simulation of material hardening
and softening in the masonry FE model, the
behaviour (stress-strain relationship) beyond
the elastic range of the masonry units and
mortar is required. Whilst the non-linear
behaviour of the masonry units and mortar
of the material groups “G1” and “G2” was
not reported in their respective sources,
Santos et al (2017) explained that the com-
plete stress-strain curves for the masonry
constituents of material group “G2” were
determined for tension and compression
from methods for estimating the stress-
strain curves of plain concrete as proposed
by Zhenhai (2014). These methods were
used in their study to simulate testing of
masonry prisms through FE modelling. The
Mortar same methods were adopted in this study
to estimate the non-linear behaviour of
0.150 b the masonry units and mortar for material
0.150 b groups “G1” and “G2”. The resultant stress-
strain curves are presented in Figure 14. For
0.210
detailed information on the methods used
0.220 in determining the stress-strain curves, the
reader is referred to Zhenhai (2014). The
non-linear material properties of the mason-
ry units and mortar for material groups
“G3” and “G4” were adopted from the values
reported by Agüera et al (2016).
The parameters adopted for the CDP
model that were used to describe the
behaviour within the plasticity model were
obtained from Santos et al (2017). The
parameters are described as follows:
QQ Dilation angle (ψ): This parameter is
defined by the internal friction of the
concrete. The suggested value of 36⁰ was
adopted (Kmiecik & Kaminski 2011).
QQ Ratio between yield strength in biaxial
state (σ b0) and yield strength in uni-
axial state (σc0): This ratio (σ b0 / σc0)
was assigned a suggested value of 1.16
(Abaqus 2017).
QQ Viscosity parameter (μ): This parameter
defines damage stabilisation that is
generally used in implicit models with
convergence difficulties. A zero value
was adopted.
QQ Eccentricity parameter (ρ): The param-
eter obtained from the DP model defines
the yield surface in the meridian plane. A
value of 0.1 was adopted for this study.
The steel defined for the anchor strap
in the FEA model has a yield strength of
220 MPa and ultimate strength of about
375 MPa. The stress-strain relationship of

40 3.5
35 3.0
Compressive stress (MPa)

Compressive stress (MPa)

30
2.5
25
2.0
20
1.5
15
1.0
10
5 0.5

0 0
0 5 10 15 20 0 0.25 0.50 0.75 1.00
(a) Strain (mm/mm) × 10 –3 (b) Strain (mm/mm) × 10 –3
G1 – Units G1 – Mortar G2 – Units G2 – Mortar G1 – Units G1 – Mortar G2 – Units G2 – Mortar

Figure 14 Stress-strain curves for masonry units and mortar of material groups “G1” and “G2” as determined from the methods proposed by Zhenhai
(2014) for (a) compression and (b) tension

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 29
Table 2 S tress-strain data defined for steel Table 3 Cohesive properties defined for masonry joints per material group
anchor strap (Qi et al 2005) Cohesive properties of joints
Material group
Stress-strain data for the Knn (N/mm3) K ss (N/mm3) Ktt (N/mm3)
steel anchor strap
G1 4 500 1 960 1 960
Stress (MPa) Strain (mm/mm × 10 –3)
G2 9 290 4 040 4 040
220 0
G3 7 760 3 210 3 210
250 5.94
G4 1 550 640 640
300 16.74

332 22.00 Table 4 Damage interface properties defined per material group
353 25.75
Quadratic damage properties with mixed-mode evolution
370 51.50 Material
group tn ts tt GI GII GIII
η
374 105.20 (MPa) (MPa) (MPa) (N/mm) (N/mm) (N/mm)

G1 0.25 a 0.17 0.17 0.012 b 0.028 b 0.028 b 2

G2 1.47 0.23 0.23 0.012 b 0.028 b 0.028 b 2
the steel was adopted from Qi et al (2005)
as presented in Table 2. G3 0.20 a 0.30 c 0.30 c 0.012 b 0.028 b 0.028 b 2
The unit-mortar bond interface
G4 0.20 a 0.30 c 0.30 c 0.012 b 0.028 b 0.028 b 2
between adjacent REs was defined by
References: a: Reddy & Gupta (2006) b: Angelillo (2014) c: BS (2005)
surface-surface contact interaction through
the following contact relationships:
QQ Tangential behaviour to accommodate
friction in masonry joints 0.40
QQ Hard contact (prevents penetration and
0.35
allows for separation of surfaces)
QQ Cohesive behaviour to accommodate 0.30
Traction stress (MPa)

joint stiffness 0.25

QQ Damage initiation and evolution to
accommodate stiffness degradation.
The tensile and shear response of the joints
was defined through traction separation laws.
A friction coefficient of 0.75 was defined
as adopted from Abdulla et al (2017). A
dummy stiffness was defined to represent
the elastic response of the masonry joint
interface and to prevent interpenetration of
the REs. The Young’s Modulus of the mortar
for each material group was adopted as the
elastic response of the masonry joints in the
normal (tensile) direction, while the elastic
response of the joint interface in the first
and second shear directions is equal to the
Shear Modulus of the mortar. The masonry
joint stiffness (K) for each material group is
summarised in Table 3. The subscripts ‘’n’’,
‘’s’’, and ‘’t’’ represent the normal, first-, and
second shear directions, respectively.
Damage initiation of the joint inter-
faces, as summarised in Table 4, was
defined through a quadratic stress criterion
that follows the initial linear response. The
traction at damage initiation is defined
by the parameters designated as ‘’t’’. The
fracture energy of the masonry joints was
defined through the parameters ‘’GI’’, ‘’GII’’,
and ‘’GIII’’ that represent the fracture ener-
gies in the normal, first and second shear
directions, respectively. The critical mixed-
mode fracture energy is obtained from
0.20
0.15
0.10
0.05
0
0 25
Van der Pluijm (1993)
Figure 15 C
 omparison between numerical (red line) and experimental (grey envelope) results for
small-scale masonry specimens subject to direct tension
the Benzeggagh-Kenane (BK) law with an
exponent (η) taken as 2 by assuming brittle
behaviour (Abdulla et al 2017).
Note that the interfacial bond strength
of material group “G2” relates to concrete
hollow-block masonry, while the interfacial
bond strengths of material groups “G1”, “G3”
and “G4” were obtained from a study con-
ducted on the bond strength of soil-cement
masonry (Reddy & Gupta 2006), and the
values are fairly lower (0.2 ‒ 0.25 MPa com-
pared to 1.47 MPa). Despite the lower bond
strength values being based on soil-cement
blocks, they were also utilised in the hollow-
block FE models. This is deemed reasonable,
as it can be regarded as the equivalent of
accounting for poor quality mortar or work-
manship, resulting in a weak interface.
50 75 100 125
Joint crack opening (mm)
Numerical simulation
A friction interaction with a coefficient
of friction of 0.5 was assigned between the
steel and the masonry wall that is in range
with the values suggested by Rabbat and
Russell (1985) for friction between steel
and concrete. The bonding of the anchor
embedded in the masonry joint was defined
through a contact interaction property. The
properties were adopted from Rabbat and
Russell (1985) and were defined as follows:
QQ Tangential behaviour to accommodate
friction between the anchor mortar
joint with a friction coefficient of 0.5.
QQ Cohesive behaviour to accommodate
anchor bond stiffness. The stiffness
value is based on the underlying ele-
ments by using the default value of
Abaqus.

30 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
 QQ Damage initiation and evolution that
2.5 represent the degradation of anchor
Van der Pluijm (1993) σ bond strength. A bond strength of
Numerical simulation
0.39 MPa and fracture energy of 25 N/m
2.0
were defined.
Shear stress (MPa)

1.5 Validation of modelling approach

σ Small-scale FEA models were used to verify
the behaviour of the modelling strategy by
1.0 comparing the result with experimental
σ = –1.0 MPa
results obtained by Van der Pluijm (1992).
Figures 15 and 16 show the comparison
0.5 σ = –0.5 MPa between the numerical and experimental
σ = –0.1 MPa
results for the tensile and shear behaviour of
the masonry joints, respectively.
0
0 0.2 0.4 0.6 0.8 1.0 A small-scale FEA that simulates an
Joint slip (mm) anchor pull-out test was conducted to vali-
date the contact interaction that represents
Figure 16 C
 omparison between numerical (solid lines) and experimental (dotted lines show a the anchor bond ‒ the horizontal section
range of values) results for small-scale masonry specimens subject to shear for three of the anchor was modelled and embedded
values of initial compressive stress 0.1, 0.5 and 1.0 MPa 70 mm between two masonry units by adopt-
ing the material properties of material group
“G1”, and a horizontal “pulling” force was
900 applied to the anchor. The peak force that
800 will cause the anchor bond to fail (referred
to as the anchor pull-out resistance) is
700 F determined from the horizontal force versus
Applied horizontal force (N)

600 horizontal displacement diagram extracted

from the FEA software. The results indicate
500
a peak horizontal force of about 830 N, as
400 presented in Figure 17.
300 Hand calculations were used to validate
the resistance of the defined contact inter-
200
action. An anchor pull-out resistance (HR)
100 of 819 N was calculated per Equation 1,
which is in close agreement with the result
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 obtained from Figure 17.
Horizontal displacement (mm)
HR = c × Ab(1)
Figure 17 S mall-scale anchor pull-out test to validate anchor bond simulation
Where:
HR is the determined peak resistance (hori-
6 000 zontal pulling force) given in Newtons
c is the cohesion or bond strength defined
5 000 as 0.39 MPa (Rabbat & Russell 1985)
Applied reaction force (N)

Ab is the calculated area of the bond surface

4 000 that is 2 100 mm2 (70 mm × 30 mm)

3 000

Results and discussion

2 000

Reaction forces determined

1 000
at the roof anchors
Figure 18 shows the maximum reaction
0
1.0 m 1.2 m 1.5 m forces obtained at the roof anchor systems
Roof support spacing for the basic wind speeds considered at
25 m/s 32 m/s 36 m/s 40 m/s 44 m/s 1.0 m, 1.2 m, and 1.5 m roof support spac-
ing. The maximum reaction forces were
Figure 18 V
 ertical reaction forces obtained at the roof anchors at the roof support spacings, and obtained from the range of LIH models that
peak basic wind speeds considered were considered in this study.

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 31
The response of the roof anchor
systems consisting of a solid brick wall 3 500
The load-displacement curves presented in
Figure 19 show the applied force versus the 3 000
free-end slip of the solid brick wall’s anchor
strap for the material groups defined in this
2 500
study. The ultimate capacity of the roof
anchor system is represented by the maxi-

Vertical reaction force (N)

mum reaction force obtained throughout 2 000
the loading history.
From Figure 19 it is evident that the 1 500
limiting resistance of about 2 100 N cor-
responds to material data groups ‘’G2’’
and ‘’G3’’. The ultimate failure of the roof 1 000

anchor system was achieved through

pull-out of the anchor strap as the bond 500
strength was overcome. The free-end slip
of the anchor is demonstrated in Figure 20.
0
It was further noted that the resistance of 0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
the roof anchor system is affected by the Anchor free-end slip (mm)
material properties of the masonry constit- G1 G2 G3 G4
uents and the brick-mortar bond strength.
The damage variable was used to cap- Figure 19 R
 esponse of the roof anchor system for the material groups defined in this study
ture the degree of damage in the elements
of the masonry brick units and mortar
throughout the loading history. Figure 21
shows the results of material group “G1” at
the instance of roof anchor pull-out.
The results show that material damage
in the masonry is present at the instant of
roof anchor pull-out. This suggests that
cracks in the masonry wall will occur
before anchor pull-out and damage are
visible on the masonry walls. A maximum
damage variable of 0.56 was observed. This
suggests that the masonry is partially dam-
aged while the anchor strap is still fixed.
The results agree with the nature of dam-
age presented in Figure 1.

The response of the roof Y

anchor systems consisting Z X Anchor free-end slip
of a hollow-block wall
The load-displacement curves presented Figure 20 S ection indicating the anchor free-end slip resulting from an anchor bond failure
in Figure 22 show the applied force versus
the free-end slip of the hollow-block wall’s
anchor strap.
SDEG
Material groups ‘’G1’’ and ‘’G4’’ failed (Avg. 75%)
+5.562e-01
through masonry joint failure. Figure 23(a) +5.098e-01
+4.635e-01
illustrates a section through the wall that +4.171e-01
+3.708e-01
indicates the masonry joint failure. The +3.244e-01
+2.781e-01
corresponding load-displacement curves +2.317e-01
+1.854e-01
from Figure 22 showed gradual stiff- +1.390e-01
+9.270e-02
ness degradation as failure occurs in the +4.635e-02
+0.000e+00
masonry joints. The curves flattened out
after the peak capacity of the masonry
Y Step: Step-1
joints was attained. Z
Increment 27966: Step Time = 9.0001E-02
X Primary Var: SDEG
Material groups ‘’G2’’ and ‘’G3’’ failed Deformed Var: U Deformation Scale Factor: +1000e+00

through anchor pull-out, the correspond-

ing load-displacement curves from Figure 21 Damage variable in brick and mortar recorded at the instant of roof anchor pull-out

32 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering

Vertical reaction force (N)
Figure 22 exhibited similar behaviour to
12 000 that of a pull-out test conducted for a bar
10 000 embedded in concrete, as proposed by
Zhenhai (2014). Figure 23(b) illustrates a
8 000 section through the wall that indicates the
6 000 free-end slip observed as the anchor pulls
out of the concrete infill.
4 000

2 000 Predicted adequacy of the

roof anchor systems
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Figure 24 presents the least resistance
Vertical displacement (mm) obtained from the FEA models, represent-
G1 G2 G3 G4 ing the capacity of roof anchor systems, as
well as the reaction forces determined at
Figure 22 Response of the roof anchor system consisting of hollow-block masonry walls for the the roof anchor systems against basic wind
material groups defined in this study speed for different roof support spacings.
The following observations were made
from the diagram:
QQ Failure of the roof anchor system for
solid brick walls occurs at a basic wind
speed of about 27 m/s for LIHs with
roof support spacings of 1.5 m.
QQ Failure of the roof anchor system for
solid brick walls occurs at a basic wind
Masonry joint failure speed of about 30 m/s for LIHs with
(a) roof support spacings of 1.2 m.
QQ Failure of the roof anchor system for
solid brick walls occurs at a basic wind
speed of about 32 m/s for LIHs with
roof support spacings of 1.0 m.
QQ Overall, the roof anchor system for solid
brick walls lacks sufficient resistance at
Anchor free-end slip the maximum basic wind speed that is
(b) expected to occur during the design life
of South African LIHs. The peak basic
Figure 23 F ailure mechanisms observed in hollow block walls: (a) material groups ‘’G1’’ and ‘’G4’’, wind speeds at failure agree with the
(b) material groups ‘’G2’’ and ‘’G3’’ suggestion that structures with a lack of
engineering input generally fail at wind
speeds of about 30 m/s as suggested by
8 000 Mahachi et al (2018).
QQ The results show that the resistance of
7 000
the roof anchor system for hollow-block
Resistance/reaction force (N)

6 000 walls provided sufficient resistance at

the roof anchors for the range of peak
5 000
basic wind speeds and roof support
4 000 spacings considered.

3 000

2 000
Conclusion
This paper investigated the structural
1 000 performance of the roof anchor systems for
LIHs with light roofs that are supported on
0
25 27 29 31 33 35 37 39 41 43 45 single-leaf masonry walls as prescribed by
Peak basic wind speed (m/s) SANS 10400-K (SANS 2015). The investiga-
1.0 m roof support spacing Resistance: solid brick wall tion was carried out through a quantitative
1.2 m roof support spacing Resistance: hollow-block wall research approach. FEA models were used to
1.5 m roof support spacing simulate the roof anchor system. The resist-
ance predicted from the FEA models was
Figure 24 P
 redicted resistance of the roof anchor systems compared to reaction forces expected compared to the expected reaction f­orces
at the roof anchor systems at the roof anchors that were determined

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 33
from a series of static analyses. The reaction
forces that are expected to occur at the roof
anchor systems are based on South Africa’s
strong wind climate. SANS 10160-3 (SANS
2018) was used to calculate the wind loading
on the roofs.
This study predicted that the prescribed
anchoring method used in conjunction with
single-leaf solid brick walls are inadequate
for LIHs and failures are likely to occur
during strong wind events. The results
predicted that roof anchor failures might
occur at basic wind speeds between 27 m/s
and 32 m/s for LIHs situated in areas with
regular vegetation or buildings, depending
on the roof support spacing. Failures might
occur at lower basic wind speeds for LIHs
located in open areas with low vegetation.
This study recommends further research
for developing robust roof anchor systems
for single-leaf solid brick walls, and that the
local codes be amended accordingly.
This study showed that the roof anchor
systems prescribed for hollow-block walls
performed well. The resistance of the roof
anchor systems provided sufficient resis-
tance against the expected reaction forces at
the roof anchors during strong wind events.
REFERENCES
Abaqus 2017. Abaqus documentation. Providence,
RI: Dassault Systèmes.
Abdulla, K F, Cunningham, L S & Gillie, M 2017.
Simulating masonry wall behaviour using a
simplified micro-model approach. Engineering
Structures, 151: 349–365.
Agüera, N, Tornello, M Frau, C D 2016. Structural
response of unreinforced masonry walls. Journal of
Civil Engineering and Architecture, 10(2): 219–231.
Angelillo, M 2014. Mechanics of masonry structures.
Udine, Italy: International Centre for Mechanical
Sciences.
34
BS (British Standard) 2005. BS EN, 1996-1-1 2005.
Eurocode 6: Design of Masonry Structures, Part1-
1: General Rules for Reinforced and Unreinforced
Masonry Structures. Brussels, Belgium: European
Committee for Standardization.
Fourie, J 2017. Characterisation and evaluation of the
mechanical properties of alternative masonry units.
MEng Dissertation. Stellenbosch University.
https://scholar.sun.ac.za/handle/10019.1/100880.
Goliger, A M 2002. Development of a wind damage and
disaster risk model for South Africa. PhD Thesis.
Stellenbosch University. https://scholar.sun.ac.za/
handle/10019.1/70419.
Kmiecik, P & Kaminski, M 2011. Modelling of
reinforced concrete structures and composite
structures with concrete strength degradation taken
into consideration. Archives of Civil and Mechanical
Engineering, 11: 623–636.
Mabuya, B & Scholes, M 2020. The three little
houses: A comparative study of indoor and
ambient temperatures in three low-cost housing
types in Gauteng and Mpumalanga, South
Africa. Environmental Research and Public
Health, 17(10): 3524. https://www.mdpi.
com/1660-4601/17/10/3524#.
Mahachi J, Bradley, R & Goliger, A 2018. Windstorm
damage to houses: Planning and design
considerations. Proceedings, Out-of-the-Box Human
Settlements Conference, October 2018, CSIR,
Pretoria.
Mahachi, J 2021. Development of a construction
quality assessment tool for houses in South
Africa. Acta Structilia, 28(1): 91‒116.
doi: 10.18820/24150487/as28i1.4.
Ogunfiditimi, O 2008. Assessment of structural quality
of houses delivered through the people’s housing
process in South Africa. MTech Dissertation.
University of Johannesburg.
Oliveira, A, Casali, J & Jr, L R P 2009. A study of the
structural behavior of masonry prisms and wallettes
built with a new type of concrete block. Proceedings,
11th Canadian Masonry Symposium, 31 May –
3 June 2009, Toronto, Canada.
Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Qi, C, Ma, Z D, Kikuchi, N, Pierre, C, Wang,
H & Raju, B 2005. Fundamental studies on
crashworthiness design with uncertainties in the
system. Proceedings, 2005 SAE World Congress
and Exhibition, 11–14 April 2005, Detroit, MI.
SAE Technical Paper.
Rabbat, B G & Russell, H G 1985. Friction coefficient
of steel on concrete or grout. Journal of Structural
Engineering (US), 111(3): 505–515.
Reddy, B V V & Gupta, A 2006. Tensile bond strength
of soil-cement block masonry couplets using
cement-soil mortars. Journal of Materials in Civil
Engineering, 18(1): 36–45.
SANS (South African National Standard) 2011. SANS
10400-L: 2011. The Application of the National
Building Regulations. Part L: Roofs. Pretoria: SABS
Standards Division.
SANS 2015. SANS 10400-K: 2015. The Application of
the National Building Regulations Part K: Walls.
Pretoria: SABS Standards Division.
SANS 2018. SANS 10160-3: 2018. Basis of Structural
Design and Actions for Buildings and Industrial
Structures Part 3: Wind Actions. Pretoria: SABS
Standards Division.
Santos, C, Alvarenge, R, Ribeiro, J, Castro, L, Silvia, R,
Santos, A et al 2017. Numerical and experimental
evaluation of masonry prisms by finite element
method. Revista IBRACON de Estruturas e
Materiais, 10(2): 477–508.
Sarhosis, V, Garrity, S W & Sheng, Y 2015. Influence of
brick-mortar interface on the mechanical behaviour
of low bond strength masonry brickwork lintels.
Engineering Structures, 88: 1–11.
Van der Pluijm, R 1992. Material properties of masonry
and its components under tension and shear.
Proceedings, 6th Canadian Masonry Symposium
15–17 June 1992, Saskatoon, Canada, pp 675–686.
Zhenhai, G 2014. Principles of Reinforced Concrete.
1st ed. Oxford: Elsevier.
Zunguzane, N, Smallwood, J & Emuze, F 2012.
Perceptions of the quality of low-income houses
in South Africa: Defects and their causes. Acta
Structilia, 19(1): 19–38.
A development cost TECHNICAL PAPER
comparison between a Journal of the South African
Institution of Civil Engineering
multi-storey mass timber ISSN 1021-2019
Vol 63 No 4, December 2021, Pages 35–44, Paper 1281

and reinforced concrete FANIE VAN DER WESTHUYZEN currently

building in South Africa

works as a structural engineer for Zutari in
Cape Town. He holds a Master’s in Civil
Engineering from Stellenbosch University.
His undergraduate thesis focused on the
behaviour of South African timber species in
fire. In his Master’s research he went on to
S Van der Westhuyzen, J Wium
focus on the economic viability of a
potential multi-storey mass timber building in South Africa. He plans to
specialise in timber engineering through research, and practical design
High-rise timber buildings have experienced a resurgence internationally during the past two and construction experience.

decades. This paper presents an investigation into the financial feasibility of a multi-storey Contact details:
mass timber building for South Africa through a development cost comparison. Two 8-storey Zutari (Cape Town Office)
1 Century City Drive
commercial buildings – a mass timber frame and a reinforced concrete frame – were first
Century City
designed by independent engineering consultants. A focus group workshop, conducted with Cape Town 7446
industry professionals, assisted with the development of construction schedules. Subsequently, South Africa
a financial model was developed to determine the overall development cost and financial T: +27 21 526 9400
E: fanie.vdw@zutari.com
feasibility of each option. Finally, a sensitivity analysis was conducted to investigate the effect of
certain variables on the overall profitability of the mass timber frame development.
PROF JAN WIUM (FSAICE) holds a PhD from
The focus group workshop identified that the construction of the reinforced concrete frame the Swiss Federal Institute of Technology in
building and mass timber frame building will take 42 weeks and 21 weeks, respectively. The total Lausanne. He worked in industry as a
capital investment required for the mass timber frame development was found to be 10% more structural engineering consultant and
project manager for 20 years. In 2003 he
than that of the reinforced concrete frame development (R115 691 000 versus R 105 118 000).
joined Stellenbosch University as specialist
A five-year internal rate of return (IRR) of 20.9% and 25.7% was calculated for the mass in structural concrete and structural
timber and reinforced concrete frame developments, respectively. A significant finding of the dynamics. Since 2010 he is responsible for
sensitivity analysis was that the mass timber frame building proved to generate a higher five- the focus area in Construction Engineering and Management at the
Department of Civil Engineering. His research interests are modular
year IRR than that of the reinforced concrete frame once the mass timber building achieved a construction, collaboration for improved constructability, design
rental premium of 7.8% or more. The sensitivity analysis further showed that the importation management, and risk management.
of the mass timber elements remains an expensive option, with a 16.4% five-year IRR for the Contact details:
imported mass timber frame (at a R17:€1 exchange rate). Department of Civil Engineering
Stellenbosch University
Private Bag X1
Matieland 7602
INTRODUCTION performer across most environmental South Africa
Rising awareness and interest in environ- impact factors when compared to build- T: +27 21 808 4348
mental and global warming challenges ing materials such as steel and concrete, E: janw@sun.ac.za

have grown significantly in recent years, with particularly good performance

leading to a call for sustainable hous-
ing technology and methods within the
construction industry (UN Habitat 2016).
This has sparked renewed worldwide
interest in the use of timber for construc-
tion (MGA 2012). Timber remains unique
as it is one of few construction materials
with a negative carbon footprint before
processing (MGA 2012). During its
‘manufacturing period’ (growth) it takes
in atmospheric CO2 and releases O2 dur-
ing photosynthesis. This contrasts with
steel and cement which were responsible
for approximately 5% (2003) and up to
7% (2002) of global greenhouse gas emis-
sions (Anderson et al 2015). Taking this
into consideration, research has shown
that timber is renewable, and is the best
in terms of greenhouse gas emissions
(Crafford & Wessels 2020; Petersen &
Solberg 2005; Sathre & O’Connor 2010;
Upton et al 2008; Wang et al 2014;
Werner & Richter 2007).
Forte Living is a 10-storey mass timber
building in Melbourne constructed with
759 cross-laminated timber (CLT) panels.
After completion it was estimated that
the building has a 22% lower carbon
footprint compared to similar reinforced
concrete structures (Arup 2019). In light
of such case studies, timber construction
is p
­ romoted as a more environmentally
friendly and sustainable building mate-
rial when sustainable forest manage-
ment is practised (as opposed to steel Keywords: mass timber, reinforced concrete, construction cost,
and concrete). construction schedule, development cost comparison

Van der Westhuyzen S, Wium J. A development cost comparison between a multi-storey mass timber and reinforced concrete building
in South Africa. J. S. Afr. Inst. Civ. Eng. 2021:63(4), Art. #1281, 10 pages. http://dx.doi.org/10.17159/2309-8775/2021/v63n4a4 35
 As a result, high-rise timber build-
ings have experienced a resurgence
30.00
internationally during the past decade.
6.00
Approximately 20 mass timber buildings,
which are six storeys and higher, have A
Office Office

3.00
been completed since 2010, while in 2017 17 m2 17 m2

6.00
more than 13 multi-storey mass timber
Office Office
buildings (seven storeys and higher) 17 m2 17 m2
were under way (Forestry Innovation B
Office Office

3.00
Investment 2017; Salvadori 2017). The 17 m2 Open plan 17 m2
office
growth in the market share of multi- Office 321 m2 Office
storey mass timber building has sparked 17 m2 17 m2

24.00
C
interest among South African property Office Office
developers, and architecture, engineering 17 m2 17 m2
and construction (AEC) professionals. Office Office
Questions have arisen – upon comple- 17 m2 17 m2
D
tion of previous research on mass timber Elevator
Shaft
elements by the authors – regarding the Kitchen & lounge Bathroom
72 m2 72 m2

HVAC
Stairs
potential development cost of multi-storey
mass timber buildings and how this would E
compare to a typical building system used
1 2 3 4 5 6
in South Africa. The new-found interest in
timber construction among South African
industry professionals merits research in Figure 1 F loor plan for building
mass timber construction (MTC). It would
be of interest to see how the application
of mass timber compares to that of rein- Concrete elevator and staircase core
forced concrete, since reinforced concrete
is the most dominant construction mate-
rial used in South Africa (Drennan 2017).
Previous studies internationally have
made comparisons between timber and
conventional construction options, but
no such comparison has been made for
the South African industry (MGA 2012;
Shear wall
Smith et al 2018; Timber Development
Association NSW 2015).
This paper presents a development cost
comparison between a multi-storey mass
timber and a reinforced concrete building
in South Africa. The aim is to assess the
feasibility of using mass timber in multi-
storey structures in South Africa.

DESIGNS
Flat slab concrete structure

Conceptual designs and design loads

The building considered for the study is a
fictitious 8-storey upmarket commercial
building situated in the Sandton central
business district (CBD). The building
footprint is 2 430 m2 with a gross floor
area (GFA) of 5 472 m2 for the eight storeys
(excluding lift and stair shaft area). The
floor-to-floor height is 3.5 m throughout
the building, resulting in a total height of
28 m. The building has a 66 m2 column
grid from floors 1 to 8 as seen in Figure 1.
The external face of the building is covered
Figure 2 3 D Revit (Autodesk 2019) model of concrete frame
by an aluminium glass façade, while
lightweight in-fill drywalls make up the
office partitions. The timber and concrete
buildings share the same basic layout for
comparison purposes.
Design loads and limit states
The buildings in this study were conceptu-
ally designed by independent structural
engineering firms. SANS 10160-2 (SANS
2011) was applied to both structures. The
reinforced concrete building was designed
using SANS 10100-1 (SANS 2000), whereas
the mass timber frame building was
designed using Eurocode 5 (EN 2004). This
is largely ascribed to the European consult-
ing firm’s familiarity with Eurocode 5
(EN 2004) and a lack of information
about CLT in SANS 10163-1 (SANS 2003).
Both designs remain conceptual and are

36 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Table 1 Summary of concrete structure is the system of choice for this particular
Element Location Depth (m) Width (m) Length (m) Quantity (no)
design. Research by Drennan (2017) has
shown that a post-tensioned slab is a
Inner 0.8 5.1 5.1 10 cheaper alternative for concrete slabs (espe-
Outer 0.7 3.2 4.1 10 cially in the case of long-spanning slabs).
However, normal reinforced concrete
Footings (30 MPa) Corner 0.6 3.0 3.0 4
remains the most common building tech-
Shear wall 0.8 4.0 8.0 1 nique in South Africa. As such, and due
Core 0.8 10.0 10.0 1 to the 66 m2 column grid, it was decided
to simplify the design by selecting normal
1–2 0.5 0.5 3.5 20
reinforced concrete. Table 1 provides a
Inner columns 3–5 0.4 0.4 3.5 20 summary of the dimensions of the main
6–8 0.3 0.3 3.5 20 concrete elements within the building.
Upon a design review it was realised that
Outer columns G–8 0.5 0.5 3.5 16
a raft foundation may have been more
Surface bed 0.12 24 30 1 appropriate, given the large dimensions of
Slab (35 MPa) the footings (resulting from the low safe
Flat slab 1–8 0.3 24 30 7
bearing pressure chosen).
Ring beam 2–Roof 0.5 0.3 108 8

Core 0.25 33.8 3.5 8 Mass timber frame building

Walls
Shear 0.25
Figure 3 3 D Revit (Autodesk 2019) model of mass timber frame
considered to be conservative (have not
been optimised).
Reinforced concrete frame building
The structural analysis and design of
the reinforced concrete building was
performed by an independent consulting
engineering firm in South Africa. Figure 2
5.5 3 8
is an image of the 3D Revit (Autodesk 2019)
model of the reinforced concrete frame
building. A steel roof structure is assumed
for both the concrete and mass timber
frame buildings.
Lateral stability is provided by the
reinforced concrete core and shear walls. A
flat slab system with an external ring beam
For this structural option a realistic tim-
ber building for South African conditions
was envisaged. It takes into account the
timber products that are currently manu-
factured in South Africa, as well as the
skills and expertise of the carpenters/con-
tractors as raised during the focus group
workshop (discussed further down in the
section titled “Focus group participants”
on page 38). As such, the design can be
considered conservative (as in the case of
the concrete structure). The structural
analysis and design of the building were
performed by an independent consulting
engineering firm (located in the Czech
Republic) with previous experience of
mass timber frame structures. Figure 3
is an image of the 3D Revit (Autodesk
2019) model of the proposed mass
timber structure.
A column-beam system with a CLT
core and lateral glulam bracing is the
system of choice for the timber build-
ing. One-way spanning, 220 mm deep
CLT floor panels span between large
glulam beams as seen in Figure 4. Each
CLT floor panel consists of seven layers
(30‑30‑30-40-30-30-30) and is 6 m long
and 3 m wide, with a total estimated mass
of 1.86 tons per panel. The CLT panels are
single span, simply supported on glulam
beams (see Figure 4).
The structure of the proposed floor
system is illustrated in Figure 5. A 50 mm
screed layer protects the CLT from moisture
damage and assists with vibration control.
Additional sound insulation, separation
membranes and sealing tapes are also
recommended to ensure the floor system
satisfies the necessary design requirements.

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 37
The floor system was designed using Stora
Enso’s Calculatis software (Stora Enso 2019). Column/beam glulam system
Initially, all the glulam beams and 6.0 m

columns were designed as GL24h timber

A
(imported timber). However, in the descrip-
tion below, South African timber species
such as SA pine and saligna (eucalyptus)

12.0 m
are considered. Varying dimensions with B

different timber species were therefore

scrutinised by the authors in an attempt to
Glulam
optimise the cost of the structure. These bracing system C
options are discussed further down in 3 m × 6 m CLT
the section titled “Development cost” (see Floor panel

page 40). The design completed by the D

Czech consultant only considered the use
of GL24h timber. Alternative conceptual
designs using South African timber spe- E
cies were verified by the authors through 1 2 3 4 5 6
calculations and structural analysis soft- CLT core
ware. The design completed by the Czech
consultant for GL24h timber is presented Figure 4 P
 lan view of the mass timber building
in Table 2.
The mass of the reinforced concrete
frame building is approximately 3.6 Structure
times more than the mass timber frame Screed
option. As a result, the use of mass timber Separating layer (optional)
as opposed to conventional reinforced Impact sound CLT wall board
concrete for the structural frame resulted Insulation
in a 68% reduction in foundation size
Fill (gravel)
from the concrete building for this
Trickle protection
particular comparison. (optional)
CLT floor board
Connection design
Beam–column and column–column
connections in the timber frame were Screed edge strap
modelled as fixed connections in the
structural model as per recommendation
from the Czech consultant. The beam
Joint-sealing
column connection experienced a tape
maximum shear force of 136 kN and
maximum bending moment of 150 kNm.
Figure 6 shows a 3D render of a proposed
connection to resist such forces. A
steel-plate-and-dowel system was
recommended by the Czech consultant.
CLT floor/ceiling board
An international supplier of mass
timber connections provided additional
assistance with the design and costing of Figure 5 P
 roposed CLT floor system (Stora Enso 2015)
the steel connections, membranes, taping
and sound-proofing. The impact of less pine and S10 SA pine. As such, different CONSTRUCTION SCHEDULE
expensive hinged connections on the cost cross-sectional dimensions are required
and beam element size was not considered to those presented in Table 2. Eurocode 5 Focus group participants
and is a subject for further refinement. (EN 2004) was used throughout the entire With the purpose of establishing a
fire design procedure. The reduced cross- construction schedule, five industry
Timber rational fire design section method was applied for both the professionals participated in a focus group
A rational fire design was performed by glulam beams and columns. The mass workshop together with the authors. Each
the Czech consultant, where the timber timber structure satisfies the necessary one of these participants practises in a
was initially assumed to be GL24h. ultimate limit state (ULS), serviceabil- different area of expertise and hence made
Subsequently, a rational fire design was ity limit state (SLS), and fire limit state valuable contributions. Table 3 provides
performed by the authors using S7 SA requirements (FLS). information regarding the profession,

38 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Table 2 Summary of mass timber frame for GL24h QQ The construction schedule for both

Element Location Depth (m) Width (m) Length (m) Quantity (no)
buildings would incorporate finishes.
QQ A five-day work week, from Monday to
Inner 0.6 3.2 3.2 10 Friday, would apply.
Outer 0.5 2.4 2.8 10 QQ Concrete for the slabs and beams was
Footings (30 MPa) to be pumped, whereas the verticals
Corner 0.4 2.2 2.2 4
(columns, shear walls and core) would
Core 0.5 8 8 1 be cast with buckets.
Columns (GL24h) 1–8 0.4 0.4 3.5 240 QQ A power float finish would be applied to
the reinforced concrete slab.
Internal beams (GL24h) 1–8 0.52 0.24 6 120
QQ A prefabricated, deliver-and-build con-
Ring beam (GL24h) 1–8 0.32 0.24 6 144 struction technique would be applied.
As such, excessive amounts of timber
Bracing beams (GL24h) 1–8 0.26 0.24 18.32 24
elements would not be stored on-site.
CLT floors (220L75-2) 1–8 0.22 3 12 126 The timber elements would be cut and
CLT floors (220L75-2) 1–8 0.22 3 6 14 shaped off-site.
QQ It was assumed that the South African
CLT core (180 C35) 1–8 28 0.18 32.08 1
multi-storey mass timber building
industry is an established industry.
Artisans would thus be familiar with
the construction technique, and manu-
facturers would be capable of supplying
material regularly and on time.
QQ It was assumed that the construction
schedule remains unaffected whether
the timber components are imported or
locally manufactured.
QQ Internal finishes were all installed on-
site, despite the fact that services could
be pre-installed in the factory for CLT
elements.

Figure 6 P
 otential beam–column connection used in timber frame Reinforced concrete frame schedule
It was assumed that construction would
qualification, experience and position of discussed and clarified. Following this, start on 6 January 2020. Figures 7 and 8
each participant. the construction schedule and Gantt show timelines of the main tasks scheduled
A week before the workshop charts of each participant were presented for construction of the concrete and timber
commenced each focus group participant and discussed. Finally, a combined structures, respectively. Concrete work fin-
was provided with the 3D conceptual construction schedule was developed for ished on 25 September 2020, while internal
models of the buildings, as well as the mass timber and reinforced concrete finishes for all floors were completed on
additional information regarding multi- frame buildings. 20 October 2020 – approximately four
storey mass timber construction. The weeks later. Overall, the total estimated
participants were required to complete Assumptions time required to finish the building frame
individual construction schedules for The following assumptions were made dur- and internal finishes was 207 working days
the mass timber and reinforced concrete ing the workshop: (approximately 42 weeks or 10 months).
frame building, respectively. On the day QQ A single fixed crane would be required
of the workshop a number of assumptions for both the concrete and timber Mass timber frame schedule
which may have been unclear were first building. The time to construct the foundations
and building substructure was assumed
Table 3 Focus group participants to be the same for both buildings for the
Profession Qualification Experience Company position purpose of this analysis, as can be seen in
Figures 7 and 8. The mass timber frame
Project Manager BEng (Civil) 38 years Managing Director
was constructed significantly faster than
Project Manager BSc (Building Management) 30 years Managing Director the reinforced concrete frame. A five-day
Contractor BSc (Building) 21 years Contracts Director per floor turnaround time was deemed
realistic for the timber building as opposed
Carpenter
Master’s degree in Carpentry 23 years Owner to the 15 days allocated for concrete. A
Glulam Manufacturer
total of 40 working days was required
Carpenter Bachelor of Architecture Studies to construct the entire timber frame. A
12 years Director
CLT Manufacturer (B.A.S)
20-day fit-out time per floor was also

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 39
 6 Jan – 17 Feb
Concrete bases 31 days

13 Feb – 5 Mar
Stub
16 days columns
13 Feb – 14 Apr
44 days Concrete core

44 days Shear wall

20 Jan – 10 Feb 6 Mar – 25 Sep

16 days Surface 146 days Reinforced concrete frame
bed
14 Apr – 20 Oct
136 days Fit-out phase

January February March April May June July August September October

Figure 7 T imeline of reinforced concrete structure

6 Jan – 17 Feb
Concrete bases 31 days

13 Feb – 5 Mar
Stub
16 days columns
20 Jan – 10 Feb 6 Mar – 30 Apr
16 days Surface 40 days Timber frame
bed
13 Mar – 28 May
55 days Fit-out phase
20 Oct

January February March April May June July August September October

Figure 8 T imeline of mass timber structure

allocated for both buildings. The focus
group saw this as a conservative estimate
as CLT panels allow for the pre-installation
of services off-site. The entire timber
structure, including internal finishes, was
estimated to be constructed in 104 working
days (approximately 21 weeks or 5 months).
This was five months earlier than the rein-
forced concrete frame, translating to a 50%
reduction in construction schedule. The
focus group noted that the true benefit of
mass timber construction comes from early
access for follow-on trades (fit-out). This
is highlighted in Figures 7 and 8, where a
total fit-out time of 136 days was required
for the reinforced concrete frame building,
whereas only 55 days were required for the
mass timber frame building.
DEVELOPMENT COST
Several stakeholders assisted in the
costing of the two buildings. Two profes-
sional quantity surveyors were involved
throughout the costing process. Various
manufacturers were contacted to assist in
the costing of the mass timber frame build-
ing due to a lack of established multi-storey
mass timber projects in South Africa.
These manufacturers included CLT manu-
facturers, glulam manufacturers, steel
connection suppliers, custom steel part
manufacturers, and international suppliers.
Procurement of timber
Different options were investigated for the
procurement of mass timber elements, as
shown in Table 4. The first three columns
with timber alternatives consider manu-
facturing the glulam elements locally in
South Africa using different timber species
and grades. The final column presents the
estimated cost of importing the timber ele-
ments from Europe, which is dependent on
the Rand-Euro exchange rate. The dimen-
sions of the timber elements vary according
to the grade and species of timber specified
in the design. Calculations were performed
to determine equivalent cross-sectional
dimensions based on similar ultimate
bending resistances (approximately
161 kNm) for the different grades and spe-
cies of the beams. A similar comparison
was made for the glulam columns based on
similar ultimate compressive resistances
(2 150 kN).
An aspect which demands considera-
tion is the availability of a specific timber
grade for a given timber species. S5 and S7
SA pine is generally available from South
African sawmills, whereas S10 SA pine is
difficult to obtain. Manufacturers therefore
resort to using saligna (eucalyptus) – a
more expensive hardwood species – to
manufacture S10 glulam beams. Table 4
presents the cost per running metre of
beam/column for the different options.
Discussions with manufacturers indicated
that it costs approximately R12 000/m3
of S7 SA pine to manufacture glulam or
CLT (Holzbau Carpentry Hess CC 2020;
XLAM South Africa (Pty) Ltd 2020). The
lack of S10 SA pine availability presents a
challenge in determining a realistic cost
per cubic metre of manufacturing glulam.
A 20% premium has been added to the S7
price for comparison purposes based on
discussions with local suppliers (Holzbau
Carpentry Hess CC 2020; XLAM South
Africa (Pty) Ltd 2020). S10 SA pine is,
however, not considered in the final Bill
of Quantities due to a lack of availability
and uncertainty surrounding the cost per
cubic metre.
Given the above information and
the assumptions that were made, S7 SA
pine and S10 SA pine are the two most

40 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Table 4 Cost per metre comparison for timber beams and columns particular multi-storey mass timber build-
ing in order to avoid potential delays. The
Description S7 SA pine S10 SA pine S10 saligna GL24h spruce
production of such high volumes of mass
Beams timber may prove to be a significant chal-
Dimension (mm) 630×240 520×240 520×240 520×240 lenge for local manufacturers due to lim-
ited available production lines. As a result,
Cost/m3 R12 000 R14 400a R17 000 R15 863b
importation of timber becomes a more
Cost per m R1 815 R1 797 R2 122 R1 980 excl treatment viable option, despite the additional associ-
ated cost. Investment into the upgrading/
Columns
upscaling of machinery in the sector will
Dimensions (mm) 394×394 368×368 368×368 380×380 inevitably alleviate the challenge regarding
Cost/m3 R12 000 R14 400a R17 000 R15 863b manufacturing. A fictitious situation was
thus assumed where large mass timber
Cost per m R1 863 R1 950 R2 302 R2 291 excl treatment
products can be manufactured within SA.
Transportation cost of beams included in price estimate Table 5 summarises the results of the
a Assumed a 20% premium on the cost of S7 SA pine.
b R17:€1 Euro exchange rate; includes customs and import taxes, transportation and commission
investigation for different CLT procure-
ment options. Upon investigation it was
found that the mechanical performance
Table 5 CLT comparison of S7 SA pine CLT requires testing to
Description S7 SA pine C24 spruce
determine how its mechanical properties
compare to those of C24 spruce CLT. It
Panel dimension 220 × 3 000 × 6 000 mm 220 × 3 000 × 6 000 mm
was assumed that the mechanical proper-
Cost/m2 R2 640 R2 761b ties were approximately equivalent in order
to perform the cost analysis. The price
Cost per panel R47 520 R49 700 excl treatment
of 220 mm thick untreated C24 spruce is
Transportation of CLT included in price estimate approximately €110/m2 depending on the
b R17:€1 Euro exchange rate; inclusive of customs and import taxes, transportation and commission
European supplier. Guided by industry
professionals from a local timber supplier,
Table 6 Total construction cost excluding P&G costs the total price (including shipping) of the
CLT from untreated spruce equated to
Cost item Timber frame (R) % Concrete frame (R) %
R2 761/m2, at an exchange rate of R17:€1.
Substructure and foundation 508 714 1 1 578 016 3 The price for producing CLT from S7 SA
pine equates to approximately R12 000/‌m3
Structural frame 22 776 018 33 10 262 224 18
which is equivalent to R2 640/m 2 for a
Non-structural components 45 156 768 66 45 156 768 79 220 mm thick panel. From Table 5 it is
evident that the S7 SA pine CLT is a more
Total 68 441 500 100 56 997 008 100
cost-effective alternative to importing CLT
from Europe, given the above assumptions.
cost-effective solutions for the glulam beams Africa, i.e. SANS 10005 (SANS 2020). The
and columns in the mass timber frame proposed mass timber structure is assumed Total construction cost
building. S10 SA pine beams could poten- to be located within the Sandton CBD The total construction cost is summarised
tially work out to be the most cost-effective which is located outside the coastal zone. in Table 6. The R11 445 000 difference in
solution given the R14 400/m3 assumption. As such, it does not require additional total construction cost translates into a
S10 saligna and GL24h spruce remain the treatment. However, timber will require 20% increase in construction cost from the
two most expensive options for this analysis. treatment if the building were located reinforced concrete frame building.
The cost of glulam imported from within one of South Africa’s coastal cities, Table 7 contains the total construction
Europe (also known as BSH) was approxi- with an associated increase in the overall cost including the P&G cost, contractor
mately €630/m3 for untreated spruce. cost of the mass timber elements. contingencies and contract escalations
Guided by industry professionals from Interviews with manufacturing profes- as per standard industry practice. Total
a local timber supplier, the total price sionals indicated that current manufactur- costs of R75 638 000 and R65 311 000
(including shipping) for the untreated ing limitations within South Africa prevent were calculated for the mass timber and
spruce equated to R15 863/m3, at an the large-scale production of large cross- reinforced concrete frame buildings,
exchange rate of R17:€1. Analysis of the sectional beams/columns typically required respectively. The mass timber frame build-
Rand-Euro exchange rate showed that any in multi-storey mass timber structures. ing is R10 327 000 more expensive than
rate below R12.6:€1 (which is unlikely) The current conceptual design contains the concrete frame building with regard to
could make importation of untreated approximately 1 600 m3 of mass timber total construction cost.
spruce the most cost-effective solution. which will need to be manufactured over a
Timber elements within buildings 70-day construction period. This translates Total capital investment
found in coastal zones require treatment as into approximately 23 m3 of mass timber A well-established property development
specified by building regulations in South to be manufactured on a daily basis for this firm assisted in developing a financial model

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 41
to determine the overall development cost Table 7 Total construction cost including P&G cost
and financial feasibility of the ventures. Cost item Mass timber frame (R) Concrete frame (R)
Capital cost items such as land costs, con-
struction costs, professional fees, marketing Total construction cost 68 441 500 56 997 008

costs and general costs (which include inter- Preliminary and general cost 4 619 801 5 699 701
est during construction) were accounted
Contingencies (2.5%) 1 826 533 1 567 418
for in the model. Overall, the total capital
investment required for the mass timber Pre-contract escalation (1.5%) 280 829 240 990
frame building is R115 691 000. This is Contract escalation (1.5%) 469 804 806 314
R10 573 000 more than the R105 118 000
Total 75 638 467 65 311 431
total capital investment required for the
reinforced concrete frame building, which
translates into a 10% increase. Table 8 Monthly return on investment
Income description Rate Area / number of bays Monthly income
Earlier return on investment
The current monthly rental fee for green Gross rentable area R150/m2 5 427 R820 800

certified office spaces in the Sandton CBD On-grade parking R500/bay 217 R108 500
is approximately R150 per m2 whereas in
Monthly net rental R929 300
the Cape Town CBD the rental is closer
to R165 per m2 (Abland 2020). On-grade
parking bays hold a further opportunity
to earn an income from the development. 100 Loan
Table 8 shows that the total monthly income 90 Loan
80
for the development is R929 300 based on a
Millions (Rands)

100% occupation Sandton CBD office rental
fee and on-grade parking. This results in
a R4 646 500 income over the five months
while the reinforced concrete frame build-
ing is still under construction. The effect of
an earlier return on investment is presented
below under the heading “Internal rate on
return” and is quantified by calculating the
internal rate of return of each development.
Discussions by the authors revealed that a
100% occupation for both buildings may
be optimistic. A sensitivity analysis on the
occupation rate should therefore be con-
ducted for future studies.
Interest during construction
Developments are typically funded through
equity provided by the developer or private
investors and a development loan obtained
from an accredited credit provider (Abland
2020). A 70:30 debt to equity ratio was
assumed for this particular development.
Money is drawn from the development loan
once equity is exhausted (Abland 2020). An
interest rate of 8% per annum was assumed
for this particular project. Proportions for
monthly expenditure were applied based
on the experience of the developer and
allowed for the development of an ‘S-curve’
for both buildings (Figure 9). The rate at
which equity was utilised and money drawn
from the development loan was therefore
determined. Equity of R14 370 000 and
R3 796 000 was calculated after subtracting
R19 152 000 equity in respect of land value
(based on the 70:30 debt to equity ratio) for
70
60
50
40
30
20 Equity
10 Equity
0
1 2
Figure 9 C
 umulative cash outflow during construction
the timber frame and reinforced concrete
frame buildings, respectively.
Figure 9 shows the expected cash out-
flow (excluding interest expense and land
cost) for both the mass timber and rein-
forced concrete buildings. From Figure 9 it
can be seen that the R14 370 000 equity is
completely exhausted within two months
of construction for the timber frame
building. Similarly, within two months
the R3 796 000 equity for the concrete
frame building is also exhausted. Interest
is effectively charged for four months for
the timber frame building and nine months
for the concrete frame building. The total
interest payable by the developer over
the construction period is R1 486 200 for
the timber frame building. Alternatively,
R 2 706 000 is payable by the developer
for interest incurred for the construc-
tion of the concrete frame building. The
five-month shorter construction schedule
results in savings of R1 220 000 in interest
for the timber frame building.
3 4 5 6 7 8 9 10
Months
Concrete S-curve Timber S-curve
Internal rate of return
The feasibility of the two developments can
be assessed through the calculation and
evaluation of the internal rate of return (IRR)
and minimum acceptable rate of return
(MARR) of each development. If the IRR is
greater than the MARR, the development is
financially justifiable (Blank & Tarquin 2014).
When evaluating two different developments,
the development with the greatest IRR is
the more profitable development from an
investor’s perspective. The MARR, or hurdle
rate, can be assumed to be 15% for the com-
mercial developments in this study (Abland
2020). The five-year IRR of the developments
were calculated through the development
of amortisation schedules in the financial
model. The mass timber building achieved a
five-year IRR of 20.9%, while the reinforced
concrete building achieved a five-year IRR of
25.7%, 4.8% higher than the IRR of the mass
timber frame building.
From the higher IRR for the reinforced
concrete frame development, it is clear

42 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering

40
35
30
IRR and yield (%)
25
20.9
20
15
9.6
10
5 5% in the rental rate increases the five-year
0
0 1 2 3
Figure 10 R
 ental rate analysis for the mass timber frame building
that the concrete alternative remains more
profitable than the timber alternative,
despite the five-month shorter construction
schedule. Notably, the five-year IRR of both
developments is above the 15% MARR,
indicating that they are both financially
justified. This is a positive result with 100%
assumed occupation for both developments,
particularly the multi-storey mass timber
building, in the context of South Africa. It
indicates that a multi-storey mass timber
building can be financially viable in South
Africa if a number of factors are addressed
throughout the entire value chain of the
mass timber products. However, it is evident
from this particular comparison – given the
assumptions that were made – that mass
timber frame commercial buildings will
struggle to achieve higher internal rate of
returns than conventional reinforced con-
crete frame commercial buildings.
SENSITIVITY ANALYSIS
The following section explores the effect of
certain variables on the overall construction
cost comparison between the two buildings
through a sensitivity analysis. The investiga-
tion is limited to the variables that have the
greatest effect on the overall construction
cost of the mass timber frame building.
Rental rate
It is of interest to explore the effect of a
potential rental premium on the mass timber
frame building, especially when considering
the potential desirability of tenants to work
in such a mass timber frame building. The
effect of adjusting the rental rate on the five-
year IRR and initial yield is summarised in
Figure 10. The rental rate for the reinforced
concrete frame building remained constant
at R150/m2 for the entire analysis.
Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021
development. This translates into an increase
of approximately 0.7% in the five-year IRR.
Similarly, a 5% increase in the cost of SA
30.0 31.1
27.3 pine results in a 0.7% decrease in the five-
25.5
23.9 year IRR. From this analysis it is clear that
fluctuation in the cost of the SA pine has a
substantial effect on the five-year IRR of the
development, although it is not as sensitive as
10.1 10.3 10.5 10.8 10.9
fluctuation in the rental rate. An increase of
IRR between 3.0% and 3.8%, which is signifi-
4 5 6 7 8 9 10 11 12 13 14 15 16 cantly more than the 0.7% increase recorded
Percentage increase in rental rate (%) for a decrease of 5% in the cost of S7 SA pine.
5 Year IRR (%) Initial Yield (%) Results showed that the price of S7 SA pine
needs to be reduced by at least 33% for the
mass timber frame development to earn a
higher five-year IRR than the reinforced con-
Analysis of the results shows the crete frame development. This is considered
significant effect of the rental rate on to be highly unlikely.
the five-year IRR of the development. A
potential 7.8% increase in rental rate would Importation of timber
result in a five-year IRR of 25.7% for the The section above dealing with
mass timber frame development. This is “Development cost” (commencing on
equivalent to the 25.7% IRR recorded for page 40) alludes to the fact that the importa-
the reinforced concrete frame building. It tion of structural timber will have to be
remains difficult to determine whether a considered for the immediate future to
7.8% rental premium is justifiable for the sustain a potential multi-storey mass timber
mass timber frame development. building market. As such, an analysis of the
potential structural cost and five-year IRR
Cost of SA pine of the development is required if all of the
The large difference in the structural frame mass timber is imported. Table 10 contains
cost between the two buildings can be partly the results of the analysis. The results of the
attributed to the high cost per cubic metre analysis are based on the cost of untreated
of S7 SA pine. Discussions with manufactur- spruce as calculated in the aforementioned
ers indicated that it costs approximately section on “Development cost”.
R12 000/m3 of S7 SA pine to manufacture At a R17:€1 exchange rate, the total
glulam or CLT (Holzbau Carpentry Hess CC structural cost increases from R23 285 000
2020; XLAM South Africa (Pty) Ltd 2020). for the original S7 SA pine option to
It remains difficult to determine how prices R30 255 000 for imported spruce. This
may differ in an established multi-storey amounts to a 30% increase in the total
mass timber market. To this end, fictional structural cost of the building. Additionally,
market prices of S7 SA pine were investi- it results in a 4.5% decrease in the five-year
gated. The results are summarised in Table 9. IRR from 20.9% to 16.4%. In the section
Table 9 shows that a 5% decrease in the above dealing with “Internal rate of return”
cost of SA pine results in a R933 606 (4%) (page 42), a 25.7% IRR is calculated for the
reduction in the overall structural cost of the reinforced concrete frame development.
Table 9 Analysis of S7 SA pine cost
Change SA pine cost Structural
Difference (R) IRR (%)
(SA pine) (R/m3) cost (R)
+15% 13 800 26 074 029 933 606 18.9%
+10% 13 200 25 140 423 933 606 19.5%
+5% 12 600 24 206 817 933 606 20.2%
0% 12 000 23 284 732 0 20.9%
–5% 11 400 22 339 606 –933 606 21.6%
–10% 10 800 21 406 000 –933 606 22.3%
–15% 10 200 20 472 394 –933 606 23.1%
43
Table 10 Analysis of importing timber
Spruce glulam Spruce CLT
Exchange rate
cost (R/m3)* cost (R/m2)*
R19:€1 17 602 3 968
R18:€1 16 733 3 816
R17:€1 15 863 3 664
R16:€1 14 994 3 512
R15:€1 14 124 3 360
*The cost does not include treatment of the spruce.
This is 9.3% higher than the 16.4% calcu-
lated for the imported mass timber frame.
The assumption regarding the locally
produced mass timber elements should be
taken into consideration. The manufactur-
ing capabilities of local suppliers, as dis-
cussed earlier, remains a concern. As such, a
fictional situation was assumed where South
African suppliers can manufacture the large
cross-sectional dimensions required for the
mass timber frame building in this research.
Without this assumption, the timber would
most likely need to be imported unless cur-
rent manufacturing capabilities improve.
CONCLUSION
The findings of this study show that there
is great potential in the multi-storey mass
timber building domain in South Africa.
This is justified by the 50% reduction in the
construction schedule for the mass timber
frame building, but also by the sensitivity
analysis where the mass timber frame
building achieved the same five-year IRR
than the reinforced concrete building with
a 7.8% rental premium. However, it has
become evident through interaction with
various industry stakeholders that great
strides are still required in the forestry sec-
tor, mass timber manufacturing sector, as
well as the AEC sector before multi-storey
commercial mass timber buildings may
prove to be more profitable than conven-
tional multi-storey reinforced concrete
buildings in South Africa.
Two of the most notable requirements
include improvement in the sourcing of
high-grade structural timber (S7, S10) and
investment into equipment that will enable
the large-scale production of large beams/col-
umns, typically required in multi-storey mass
timber structures. Development and invest-
ment of capital and resources are required
throughout the entire value chain of mass
timber production to establish a sustainable
multi-storey mass timber market.
44
Structural Five-year
cost (R) IRR (%)
32 520 480 15.2
31 387 958 15.8
30 255 437 16.4
29 122 915 17.0
27 990 393 17.7
The future success of mass timber con-
struction in South Africa is thus dependent
on the collaboration and teamwork of indus-
try stakeholders and research institutions
throughout the public and private domain.
REFERENCES
Abland 2020. Abland (Pty) Ltd 2020 online interview,
1 June, Johannesburg.
Arup 2019. Rethinking timber buildings. https://www.
arup.com/perspectives/publications/research/
section/rethinking-timber-buildings.
Anderson, J E, Wulfhorst, G & Lang, W 2015. Energy
analysis of the built environment: A review and
outlook. Renewable and Sustainable Energy Reviews,
44: 149–158. doi: 10.1016/j.rser.2014.12.027.
Autodesk 2019. Revit (Version 2019.2.1) [Computer
Program]. Autodesk, San Rafael, CA.
Blank, L & Tarquin, A 2014. Rate of Return analysis. In:
Basics of Engineering Economy. 2nd ed. New York:
McGraw Hill, 143.
Crafford, P L & Wessels, C B2020. South African log
resource availability and potential environmental
impact of timber construction. South African Journal
of Science, 116(7/8): 1–8. doi: 10.17159/sajs.2020/6419.
Drennan, M B 2017. Comparative construction costs
of typical low-rise office buildings in South Africa.
Master’s Thesis. University of Stellenbosch.
EN (European Standard) 2004. EN 1995-1-1: Eurocode
5: Design of Timber Structures. Brussels, European
Committee for Standardization (CEN).
Forestry Innovation Investment Ltd (FII) 2017.
Brock Commons Tallwood House, Naturally
Wood. Vancouver, Canada: FII. https://
www.naturallywood.com/wp-content/
uploads/2020/08/brock-commons-tallwood-
house_factsheet_naturallywood.pdf.
Holzbau Carpentry Hess CC 2020. Online interview,
5 May, Stellenbosch.
MGA/Michael Green Architecture 2012. The Case for
Tall Wood Buildings, 2nd ed. Vancouver, Canada,
p 240. http://cwc.ca/wp-content/uploads/
publications-Tall-Wood.pdf.
Petersen, A K & Solberg, B 2005. Environmental and
economic impacts of substitution between wood
products and alternative materials: A review of
Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
micro-level analyses from Norway and Sweden.
Forest Policy and Economics, 7(3): 249–259.
http://linkinghub.elsevier.com/retrieve/pii/
S1389934103000637.
Salvadori, V 2017. The development of a tallwood
building. Master’s Thesis. Vienna, Austria:
Politecnico Milano & Technische Universitat Wien
(TU Wien).
SANS 2000. SANS 10100-1: 2000: The Structural Use
of Concrete, Edition 2.2. Pretoria: SABS Standards
Division.
SANS 2003. SANS 10163-1:2003: Structural Use of
Timber. Part 1: Limit-states Design, Edition 2.3.
Pretoria: SABS Standards Division.
SANS (South African National Standard) 2011. SANS
10160-2 2011: Basis of Structural Design and Actions
for Buildings and Industrial Structures. Part 2: Self
Weight and Imposed Loads, Edition 1.1. Pretoria:
SABS Standards Division.
SANS 2020. SANS 10005: 2020: The Preservative
Treatment of Timber. Edition 9. Pretoria: SABS
Standards Division.
Sathre, R & O’Connor, J 2010. Meta-analysis of
greenhouse gas displacement factors of wood
product substitution. Environmental Science &
Policy, 13(2): 104–114. http://linkinghub.elsevier.
com/retrieve/pii/S1462901109001804.
Smith, R E, Griffin, G, Rice, T & Hagehoffer-Daniell,
B 2018. Mass timber: evaluating construction
performance. Architectural Engineering and Design
Management, 1(1-2): 27‒138. doi/abs/10.1080/17420
07.2016.1273089.
Stora Enso 2019. Calculatis (Version 2019). [Computer
Program]. Helsinki, Finland. https://www.storaenso.
com/en/products/wood-products/calculatis.
Timber Development Association NSW 2015.
Rethinking Apartment Building Construction –
Consider Timber. Melbourne: Forest and Wood
Products Australia.
Upton, B, Miner, R, Spinney, M & Heath, L S 2008. The
greenhouse gas and energy impacts of using wood
instead of alternatives in residential construction
in the United States. Biomass and Bioenergy,
32(1): 1–10. http://www.sciencedirect.com/
science/article/pii/S0961953407001109.
UN Habitat 2016. World Cities Report 2016 –
Urbanization and Development: Emerging Futures.
https://unhabitat.org/world-cities-report.
Wang, L, Toppinen, A & Juslin, H 2014. Use of wood in
green building: A study of expert perspectives from
the UK. Journal of Cleaner Production, 65: 350–361.
http://linkinghub.elsevier.com/retrieve/pii/
S0959652613005623.
Werner, F & Richter, K 2007. Wooden building
products in comparative LCA: A literature review.
International Journal of Life Cycle Assessment,
12(7): 470–479.
XLAM South Africa (Pty) Ltd 2020. Online interview,
5 May, Stellenbosch.
Load spreading in TECHNICAL PAPER
ultra‑thin high-strength Journal of the South African
Institution of Civil Engineering

steel-fibre-reinforced ISSN 1021-2019

Vol 63 No 4, December 2021, Pages 45–52, Paper 1168

concrete pavements PHIA SMIT (MSAICE) received her BEng in

Civil Engineering from the University of
Pretoria in 2012. This paper is based on the
research done for her PhD at the University
M S Smit, E P Kearsley of Pretoria. She is currently employed as a
lecturer at the same university, and her
fields of interest are concrete technology,
pavement engineering and soil-
Ultra-Thin Continuously Reinforced Concrete Pavement (UTCRCP) consists of a 50 mm thin High- structure interaction.
Strength Steel-Fibre-Reinforced Concrete (HS-SFRC) overlay placed on existing pavements as
Contact details:
rehabilitation or used as part of new pavements. Difficulties have been experienced with the Department of Civil Engineering
construction of UTCRCP. Additionally, the thin HS-SFRC has superior fatigue properties, but University of Pretoria
poor load-spreading ability compared to conventional concrete pavements due to its reduced Pretoria
0002
thickness. This results in high deflections when the pavement is loaded. The substructure
South Africa
of UTCRCP plays an important role in its performance. Cement-stabilised granular materials T: +27 12 420 2179
can be used to ensure gradual load spreading with depth, but its behaviour under flexible E: phia.smit@up.ac.za
concrete layers is not yet well understood. In this study the effect of increasing the HS-SFRC
layer thickness and the effect of incorporating cement-stabilised base layers were investigated PROF ELSABÉ KEARSLEY (Pr Eng, FSAICE)
graduated with a degree in Civil Engineering
using linear elastic finite element modelling. From stress levels calculated, it was found that C1
from the University of Pretoria in 1984, and
and C2 materials perform well underneath a 50 mm HS-SFRC layer subjected to standard axle she holds a PhD from the University of
loads of 80 kN, while C3 and C4 would deteriorate faster. Stabilised layers placed below a thin, Leeds. She worked as a Structural Design
flexible concrete layer may however crack, resulting in increased damage to supporting layers. It Engineer in both South Africa and the
United Kingdom before joining the staff in
is recommended that the response of UTCRCP should be investigated using advanced material
the Department of Civil Engineering at the
models for the cement-stabilised base and other substructure layers. University of Pretoria. For the last 26 years she has been involved with
cement and concrete materials research. In 2009 she served as President
of the South African Institution of Civil Engineering (SAICE).

Introduction be increased to 76 mm. Current specifica- Contact details:

Ultra-Thin Continuously Reinforced
Concrete Pavement (UTCRCP) consists of a
50 mm High-Strength Steel-Fibre-Reinforced
Concrete (HS-SFRC) overlay that is cast
continuously. The concrete overlay is typi-
cally additionally reinforced with 5.7 mm
diameter, 50 × 50 mm welded deformed
steel bar mesh. The continuous nature of
UTCRCP reduces potential problems with
joints by reducing the number of movement
joints to construction joints. The HS-SFRC
makes it possible to reduce the concrete slab
thickness (Briggs et al 2016). The steel mesh
and HS-SFRC control crack widths, prevent-
ing moisture ingress, and provide superior
post-crack carrying capacity which mitigates
edge punchout failures.
Although environmental loading has a
significant influence on the performance
of UTCRCP by causing curling, warping
and blow-up, the focus of this study is on
traffic loading (Bredenhann et al 2018; Rao
& Roesler 2005). Other problems that have
been experienced with the implementation
of UTCRCP include the constructabil-
ity of its 50 mm layer, for which it was
proposed that the layer thickness should
tions also require that cement-stabilised Department of Civil Engineering
University of Pretoria
bases are incorporated to ensure gradual Pretoria
load spreading. 0002
UTCRCP design methodology develop- South Africa
T: +27 12 420 2176
ment for traffic-associated failure has
E: elsabe.kearsley@up.ac.za
been focused on modelling the fracture
of HS-SFRC more accurately to predict
the system’s performance (Denneman
2011; Elsaigh 2007). Less attention has
been given to the consequences of the
reduced flexural stiffness of the thin
HS-SFRC layer of UTCRCP and how this
should inform the design approach of the
UTCRCP substructure.
In this paper the effect of increasing
the ultra-thin HS-SFRC layer thickness
from 50 mm to 76 mm on pavement
response to traffic loading, and the effect
of incorporating a cement-stabilised base
underneath the 50 mm HS-SFRC layer on
road pavement response to traffic loading
were investigated.
Three-dimensional linear elastic finite
element modelling was used to investigate
the pavement response, by considering
the overall pavement response as well as Keywords: ultra-thin concrete pavements, cement-stabilised bases,
the substructure response. The overall finite element modelling, load spreading

Smit MS, Kearsley EP. Load spreading in ultra-thin high-strength steel-fibre-reinforced concrete pavements.
J. S. Afr. Inst. Civ. Eng. 2021:63(4), Art. #1168, 8 pages. http://dx.doi.org/10.17159/2309-8775/2021/v63n4a5 45
pavement response was described by
selected critical parameters such as verti-
cal displacement, stress and strain in the
concrete layer and substructure.
Background
Pavement design
Pavements consist of a system of layers of
unbound and bound materials placed on
each other and supported by the subgrade
(Huang 1993). The purpose of pavements is
to allow wheeled vehicles to operate safely
(Brown & Selig 1991). Werkmeister et al
(2004) summarised that pavement design
is a process intended to find the most eco-
nomical combination of layer thickness and
material types for pavements, considering
the properties of the subgrade and the
environmental and traffic loading during
the service life of the road.
The uppermost layer of pavements
usually consists of a bound material such
as asphalt or concrete. The subbase and sub-
grade are considered as the foundation layers
of pavements. The system of layers placed
on the foundation is considered to be the
structural layer of pavements (Brown & Selig
1991). Broadly there are two types of pave-
ment ‒ flexible and rigid pavement. Typically,
concrete pavements are deemed to be rigid
pavement and asphalt pavements are deemed
to be flexible pavement. Conventional
concrete pavements use Normal-Strength
Concrete (NSC) which fails in a brittle man-
ner, has a cube compressive strength smaller
than 80 MPa and flexural strength smaller
than 8 MPa (Domone & Illston 2010; Neville
& Brooks 2010).
Rigid pavements are designed to limit
fatigue cracking by determining the load-
induced tensile stresses in the concrete
layer. This stress is used to calculate the
stress level, which is the ratio of the calcu-
lated tensile stress to the flexural strength.
The stress level is limited to ensure that
the desired number of load cycles can be
absorbed. The horizontal tensile stress in
the concrete layer can be reduced by alter-
ing the concrete layer properties or increas-
ing the strength and stiffness of the foun-
dation layer. If the concrete layer contains
relatively large volumes of steel reinforcing,
as is the case with UTCRCP, the flexural
strength (measured to be in the region of
11.7 MPa) is significantly higher than that
of NSC that is not reinforced. Due to the
relatively high steel-reinforcing content, the
load-carrying capacity of the concrete layer
46
will not reduce as a result of tensile crack-
ing, and any cracks that form when the
tensile strength of the concrete is exceeded
by the stress caused by wheel loads will
be prevented from opening up by the steel
reinforcing (Kearsley & Mostert 2010).
UTCRCP design approach
The thin, heavily reinforced HS-SFRC
overlay is placed on pavement systems
that require rehabilitation or forms part
of a new pavement system. The design
methodology of UTCRCP was extrapolated
from conventional concrete pavement
design methodology where concrete layer
thickness is typically greater than 150 mm
(SANRAL 2013). The applicability of
conventional concrete pavement design
methodology to the innovative pavement
system has been questioned in the past, the
main critique (and focus) being that the
fracture of HS-SFRC should be modelled
more accurately to predict the system’s
performance (Denneman 2011; Elsaigh
2007). This focus in terms of UTCRCP
and HS-SFRC agrees with the statement by
Ioannides (2006) that fracture mechanics
is one of the future directions of concrete
pavement research and design.
Less attention has been given to the
effect of the reduced flexural stiffness of
the thin HS-SFRC layer of UTCRCP in
comparison to the relatively thick NSC
layer of conventional concrete pavements
and how this should inform the UTCRCP
design approach. The principal traffic-
associated failure mechanism of conven-
tional concrete pavement is fatigue crack-
ing of the concrete layer. Rutting is not
considered as a traffic-associated failure
mechanism for conventional concrete pave-
ments, because the load spreading through
the thick concrete layer reduces the stress
that is experienced by the substructure
to such a low level that the accumulation
of permanent deformation is deemed to
be negligible. The response of the sub-
structure to traffic loading is considered
unimportant, as long as the variability of
the substructure is limited. During the
mechanistic analyses of conventional con-
crete pavement, the substructure is often
reduced to an array of springs and complex
load configurations are usually ignored,
with the load location on the concrete slab
being of greater importance.
Balanced pavements
The modular or modulus ratio was initially
introduced by Burmister (1945) in his
Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
paper The general theory of stresses and
displacements in layered systems I. It is
the ratio of the Young’s Modulus of each
layer divided by the Young’s Modulus of
the layer underneath it. It is a measure of
relative material stiffness and the relative
load spreading ability of adjacent layers in
pavements. The concept of limiting the
modular ratio has been extended to other
design methods. Pavements designed to
limit the modular ratio between adjacent
layers, spread load progressively with
depth, ensuring that layers with decreas-
ing strengths are not overloaded. These
pavements are referred to as balanced pave-
ments (SANRAL 2013) and this principle is
used to design flexible pavements.
The Pavement Number design method
incorporates a modular ratio limit to
ensure that balanced pavements are built.
This design method is used for Category A
and B roads, designed for traffic between
1 and 30 million equivalent standard axles
and roads with thin asphalt surfacing. The
Pavement Number design method uses an
effective long-term stiffness, which has a
maximum allowable limit, to determine
modular ratios. Typical modular ratio
limits range from 2 to 1.2 for unbound
granular materials (G1 to G10), 9 to 3 for
cement-bound granular materials (C1 to
C4) and 5 to 2 for materials that incorpo-
rate asphalt and bitumen (SANRAL 2013).
The modular ratio between conventional
concrete layers and granular material or
subgrade is typically two orders of magni-
tude higher than these values.
Cement stabilisation
Cement stabilisation is often used in pave-
ment layers because it is an economical way
of improving marginal granular materials
(De Beer 1990). The resulting materials fall
under the umbrella of “cement-modified
soils” (O’Flaherty 1967), referred to as
cemented natural gravel and classified
as C3 and C4 in South Africa (SANRAL
2013). Cement stabilisation is also used in
inverted pavements where a granular base
is placed on a cemented subbase. This is
done not only to create an anvil on which
the granular material can be densely com-
pacted, but also to prevent cemented layer
cracking from progressing through the
base to the asphalt and creating reflective
cracking on the surface. As with cemented
natural gravel, the material is expected to
crack and assume the characteristics of a
granular material (referred to as “equivalent
granular state”). This is a form of traffic
Table 1 S trength properties of cement-stabilised material (adapted from Department of Transport 1996)
UCS# ITS* Flexural strength
Material
(MPa) (kPa) (MPa)

C1: Cemented crushed stone or gravel 6.0. to 12.0 - 1.2 to 2.4$

C2: Cemented crushed stone or gravel 3.5 to 6.0 > 400 0.7 to 1.2$

C3: Cemented natural gravel 1.5 to 3.5 > 250 0.5 to 1.17$$

C4: Cemented natural gravel 0.75 to 1.5 > 200 0.25 to 0.5$$

# @ 100% Mod AASHTO (MPa) (Department of Transport 1979)

* @ 95–97% Mod AASHTO compaction, (SABITA 1993)
$ High-strength materials (C1 & C2): Flexural strength = 0.2*UCS
$$ Low-strength materials (C3 & C4): Flexural strength = 0.33*UCS

moulding, and a balanced pavement (where
the strength of the pavement layers reduces
with depth) is formed in the process.
Cemented crushed stone or gravel,
classified as C1 and C2, are not designed to
become granular-like materials with traf-
fic loading. They tend to crack in a more
discrete fashion and do not result in bal-
anced pavements (Jordaan 1984). Reflection
cracking is caused by the cement-stabilised
base layers that cause stresses in the overlay-
ing asphaltic layers (Visser 2017) and often
occurs when soil-cement materials are used.
It has been established that the sub-
structure of pavements with thin HS-SFRC
should be designed for poorer load spread-
ing and high deflections. The use of soil-
cement materials, C1 and C2, to ensure
gradual load spreading with depth could be
considered as a design solution. The pos-
sibility of an effect analogous to reflection
cracking should, however, be recognised.
If a C1 or C2 material in the base cracks, it
is unlikely that the stresses caused in the
HS-SFRC layer would cause the surface
layer to crack. However, it is possible that
the cracked base could cause stress con-
centrations in the underlying granular layer
and result in accelerated deterioration of
the subbase.
Material properties of cement-
stabilised materials
The Unconfined Compressive Strength
(UCS), Indirect Tensile Strength (ITS) and
flexural strength of the different strength
classes of cement-stabilised materials are
summarised in Table 1 (Department of
Transport 1996). The flexural strength
is determined as a fraction of the UCS.
Although the fatigue performance of
cement-stabilised materials is complex
(Lv et al 2021), material with cement
content from 3% to 4%, loaded to a stress
level of 0.5 generally reached between
1 and 10 million load cycles in terms of
Table 2 Material stiffness of cement-stabilsed material (adapted from Department of Transport 1986)
Material Pre-cracked
phase
C1: Cemented crushed stone or gravel 7 to 30
C2: Cemented crushed stone or gravel 4 to 14
C3: Cemented natural gravel 3 to 10
C4: Cemented natural gravel 2 to 7
flexural fatigue after 28 days of curing (Xie
et al 2018).
The material stiffness, in terms of
Young’s Modulus, of the different strength
classes of cement-stabilised materials is
summarised in Table 2. The material stiff-
ness of cracked cement-stabilised material
is significantly reduced. When the material
is cracked, the material stiffness is affected
by whether the overlaying layer is bound
or unbound. This is in part because con-
finement is influenced by the state of the
overlaying layer. The Young’s Modulus of
uncracked cement-stabilised granular mate-
rial ranges between a minimum of 2 GPa
for C4 and 30 GPa for C1. The Young’s
Modulus range for C1 materials is also wide
and falls between 7 GPa and 30 GPa.
The material stiffness can be esti-
mated from the UCS or flexural strength.
Relationships have been established for
cemented crushed stone or gravel and
for cemented natural gravel. Equations 1
and 2 show the relationship of the Young’s
Modulus (E) to flexural strength (σb )
and UCS (σc ) for C1 and C2 materials,
respectively. Equations 3 and 4 show the
relationship of the Young’s Modulus (E) to
flexural strength (σb ) and UCS (σc ) for C3
and C4 materials.
E = 8σb + 3 500 (1)
Young’s Modulus (GPa)
Post-cracked phase
Under cracked
Under bound
or untreated
materials
materials
1.5 1.2
1 0.75
0.75 0.5
0.5 0.3
E = 4.16σc0.88 + 3 484 (2)
E = 10σb + 1 000 (3)
E = 5.13σc0.88 + 1 098 (4)
It has been demonstrated through back-
calculation of deflection that the initial
material stiffness, in terms of Young’s
Modulus, of cement-stabilised layers is in
the order of 3 to 4 GPa (Department of
Transport 1986). The Pavement Number
design method uses a maximum effective
long-term stiffness of 1 500 MPa for C1
and C2, 550 MPa for C3 and 400 MPa for
C4 (SANRAL 2013).
Accelerated pavement testing of
UTCRCP with cement-stabilised bases
Internationally, ultra-thin SFRC pave-
ments fall under the category of ultra-thin
white-topping (Chen et al 2016; Pereira et
al 2006). Accelerated Pavement Testing
(APT), using a Heavy Vehicle Simulator
(HVS), has been used to investigate the
response of ultra-thin HS-SFRC slabs on
weak, medium and strong substructures,
as well as slabs with partial support
(Kannemeyer et al 2007). The weak sub-
structure consisted of ripped and recom-
pacted in-situ material. The medium and
strong substructures consisted respectively

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 47
of 150 mm and 300 mm thick, 4% cement-
stabilised in-situ material. The par-
tially supported pavement section had an
800 mm wide transverse cavity. The pave-
ments were subjected to a range of loads
that included 80 kN dual-wheel loading and
140 kN aircraft-wheel loading. The surface
displacements were measured using Joint
Deflection Measurement Devices (JDMDs).
Kannemeyer et al (2007) found that all the
pavement models performed well in dry
conditions and cycles of wetting had to be
introduced to ensure pavement failure.
The permanent deformation accu-
mulated steadily in the pavement with
the weak substructure. It was reported
that cracks formed parallel to the loading
direction, approximately 300 mm from the
wheel path, and loss of support was identi-
fied between the concrete and the recom-
pacted in-situ material. The deflection and
permanent deformation in the medium and
strong substructure were similar, with the
strong substructure performing margin-
ally better. The formation of longitudinal
cracks adjacent to the loaded area indicates
that permanent damage of the supporting
layers resulted in the increased surface
deflections measured.
Experimental setup
Three-dimensional finite element (FE)
modelling was used to investigate the effect
of increasing the ultra-thin HS-SFRC layer
from 50 mm to 76 mm and the effect of
incorporating cement-stabilised materials
in the base layer of UTCRCP.
The general-use FE analysis program
ABAQUS/Standard (Dassault Systemes
Simulia Corp 2016) was used. A three-layer
pavement model with an HS-SFRC layer,
base layer and subgrade with an axle load
configuration was adapted from literature
(Kim 2007). An FE analysis program was
used to allow control over all boundary
conditions and other assumptions.
Quarter symmetry and an axle load
configuration were used. Similar to Kim
(2007), the total depth and longitudinal
dimension were 21.336 m and 3.048 m
respectively. The transverse dimension
was adjusted to 4.038 m to change the load
configuration to axle loading. A pressure of
550 kPa was used to represent a standard
axle load of 80 kN, where 40 kN is applied
per side of the axle. An axle length of
1.98 m was assumed. The distance between
the wheel centreline and axle centreline
was 0.99 m. A single wheel at each axle
48
3.048 m
3.048 m
Load location
0.99 m
(a)
Figure 1 A
 xle model with (a) plan view and (b) isometric view
end was modelled. Load was applied as a
pressure on a circular area with a radius
of 152.4 mm. Figure 1 shows the geometry
of the 3D FE model in plan and isometric
view. The circular load is also indicated.
The element type used was 20 node
quadratic brick elements with reduced
integration (C3D20R). Swept meshing was
used. The mesh fineness decreased further
from the zone of interest where the pres-
sure is applied and in the axle centreline.
Figure 1(a) also gives a plan view of the
mesh. In terms of depth, six elements
were fitted into the 50 mm concrete layer,
resulting in an element height of 8.3 mm.
The same element height was used for the
base. The element height was progressively
increased to 5 m in the subgrade toward
the bottom of the model. Figure 1(b)
shows how the aspect ratio of the elements
further away from the zone of interest falls
outside the normally accepted limits with
an aspect ratio smaller than 5 generally
deemed to be acceptable (MacDonald
2011). The use of quadratic elements par-
tially mitigates the effect of element aspect
ratio (Cho et al 1996).
The HS-SFRC layer was 50 mm thick
and the base layer was 305 mm thick. The
subgrade was 20.955 m deep. Isotropic, LE
constitutive material models were used.
Table 3 Layer thicknesses and material properties
Thickness
Layer
(mm)
HS-SFRC layer 50 & 76
Base layer 305
Subgrade 20 955
Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
(b)
To simulate the effect of having a thin
HS-SFRC layer, typical material proper-
ties for HS-SFRC were used. A Young’s
Modulus of 40 000 MPa and a Poisson’s
Ratio of 0.17 were selected (Kearsley et
al 2014). The effect of cracking was not
considered. Kim (2007) used a Young’s
Modulus and Poisson’s Ratio of 207 MPa
and 0.4 for the base, and 41 MPa and 0.45
for the subgrade, and these values were
adopted for the current study.
This investigation consisted of two
parts. In the first part the effect of increas-
ing the HS-SFRC layer thickness from
50 mm to 76 mm was considered. The
substructure layer thicknesses and all
material properties remained constant. In
the second part the base Young’s Modulus
was increased incrementally, while the
concrete layer thickness and subgrade
material stiffness remained constant.
Increasing the base material stiffness gave
an indication of the effect of the use of
cement-stabilised base layers underneath
the 50 mm ultra-thin HS-SFRC layer on the
pavement response. Table 3 summarises
the layer thickness, material properties and
varied parameters.
The material stiffness of cement-
stabilised material is variable, influenced
by the extent of cracking of the pavement
Young’s Modulus
Poisson’s Ratio
(MPa)
40 000 0.17
207, 1 400, 12 600 & 37 800 0.4
41 0.45
Table 4 Critical parameters of 50 mm concrete layer and 76 mm concrete layer models The deflection bowl as shown in
Critical parameter
Figure 2 indicates that the deflection
around the load location decreased as
Thickness δv surface σh bottom of BL σv top of subgrade εv top of subgrade
a result of the increased concrete layer
(mm) (kPa) (kPa) (μm/m)
thickness. The difference in deflection
50 mm –0.862 6 914 –33.8 –749 diminishes between an offset of 0.5 m and
1 m to the right of the load location. In the
76 mm –0.692 4 637 –22.3 –466
axle centreline the deflection was slightly
greater for the 76 mm thick concrete
layer model.
Axle CL Wheel CL
Effect of cement-stabilised bases
Offset from wheel centreline (m)
–1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 The effect of including bound granular
0 material, in the form of cement-stabilised
–0.1 granular material, in the design of
–0.2 UTCRCP is of interest. In this section the
–0.3
effect of varying the base material stiffness
Deflection (mm)

to include the range of cement-stabilised

–0.4
–0.5
–0.6
–0.7
–0.8
–0.9
–1.0
Figure 2 T ransverse deflection bowls of models with 50 mm and 76 mm concrete layers
layer, as well as whether the overlaying layer
is bound or unbound. A wide range of base
material stiffness values (Young’s Modulus)
was used, varied in multiples of 1 400 MPa.
The maximum value was close to that
used as the concrete Young’s Modulus
(40 000 MPa).
The pavement response was evaluated
by using selected critical parameters,
transverse deflected shapes and the stress
in the base. The critical parameters were
the vertical deflection of the surface, the
transverse horizontal tensile stress at the
bottom of the HS-SFRC layer, and the
vertical compressive stress and strain at
the top of the subgrade. All the parameters
were in the wheel centreline.
Results
50 mm
76 mm
2010). The ratio of the transverse horizontal
tensile stress in the 50 mm thick concrete
layer and the flexural strength of NSC is
greater than 0.85. Although the stress in the
concrete layer does not exceed the flexural
strength, it indicates that the concrete layer
would crack after a limited number of load
cycles. This enforces that HS-SFRC should
be used for thin concrete layers supported
by unbound granular substructures. The
high strength of HS-SFRC ensures that the
stress level (ratio of tensile stress of flexural
strength) remains low, while incorporation
of steel fibres reduces stress concentrations
at crack tips when they do occur. The lower
stress and strain at the top of the subgrade
show how the thicker concrete layer radiates
(or spreads) the load (and stress) further
from the load location, also resulting in a
lower vertical deflection.
materials was investigated.
Table 5 shows the critical parameters
of the respective base materials stiffnesses
(in term of Young’s Modulus) used. The
response of a pavement model with base
Young’s Modulus of 207 MPa (typical for
granular material not cement-stabilised)
is included for reference. The subgrade
properties remained constant with a depth
of 20.955 m, Young’s Modulus of 41 MPa
and Poisson’s Ratio of 0.45.
All the critical parameters decreased as
the base material stiffness was increased.
This is because load spreading through the
base improved. The horizontal stress at
the bottom of the concrete layer became
a compressive stress as the base material
stiffness was increased.
For unreinforced concrete pavements,
it is necessary to limit the tensile stresses
in the concrete to, in turn, limit the forma-
tion of cracks, but with UTCRCP, which
is optimised to contain sufficient steel
reinforcing for the post-cracking strength
to match or exceed the cracking strength
of the concrete, tensile stresses can be
resisted without cracks opening up, even
after the concrete has cracked. These
results indicate that the advantage of the
relatively high tensile strength and ductility

Increasing the HS-SFRC Table 5 Critical parameters of models with increasing base material stiffness
layer thickness Critical parameter
The difference in the pavement response Material
stiffness δv surface σh bottom of BL σv top of subgrade εv top of subgrade
of 50 mm and 76 mm thick concrete
(mm) (kPa) (kPa) (μm/m)
layers was determined by evaluating
the critical parameters and deflected 207 MPa –0.862 6914 –33.8 –749
shape. Table 4 shows that all the critical
1 400 MPa –0.515 2185 –14.6 –294
parameters decreased when the thickness
was increased. 12 600 MPa –0.317 –394.6 –5.07 –71.0
The flexural strength of NSC is typically
37 800 MPa –0.273 –319.6 –3.38 –38.1
smaller than 8 000 kPa (Domone & Illston

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 49
 of the thin SFRC layer used in UTCRCP
700 can be utilised when the supporting layer
37 800; 605
lacks stiffness.
600
Horizontal tensile stress (kPa)

Figure 3 shows the effect of increasing

500 12 600; 471
the base material stiffness on the trans-
verse horizontal stress at the bottom of the
400
base layer. A tensile stress is induced, and
300 1 400; 266 the relationship is logarithmic and ranges
between 86.9 kPa and 605 kPa. As the
200
stiffness of the base increases, the tensile
207; 86.9
100 y = 98.117ln(x) –441.41 stress increases and the ability of the layer
R2 = 0.9975 to not crack (tensile strength) becomes
0
100 1 000 10 000 100 000 more important.
Base material stiffness (MPa) In Figure 4 the horizontal tensile stresses
in the base versus material stiffness are
Figure 3 H
 orizontal tensile stress at bottom of base compared to the uncracked material stiff-
ness and flexural strength combinations
summarised in the background on cement
3 000 stabilisation from TRH 4: Structural Design
Horizontal tensile stress & flexural strength (kPa)

Horizontal tensile stress

Stress level (S) = of Flexible Pavements for Interurban and
Flexural strength
2 500 C1: S = 0.24 Rural Roads (Department of Transport
1996) and TRH 13: Cementitious Stabilizers
in Road Construction (Department of
2 000
Transport 1986). The function reported
in Figure 3 was used to determine the
1 500
load-induced horizontal tensile stress for
C3: S = 0.40
C1: S = 0.36 the material stiffnesses of the respective
C2: S = 0.41
1 000 strength classes in Table 1. These stresses
C2: S = 0.53 were divided by the flexural strength of the
C4: S = 0.85
500 C3: S = 0.69 strength classes to calculate the stress level
(S), which was also included in the Figure 4.
C4: S = 1.22 The range of the y-axis of Figure 4 is greater
0
100 1 000 10 000 100 000 (0 to 2 500 MPa) than in Figure 3 (0 to
Base material stiffness (MPa) 700 MPa).
The stress level ranged from a
C1 material (flexural strength range) C2 material (flexural strength range)
minimum of 0.24 for C1 materials to a
C3 material (flexural strength range) C4 material (flexural strength range) maximum of 1.22 for C4 materials, with
Horizontal tensile stress at bottom of base the number of cycles that can be endured
decreasing as the strength decreased. The
Figure 4 C
 omparison of material stiffness – flexural strength ranges to tensile stress values from fatigue life of cemented materials loaded
models to a stress level under 0.50 is generally
good, being able to absorb millions of load
cycles (Xie et al 2018). For the three-layer
Axle CL Wheel CL pavement system modelled using 3D LE
FE, the stress levels calculated for the base
Offset from wheel centreline (m)
–1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 layer underneath a 50 mm HS-SFRC layer
0 indicated that C1 and C2 materials would
–0.1 perform well under 80 kN axle loads. C3
–0.2 and C4 materials would deteriorate faster.
–0.3
Figure 5 shows the effect of increasing
Deflection (mm)

the base material stiffness on the deflection

–0.4
bowl, where the concrete layer and base
–0.5
start acting as one layer and the deflection
–0.6 bowl shape becomes similar to that of the
–0.7 207 MPa thick concrete layer models.
–0.8 1 400 MPa The LE FE analysis showed that the
12 600 MPa
–0.9 inclusion of cement-stabilised granular
37 800 MPa
–1.0 material would reduce all critical param-
eters. However, this analysis does not cap-
Figure 5 D
 eflected bowls of increasing base material stiffness ture the possibility of the cemented material

50 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
cracking due to material and traffic-loading
variability. The effect of discretely cracked
C1 and C2 materials underneath a 50 mm
HS-SFRC layer could be detrimental to the
layer underneath it, because cemented layer
crack edges may cause stress concentrations
in the adjacent layers. Physical model-
ling and FE modelling, that incorporates
advanced material models for cemented
granular materials, should be used to fur-
ther investigate the use of cement-stabilised
granular materials underneath thin concrete
layers. Further investigations should also
incorporate the repeated loading of traffic
into numerical analyses.
Conclusion
Increasing the HS-SFRC layer thickness
reduced all the critical parameters. The
effect of including bound granular material,
in the form of cement-stabilised granular
material, was investigated by varying the
base material stiffness over a wider range.
By doing this the relative stiffness was var-
ied. Increasing the base material stiffness
reduced all the critical parameters. It was
found that there was a logarithmic rela-
tionship between the tensile stress at the
bottom of the base layer and base material
stiffness. The linear elastic finite element
modelling suggests that cemented bases
underneath a 50 mm HS-SFRC layer that
use C1 and C2 materials would perform
well under standard axle loads of 80 kN,
because the stress level they are subjected
to is relatively low. The stress level of C3
and C4 materials was higher and these
materials are more likely to deteriorate
faster underneath a 50 mm HS-SFRC layer.
According to the LE FE modelling done,
the incorporation of cement-stabilised bases
would reduce all the critical param­eters,
and the stress levels in the base would fall
within acceptable ranges. However, the
possibility of the cement-stabilised mate-
rial cracking due to material variability
and/or overloading should be taken into
consideration before a conclusion can be
made of cemented base usefulness. If C1 or
C2 materials are used underneath a 50 mm
HS-SFRC layer and the cemented layer does
crack discreetly, the unbound granular layer
of subgrade underneath it will be subjected
to stress concentrations.
As pavement design is a process
intended to find the most economical
combination of layer thickness and mate-
rial types for pavements during the service
life of the road, it is of concern that the
Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021
high-tensile-strength capacity of the
HS-SFRC does not seem to be utilised to its
full potential when the thin concrete layer
is supported by a stiff supporting layer.
The use of cement-stabilised base in
UTCRCP should thus be further investigated
to ensure that the most economical com-
bination of materials is used in UTCRCP.
Advanced material models for cement-
stabilised granular materials should be used
to verify the behaviour of cemented bases
directly underneath thin HS-SFRC layers.
References
Bredenhann, S, Van Heerden, J, Strauss, P & Joubert, P
2018. Design and construction of Ultra-Thin
Continuously Reinforced Concrete (UTCRC) on N1
near Hugenot Tunnel. Proceedings, International
Conference on Concrete Repair, Rehabilitation and
Retrofitting, 31 October 2018, Cape Town. MATEC
Web Conference, Vol 199.
Briggs, M A, Valsangkar, A J & Thompson, A 2016.
Behaviour of fibre-reinforced concrete beams on
grade. International Journal of Physiscal Modeling in
Geotechnics, 16(4): 152‒159.
Brown, S F & Selig, E T 1991. The design of pavement
and rail track foundations. In O’Reilly, M P & Brown,
S F (Eds). Cyclic Loading of Soils: From Theory to
Design, 1st ed. London: Blackie, pp 249‒305.
Burmister, D M 1945. The general theory of stresses
and displacements in layered systems I. Journal of
Applied Physics, 16 (February): 89‒95.
doi.org/10.1063/1.1707558.
Chen, D H, Won, M, Chen, X & Zhou, W 2016. Design
improvements to enhance the performance of
thin and ultra-thin concrete overlays in Texas.
Construction and Building Materials, 116(July): 1–14.
Cho, Y-H, McCullough, B F & Weissmann, J 1996.
Considerations on finite-element method
application in pavements. Transportation Research
Record, 1539(1): 96‒101.
Dassault Systemes Simulia Corp 2016. ABAQUS.
Mayfield Heights, OH: Simulia Corp.
De Beer, M 1990. Aspects of the design and behaviour
of road structures incorporating lightly cementitious
layers. PhD Thesis. University of Pretoria.
Department of Transport (DoT) 1979. TMH 1:
Standard Methods of Testing Road Construction
Materials, Method A14. Pretoria: DoT.
Department of Transport 1986. TRH13:Cementitious
Stabilizers in Road Construction. Pretoria: DoT.
Department of Transport 1996. TRH 4: Structural
Design of Flexible Pavements for Interurban and
Rural Roads. Pretoria: DoT.
Denneman, E 2011. Fracture in high performance fibre
reinforced concrete pavement materials. PhD Thesis.
University of Pretoria.
Domone, P & Illston, J 2010. Construction Materials,
4th ed. Abingdon, UK: Spon Press.
Elsaigh, W A 2007. Modelling the behaviour of steel
fibre reinforced concrete pavements. PhD Thesis.
University of Pretoria.
Huang, Y H 1993. Pavement Analysis and Design, 2nd
ed. Englewood Cliffs, NJ: Pearson Prentice Hall.
Ioannides, A M 2006. Concrete pavement analysis: The
first eighty years. International Journal of Pavement
Engineering, 7(4): 233‒249.
Jordaan, G J 1984. Bituminous pavement rehabilitation
design. Master’s dissertation. University of Pretoria.
Kannemeyer, L, Perrie, B D, Strauss, P J & Du Plessis,
L 2007. Ultra-thin continuously reinforced concrete
pavement research in South Africa. Proceedings,
International Conference on Concrete Roads, 16‒17
August 2007, Midrand, South Africa. Cement &
Concrete Institute report, pp 97‒124.
Kearsley, E P & Mostert, H F 2010. Enabling the
effective use of high-performance fibre reinforced
concrete in infrastructure. In Van Zijl, G P A G
& Boshoff, W P (Eds). Advances in Cement-Based
Materials. London: CRC Press, pp 53‒58.
Kearsley, E P, Steyn, W J VdM & Jacobsz, S W 2014.
Centrifuge modelling of Ultra Thin Continuously
Reinforced Concrete Pavements (UTCRCP). In
Gaudin, C & White D (Eds). Proceedings, 8th
International Conference, Physical Modelling in
Geotechnics, 14‒17 January 2014, Perth, Australia:
Taylor & Francis, pp 1101–1106.
Kim, M 2007. Three-dimensional finite element
analysis of flexible pavements considering
nonlinear pavement foundation behaviour.
PhD Thesis. University of Illinois at Urbana-
Champaign, US.
Lv, S, Xia, C, Liu, H, You, L, Qu, F, Zhong, W et al
2021. Strength and fatigue performance for cement-
treated aggregate base materials. International
Journal of Pavement Engineering, 22(6): 690‒699.
https://www-tandfonline-com.uplib.idm.oclc.org/
doi/full/10.1080/10298436.2019.1634808.
MacDonald, B J 2011. Practical Stress Analysis with
Finite Elements, 2nd ed. Dublin, Ireland: Glasnevin
Publishing.
Neville, A M & Brooks, J J 2010. Concrete Technology,
2nd ed. Harlow, UK: Pearson.
O’Flaherty, C A 1967. Highway Engineering. 2nd ed.,
Volume II. London: Edward Arnold.
Pereira, D D S, Balbo, J T & Khazanovich, L 2006.
Theoretical and field evaluation of interaction
between ultra-thin whitetopping and existing
asphalt pavement. International Jounral of Pavement
Engineering, 7(4): 251‒260.
Rao, S & Roesler, J R 2005. Characterization of effective
built-in curling and concrete pavement cracking on
the Palmdale test sections. Working paper. Davis,
CA: University of California, Davis Institute of
Transportation Studies.
SABITA (South African Bitumen and Tar Association)
1993. GEMS – The design and use of granular
emulsion mixes. Cape Town: South African Bitumen
and Tar Association.
51
SANRAL (South African National Roads Agency
Limited) 2013. Pavement design. In South African
Pavement Engineering Manual. Pretoria: SANRAL.
Visser, A 2017. Potential of South African road technology
for application in China. Journal of Traffic and
Transportation Engineering (English Ed), 4(2): 113‒117.
52
Werkmeister, S, Dawson, A R & Wellner, F 2004. of cement-treated aggregate base materials
Pavement design model for unbound granular under different stress states. Applied Sciences,
materials. Journal of Transportation Engineering, 8: 1‒23.
130(5): 665‒674.
Xie, J, Tang, L, Lv, S, Zhang, N, Huang, T & Liu, H
2018. Standardization of fatigue characteristics
Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Priority infrastructure for TECHNICAL PAPER
minibus-taxis: An analytical Journal of the South African
Institution of Civil Engineering

model of potential ISSN 1021-2019

Vol 63 No 4, December 2021, Pages 53–65, Paper 1228

benefits and impacts LOURENS DE BEER is a PhD student at the

University of Pretoria. He completed his
Honours degree in 2018, followed by his
Master’s, with distinction, in 2019. His
L R De Beer, C Venter research interests lie in public transport
network modelling and optimisation,
public transport infrastructure, and
logistics.
Many governments in the global south are grappling with challenges of improving the quality
of informal transport, and an inability to pay for service improvements. This paper asks the Contact details:
Department of Civil Engineering
question whether efficiency benefits might be gained through strategic implementation of University of Pretoria
once-off infrastructure interventions providing priority to informal vehicles at intersections. We Hatfield 0002
note that informal drivers already indicate this demand through (illegal) driving behaviour in South Africa
T: +27 71 479 1873
traffic. We use a drone to observe indicative behaviours among minibus-taxi drivers in South
E: lourensrdb@gmail.com
Africa. We identify interventions that would formalise this behaviour: a single lane pre-signal
strategy, queue-jumping lane, and dedicated public transport lane. The objective of the paper PROF CHRISTO VENTER (Pr Eng, FSAICE) is a
is to quantify the potential economic impacts of such treatments on minibus-taxi operators, Professor of Transportation Engineering in
passengers and other road users. The findings indicate that substantial savings could be the Department of Civil Engineering at the
University of Pretoria. His teaching and
realised in terms of travel time, user cost, and operating cost to taxi passengers and drivers
research activities focus on public transport,
without additional costs being incurred by other road users. The single-lane pre-signal strategy, transport policy and planning, travel
the queue-jumping lane and the dedicated taxi lane saw a decrease in total hourly cost of 12%, demand modelling, and social aspects of
14% and 30% respectively, including construction cost, user cost, and agency cost, indicating a mobility.

net social benefit. If part of these savings were passed on to passengers, priority infrastructure
could serve as an implicit subsidy to public transport users.
INTRODUCTION
The paratransit industry in South Africa
has grown from a modest provider of
public transport to the largest supplier of
mobility to the urban public. Small-scale
ownership of minibus-taxis enabled the
industry to develop in an adaptive and flex-
ible way where the fares remain low, and
the services respond rapidly to any change
in need from the passengers (Jennings &
Behrens 2017).
Recent initiatives to overhaul South
Africa’s entire public transport system,
partly to address the deficiencies of the
minibus-taxi system, have often resulted
in a complex set of formal and paratransit
operations which are independent of each
other, subject to a regulatory framework
that is disconnected (Salazar Ferro et al
2012). There have been some efforts to
improve the infrastructure for minibus-
taxi facilities and operations, including
undercover loading areas, public toilets,
and office space (Schalekamp & Klopp
2018). In the early 2000s the City of
Johannesburg took a step towards imple-
menting dedicated taxi lanes as a part of
its Strategic Public Transport Network
Contact details:
Department of Civil Engineering
Centre for Transport Development
University of Pretoria
Hatfield 0002
(SPTN), but this was abandoned in favour South Africa
of Bus Rapid Transit (BRT). However, the T: +27 12 420 2184
E: christo.venter@up.ac.za
realities of the slow and expensive roll-out
of BRTs, coupled with the realisation that
the minibus-taxi has a continuing role to
play in a hybrid public transport system,
has turned the attention of some authori-
ties back towards dedicated infrastructure
for this mode.
Unfortunately, the evidence base on
which to find the planning and design
of priority infrastructure for minibus-
taxis is very thin. Qualitative studies
have documented the response of the
minibus-taxi industry to proposed changes
and formalisation of the industry fairly
well (Schalekamp & Behrens 2010; 2013).
Research on the driving behaviour of
minibus operators is limited. Some simula-
tion tools have been developed to help
model driver behaviour and route evolution
of taxis (Gu et al 2012; Hager et al 2015;
Neumann et al 2015; Zheng et al 2020).
However, no systematic exploration of
alternative infrastructure-based interven-
tions for minibuses has been done. It is
the contention of this research that such Keywords: paratransit, minibus-taxi, economic model,
interventions, when applied judiciously, priority infrastructure, operating cost

De Beer LR, Venter C. Priority infrastructure for minibus-taxis: An analytical model of potential benefits and impacts.
J. S. Afr. Inst. Civ. Eng. 2021:63(4), Art. #1228, 13 pages. http://dx.doi.org/10.17159/2309-8775/2021/v63n4a6 53
may raise the overall cost-effectiveness of
minibus operations, and deliver benefits
to users and operators. Moreover, it may
be possible to do so without substantially
degrading the level of service offered to
other road users. Accordingly, the aim
is to quantify, using relatively simple
mathematical modelling, the benefits that
minibus-taxi operators and passengers
receive when they skip traffic queues at
intersections during congested periods of
the day. An analytical approach was devel-
oped for a single bi-directional corridor
with intersections, avoiding for now the (a) t = 0 s (b) t = 8 s (c) t = 10 s
complexities of simulating entire networks.
The model is a first effort to derive metrics
for the costs and benefits of operators, pas-
sengers and private car users, and does not
address issues of safety or design.
The paper starts with the observation
that minibus-taxi drivers already display
driving behaviour that simulates priority
access, even in the absence of such infra-
structure (and therefore often under unsafe
and illegal conditions). We use remote
detection to identify such behaviour and
suggest intersection treatments to formal-
ise such priority access. Then follows a (d) t = 11 s (e) t = 14 s (f) t = 16 s
description of the public transport priority
Figure 1 Minibus-taxi creating own informal priority, Case 1
measures (including pertinent literature), of
the analytical model used to evaluate them,
and the results. Lastly, we present conclu-
sions and recommendations for implemen-
tation and further research.

ILLUSTRATIVE OBSERVATIONS
OF TAXI DRIVER BEHAVIOUR
Minibus-taxi operators often try to cut
corners (literally and figuratively) in their
efforts to save time – this is mainly due
to pressure being put on them by their
passengers and their need to survive finan-
cially. The need to maximise income by
(a) t = 0 s (b) t = 4 s (c) t = 6 s
finding more passengers and reducing cycle
times to complete more round trips dur-
ing the peak period means that it is often
in their best interest to weave their way
through traffic to get ahead of congestion
(Govender & Allopi 2007).
With the use of an unmanned aerial
vehicle, commonly referred to as a “drone”,
this behaviour was observed along various
corridors in the Pretoria area. This is meant
as an observational study to find exemplars
of such behaviour and their implications,
rather than an exhaustive survey of behav-
iours. The following three cases illustrate
(d) t = 8 s (e) t = 10 s (f) t = 12 s
the delay advantage that operators try to
gain at intersections with long queues. Figure 2 M
 inibus-taxi creating own informal priority, Case 2

54 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering

(a) t = 0 s
(d) t = 4 s
Figure 3 M
 inibus-taxi creating own informal priority, Case 3
Direction of bus movement
Bi-directional lane segment
Figure 4 ( a) Intersection with single-lane approaches, (b) pre-signal strategy (adapted from
Ilgin Guler et al 2015)
Case 1 (queue-skipping behaviour)
In Figure 1 a minibus-taxi on a through-
movement is observed driving in the right-
turn lane. After the traffic signal turns green,
the taxi is seen cutting into the lane adjacent
to it, thereby effectively skipping eight vehi-
cles in the queue – this is due to the adjacent,
right-turning lane having a shorter queue
length. The behaviour is like a queue-jump-
ing lane type of infrastructure, and jumping
past such a long queue of vehicles saves this
particular taxi approximately 24 seconds.
Case 2 (queue-skipping behaviour)
The second case, as Figure 2 illustrates,
is like the first in that the operation of an
informal queue-jumping lane is observed.
This time, however, two minibus-taxis
skip the queue as soon as the traffic signal
Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021
(b) t = 2 s (c) t = 3 s
(e) t = 5 s (f) t = 6 s
Main
signal
Opposite direction
Upstream x2u
pre-signal
Downstream
pre-signal
x2d
x1
turns green. From their behaviour the taxi
travelling behind attempts to push in first
in the centre lane after which allowing
the taxi in front of it to do the same. This
illustrates the sense of community minibus-
taxi operators have, attempting to help the
other out when the opportunity arises. In
this case, the two taxis skip a queue of over
12 vehicles and can save approximately 66
seconds because they avoid being stuck
in the overflow queue at the end of the cycle.
Case 3 ( opposite-lane driving
behaviour)
In the final case observed, as illustrated in
Figure 3, a minibus-taxi is seen travelling
in the lane of the oncoming traffic after
which it makes a right turn. This behaviour
is more dangerous than the previous two
cases. Only one second is saved in this
process, as the queue that forms at the
intersection only amounts to the single
vehicle travelling in front of it.
Formalising the driving behaviour as
depicted in the cases illustrated might, in
theory, reduce the delay experienced by
minibus-taxi drivers and passengers, while
mitigating the problems with safety and
habitual flaunting of traffic rules. In the
next section three potential strategies are
identified to formalise priority to public
transport vehicles.
PUBLIC TRANSPORT
PRIORITY MEASURES
Public transport priority measures are inter-
ventions made to provide public transport
vehicles with a competitive time advantage
over private vehicles. These interventions can
be either physical or policy-related, like a bus-
only roadway or legislation requiring private
vehicles to yield to buses (Halifax Regional
Municipality (Canada) 2018). This research
considers the currently available public
transport priority measures that have proved
to be effective in the public transport sphere,
particularly pertaining to buses. These
infrastructure forms include the s­ ingle-lane
pre-signal strategy, queue-jumping lane, and
dedicated public transport lane.
Kerbside bus stops
The most basic form of infrastructure
intervention is the construction of passen-
ger loading bays. Although much provision
has been made for bus stops, little attention
has been paid to providing stopping facili-
ties for taxis. Bus service times at a bus
stop occupy a large proportion of the total
operational time the bus spends on the
road, and the occurrence of queues form-
ing at the entry and departure area of a
kerbside bus stop is frequent. Regarding the
bus stop design, bus size, and congestion,
Tirachini (2014) stated that buses have the
lowest capacity at the bus stop component
of a bus route, and the first element subject
to congestion.
Single-lane pre-signal strategy
Ilgin Guler et al (2015) proposed a strategy
whereby buses are given priority at signalised
intersections with single-lane approaches by
adding traffic signals to the road such that a
bus can jump a portion of the car queue by
making use of the travel lane in the opposite
direction. Two additional pre-signals are
placed upstream at a distance x2u km and
55
downstream at a distance x2d km from the
main signal. These two signals then operate
together to create an intermittent bus prior-
ity lane. When there is no bus present both
the pre-signals will remain green, and cars
will be able to discharge through the inter-
section normally. When a bus approaches
and reaches a distance x1 km from the main
signal, both pre-signals at x2u and x2d turn
red, indicating to cars from both directions
to stop. The bi-directional segment is now
cleared, and the bus is free to drive onto
the opposite lane and travel without being
impeded until it can merge back onto its Figure 5 S chematic representation of the kerbside taxi stop
original lane. Figure 4 (p 55) illustrates
the setup.
The authors quantified the delay sav-
ings that the buses achieved, as well as
the negative impact that cars experienced
when this method was applied. The study
found that, in the under-saturated case, sig-
nificant bus delay savings and/or improved
system-wide delays overall can be achieved
Minibus-taxis queue in
with single-lane approaches under the fol- left lane, private vehicles
lowing conditions: queue in centre lane.
QQ V/C less than 0.85
QQ A distance of at least 7 m between the
pre-signal location and the intersection
QQ When a turning ratio, from the cross-
street, of less than 25% is observed.
A theoretical analysis of an over-saturated
case, however, suggests that, although the
average bus delay savings can be up to
30 seconds, the loss in capacity can be as Dedicated green to taxis
allows their queue to
much as 25%. dissipate. Left turns allowed.

Queue-jumping lane
Extensive research has been conducted in
the functioning and operation of queue-
jumping lanes (Bhattacharyya et al 2019;
Zhou & Gan 2009). A queue-jumping
lane allows the high-occupancy vehicle
to bypass queued traffic, giving them
the opportunity to gain an advantage at Green to all vehicles.
a signalised intersection. As the vehicle Only left turns allowed
in left lane.
approaches the intersection, it leaves
the queue and enters the queue-jumping
lane. A priority signal, thereafter, allows
the queue to clear before the main green Figure 6 S chematic representation of the queue-jumping lane
stage commences.
Zlatkovic et al (2013) evaluated the The greatest benefit to the BRT is observed The largest drawback in the implemen-
individual and combined effects of a with the combination scenario ‒ the BRT tation of the public transport preferential
queue-jumping lane and public transport travel times are reduced by between 13% treatment is the deterioration of the vehic-
signal priority on the performance of a and 22%, there is a significant improvement ular traffic performance on a network-wide
BRT (Bus Rapid Transit) system. They of the progression of the BRT vehicles level, the majority of which was observed
found that for each case, namely, queue- through the networks, intersection delays on cross-streets.
jumping, public transport signal priority, and waiting time are reduced, speed
and a combination of the two, the BRT is increases significantly by 22%, and the trav- Dedicated bus lane
offered significant benefits whereas certain el time, reliability, and headway adherence Dedicated bus lanes fundamentally improve
impacts are imposed on vehicular traffic. are better than the other two scenarios. the effectiveness of public transport when

56 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering

A mix of minibus-taxis and private
vehicles form. Taxis adjacent to
the red lane can cross over and
pick up and drop off passengers.
Minibus-taxis receive a priority
green. Only taxis in the red
lane can use priority.
The queue of private
vehicles can dissipate.
Figure 7 S chematic representation of the single-lane pre-signal strategy
implemented at a city level (Brasuell 2019;
Glambrone & Acitelli 2019). Ben-Dor et al
(2018) exploited MATSim’s capabilities to
emulate how a traveller would adapt to vary-
ing transportation possibilities and found
that not only do dedicated bus lanes result
in the same public transport characteristics
to be observed during peak hours as with
off-peak hours, but an increase of 20% in
public transport use was also observed dur-
ing congested conditions.
Stamos et al (2012) evaluated the HOV
(high occupancy vehicle) lane in the central
business district of Thessaloniki, Greece,
where the primary objective was to allevi-
ate the impacts of traffic and congestion in
the city. The implementation of the HOV
lane saw a 6% drop in traffic due to the
decreased number of vehicles transporting
more than two passengers that can use the
lane. The decrease was partially balanced
by the demand that was induced due to
the attractiveness of the lane. The slight
decline in traffic, together with the prohib-
ited turning movements in the lane, caused
Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021
a 62% reduction in delay and an increase in
speed of 129%.
MODEL DESIGN
Four forms of infrastructure are modelled,
namely, a kerbside taxi stop, a queue-jump-
ing lane, a single-lane pre-signal strategy,
and a dedicated taxi lane. The objective
of the model is to quantify the high-level
economic impact that the selected priority
infrastructure would have on the paratransit
operators, taxi passengers, other road users,
and the agency providing the infrastructure.
This is in keeping with the definition of total
cost as including costs to both users and
operators/infrastructure owners, consistent
with the notion of net social welfare. This
means that the model will consist of four
main sections which include:
1. The signalised intersection design
which determines the cycle length, red
phase length, and green phase length.
2. The user cost which entails the time
passengers in the minibus-taxis, as well
as private vehicle owners, spend on
the road.
3. The operating cost, which is based
on time spent on the road as well as
the distance covered and includes all
the costs associated with operating a
minibus-taxi or a private vehicle.
4. The capital cost, which is the cost
associated with constructing each
of the four forms of public transport
infrastructure.
Model parameters
The intersection consists of a north-
south and an east-west two-lane road.
The minibus-taxis and regular vehicles
travel mixed, as there is no priority for
the paratransit vehicles at the intersection
pertaining to the kerbside stop. This form
of infrastructure will be considered as the
base case against which to compare all the
subsequent forms of priority infrastructure.
Figure 5 illustrates the schematic model
upon which calculations are based. All taxi
stops in the subsequent figures are indicat-
ed with a red triangle. For simplicity’s sake
only the west-to-east and north-to-south
movements are modelled, but the results
can easily be generalised for all directions.
The second public transport priority
infrastructure, the queue-jumping lane,
allows minibus-taxis to skip the entire
queue at the intersection by providing them
with a dedicated section of road. During the
red cycle phase, taxis can drop off and pick
up travellers in the dedicated lane but are
not able to make stops during the priority
green phase or the all-green phase. For this
purpose, a far-side kerbside stop is retained
to allow for loading and unloading during
the green cycle phase. The percentage of
taxis stopping to pick up or drop off passen-
gers is based on an input value in the model.
The operation of the infrastructure in its
three stages is illustrated in Figure 6.
The third priority infrastructure, the
single-lane pre-signal strategy, provides
taxis with a time advantage without incur-
ring significant construction costs. The
length of the priority section of road is
designed to account for the number of pri-
vate vehicles that queue over the duration
of the east-west green phase. Only taxis
adjacent to the priority section of road are
permitted to use it to gain a time advan-
tage. The three phases of the operations are
illustrated in Figure 7. It is noted, however,
that boarding or alighting a minibus-taxi
in the middle of the road is dangerous, and
that a raised kerb in the centre of the road
57
Table 1 Input variables used in the signalised intersection design
Variable
QQ
Average delay per vehicle QQ
(private vehicles), davg
Cycle length in seconds, C QQ
QQ
Arrival rate in vehicles/second, v
Departure rate in vehicles/second, s QQ
may have to be constructed to account for
this issue.
The final priority infrastructure, the
dedicated taxi lane, is expected to provide
public transport with the greatest amount
of time saving whilst minimising the
delay experienced by regular traffic. The
representation of the dedicated taxi lane is
illustrated in Figure 8.
Signalised intersection design
The design of the intersection forms the
base of the model development ‒ the signal
plan determines the waiting time at the
intersection, as well as the queue lengths
that form as a result. These values are then
used to determine the subsequent user
costs and operating costs.
Table 1 provides the input variables
used in the signalised intersection design.
Each variable is briefly explained.
A key assumption is that the average
delay for private vehicles is kept constant
at 12 seconds, corresponding to a Level
of Service B. In normal traffic analysis
the delay is estimated as a function of
arrival and departure flow rates, red times,
and cycle lengths. However, we turn this
analysis around by fixing the delay, and
calculating the red time that is needed for
a given cycle length and departure rate.
This imposes limits on the capacity of the
intersection, but allows us to focus on cases
where the minibus-taxis are provided with
some form of priority without deteriorating
conditions for private vehicle users.
Table 2 summarises the arrival rates
assumed for private vehicles and minibus-
taxis at high and low-flow scenarios, obtained
from a typical corridor in the Pretoria CBD.
For the base case (kerbside taxi stop)
the average delay per vehicle is given by the
following standard equation for undersatu-
rated signal approaches (Transportation
Research Board 2013):
2c⎫⎪⎭1 – s ⎫⎪⎭
r2
davg = (1)
v Flow (north-south)
58
Description Value used
Used as an input value to determine the red cycle time for each case.
In the infrastructure forms where minibus-taxis receive a priority signal, their average delay 12 seconds
is calculated separately.
The time to complete a full traffic intersection cycle. 80 seconds
The arrival rate was based on traffic counts that were carried out on a road corridor where
Varied (see Table 2)
different transportation modes operate.
Minibus-taxis and private vehicles are assumed to have the same saturation flow rate. 0.5 veh/s (1 800 veh/h)
Figure 8 S chematic representation of the dedicated taxi lane
Where: private vehicles and minibus-taxis. For the
r : Effective red time for a traffic move- low-flow case, the corresponding red time
ment in seconds and queue length are 36 seconds and 6.2
C : Cycle length in seconds vehicles respectively. Note that, for the high-
v : A rrival rate in vehicles/seconds of flow case, the entire queue just dissipates
taxis and private vehicles by the end of the green; this delivers a lower
s : Departure rate in vehicles/seconds performance boundary for undersaturated
operations. As a first approximation we
Rearranging in terms of r gives an expression ignore stochastic effects that might cause
for the effective red and effective green times: extra delay due to occasional oversaturation.
For the queue-jumping lane and the
r = davg ∙ 2C ∙ ⎫⎪ ⎫⎪(2)
⎭s⎭
v single-lane pre-signal strategy, the queu-
ing diagrams depict two red phases (one
for taxis and one for private cars) and
g = C – r(3) two green phases (one for taxis only and
one for all vehicles) (Figure 10). The same
The queueing diagrams for both the high design applies to both forms of infra-
and low-flow cases in the west-to-east structure, as their methods of providing
direction are shown in Figure 9. minibus-taxis with a pre-signal priority
The maximum queue and approach are similar. In the red phase both the
capacity is easily estimated from the minibus-taxi (t) and the private vehicle (c)
graphs. Over the 23-second red phase queues start to build. The minibus-taxis
of the high-flow traffic case, a queue of then receive a priority green of g t seconds,
9.3 vehicles forms from a combination of after which the minibus-taxis and private
Table 2 Arrival rate of private vehicles and minibus-taxis
Private vehicle arrivals Minibus-taxi arrivals
Traffic flow rate
vc (veh/h) v t (veh/h)
High-flow case (east-west) 1 090 350
Low-flow case (east-west) 534 81
534 81
Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
 40 40
(a) High-flow case (b) Low-flow case
35 35
Cumulative number of vehicles

Cumulative number of vehicles

30 30

25 25

20 20

15 15

10 10

5 Queue length 5
Queue length
0 0
0 20 40 60 80 0 20 40 60 80

r g r g
Time (s) Time (s)
Total arrivals Total departures Total arrivals Total departures

Figure 9 S ignalised intersection queuing graph for the kerb-side taxi stop (“g” denotes green, and “r” denotes red)

vehicles travel in the same lane as mixed rather by shortening the red time. This the delay is no longer given by Equation 1,
traffic. This introduces an inflection reduces the green time for the cross traffic but can be shown to be equal to:
point on the arrival curve after rc seconds, in the north-south direction, as well as its
corresponding to the sum of the private capacity, and possibly its level of service. r c vc
davg = (4)
vehicle and taxi arrival flows. As before, at This reduction is easily estimated using the 2(vc + t)
these flow rates the entire queue dissipates cycle length and red and green times.
by the end of the cycle. The length of the red cycle for mixed Where:
The dedicated green phase for the traffic is once again determined to keep the rc : Red phase for cars (and mixed
minibus-taxis is not granted at the cost average delay at 12 seconds per vehicle. Due traffic after priority green phase for
of green time for the private vehicles, but to the inflection point on the arrival curve, minibus-taxis) (s)

35 35
(a) High-flow case (b) Low-flow case

30 30
Cumulative number of vehicles

Cumulative number of vehicles

25 25

20 20

15 15

10 10
t Queue length t Queue length
5 5
c Queue length c Queue length
0 0
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80

rt gt gc+t rt gt gc+t

rc Time (s) rc Time (s)

Mixed-traffic arrivals Minibus-taxi arrivals Mixed-traffic arrivals Minibus-taxi arrivals
Mixed-traffic departures Minibus-taxi departures Mixed-traffic departures Minibus-taxi departures

Figure 10 S ignalised intersection queuing diagram for the queue-jumping lane and the single-lane pre-signal strategy (“g” denotes green,
“r” denotes red, “c” denotes cars, and “t” denotes taxis)

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 59
 30 30
(a) High-flow case (b) Low-flow case

25 25
Cumulative number of vehicles

Cumulative number of vehicles

20 20

15 15

10 10

t Queue length t Queue length

5 5
c Queue length
c Queue length

0 0
0 20 40 60 80 0 20 40 60 80
Time (s) Time (s)
Mixed-traffic arrivals Minibus-taxi arrivals Mixed-traffic arrivals Minibus-taxi arrivals
Mixed-traffic departures Minibus-taxi departures Mixed-traffic departures Minibus-taxi departures

Figure 11 S ignalised intersection queuing diagram for the dedicated taxi lane (“g” denotes green, “r” denotes red, “c” denotes cars, and “t” denotes taxis)

vc : Arrival rate in vehicles/second for cars (gt) time allows an average of two taxis to User cost
rc+t : Combined arrival rate in vehicles/ skip the queue over each cycle and for the Determining the user cost depends on the
second for cars and minibus-taxis queue to dissipate. The length of the sec- relevant vehicle characteristics for both
tion of road on which minibus-taxis queue private vehicles and minibus-taxis. Table 3
With davg known, the length of the red should be at least 11 m long to accom- shows typical values determined from
phase for mixed traffic rc can be extracted. modate these queues. Over the first part of observations performed on traffic footage
The length of the dedicated green for taxis the cycle, taxi delay averages 10.8 seconds, obtained in the Hatfield area in Pretoria, or
is set simply at the value required to dis- dropping to 8.8 seconds in the mixed-traffic as suggested by relevant literature.
charge the taxi queue, thus rt = rc – rc(vt /s). phase. The user cost for minibus-taxis consists
Average delay for minibus-taxis is now a The queuing diagram for the dedicated of the sum of the estimated service time,
combination of the delay of taxis using taxi lane intersection is shown in Figure 11. waiting time at the red traffic signal phase,
the dedicated green signal, and the delay The additional lane means that taxis and time taken to accelerate and decelerate,
of taxis arriving after the dedicated signal private vehicles have independent arrival and travel time. For cars, this variable is
ends. The former delay value is estimated and departure curves. The red time is still the same as that for minibus-taxis, except
using Equation 1 with r = rt , v = vt, and determined from Equation 1, keeping davg service time is excluded. Taxi fares are
C = rc. The latter delay comes from cal- at 12 seconds for private vehicles. The excluded to avoid double-counting of
culating the area between the arrival and delay for taxis is lower due to their lower operating costs. The total travel time for
departure curves for the combined green arrival rate. minibus taxis is given by the equation:
phase and dividing it by the total arrivals The private vehicle queue of 13.7 vehi-
during this phase which follows Equation 1. cles on average dissipates after 47 seconds, TT = Ts + Tr + Ta + Tt(5)
Figure 10 shows that, for the high-flow whereas the minibus-taxi queue of
case, providing minibus-taxis with a pre- three vehicles per cycle dissipates after Where:
signal priority of 4.2 seconds effective green 27 seconds. Ts : Total service time (in hours)

Table 3 Input variables used in calculating user cost

Variable Description Value used

Rates are assumed to be equal and the same for private vehicles and
Acceleration and deceleration rate, a 3.5 m/s2
minibus-taxis

Cars: 1.5 passengers

Vehicle capacity, pax The number of passengers transported by the vehicle
Minibus-taxi: 18 passengers

The passenger handling time includes the time a passenger takes to board
Passenger handling time, H 8 seconds/passenger
and alight a minibus-taxi

3 seconds
Time for opening and closing doors, Cd Value assumed to equal that of a BRT
(Transportation Research Board 2013)

This speed forms part of the calculations determining the total service time
Speed on entering the kerbside stop 3 m/s
of a minibus-taxi on the kerbside stop type of service infrastructure

60 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
Table 4 P
 rivate vehicle and minibus-taxi user Table 5 Input variables used in calculating operator cost
income group, value of time, and Variable Description Value used
proportion
The monthly salary of a minibus-
Value of Proportion Vehicle operator salary, Vs R20 000.00/month
Income group taxi operator
time (VOT) (%)
Tyres and other Contingencies and the cost of tyres
Private vehicle user R5 735.00/month
expendables, Vt per month

Low income R4.00* 7.4 The cost of maintaining a minibus-

Vehicle maintenance, Vm R4 303.00/month
taxi over a month period
Middle income R18.00 18.5 The cost to rent the premises
Facility maintenance, Vf where the minibus-taxis are stored, R811.00/month
High income R31.00 74.0 costs per month

This cost consists of

Minibus-taxi user unemployment insurance fund,
Administrative costs, Va R1 168.00/month
a cell phone payment, and a
Low income R4.00 28.1 bookkeeping cost

Middle income R18.00
High income R31.00
* The exchange rate on 12/02/2020 was
0.068 USD to 1 ZAR.
Tr : Wait time at red intersection phase
(in hours)
Ta : Time to accelerate and decelerate (in
hours)
Tt : Travel time (in hours)
In the case of the service time at a kerbside
taxi stop, the minibus-taxis make their
stops according to the following equations
45.9 Supervision and control
centre, Vc
26.0
Cost of fuel, Fc
Fuel consumption, fct, f cc
Fuel idling, fit, f ic
Vehicle-Hours, VH
Vehicle-Distance, VD
Satellite tracking and the cost of
R1 104.00/month
the vehicle
R16.48/ℓ
The price as of the 1 June 2019
(Automobile Association 2019)
7 ℓ/100km for private vehicles
The travelling component of fuel
12 ℓ/100km for minibus taxis
consumption
(Automobile Association 2013)
The idling component of fuel 1.2 ℓ/hour for private vehicles
consumption 1.5 ℓ/hour for minibus-taxis
The number of hours that a
176 hours
minibus-taxi travels in a month
The distance that the average
18 000 km
minibus-taxi operator travels in a
(Department of Transport 2008)
month

⎫ Vf ⎫
⎪ ⎪
(adapted from Bian et al 2015): and minibus-taxis respectively are from

Ta = 2 × ⎭3.6⎭
the National Household Travel Survey
⎫⎪ a ⎫⎪
(9)
⎭ 3 600 ⎭
Ts = Td + Tm(6) (Department of Transport/Statistics 2013).
The user cost is the total travel time
Td = Cd + {pax ∙ H} + t we + t wl multiplied by the value of time for each
= T + t we + t wl(7) Where: income bracket of the respective mode:
Vf : Final velocity (km/h)
Tm = te + tl(8) a : Acceleration/deceleration rate (m/s2) UC = TT ∙ VOT ∙ pax(10)

Where:
T : Minibus-taxi dwell time at stop
Cd : Time for opening and closing doors
Ts : Service time at the stop
Td : Dwell time in and/or out of the stop
Tm : Time in which minibus-taxis move
in and out of the stop
t we : Time in which minibus-taxis wait to
enter the stop
t wl : Time in which minibus-taxis wait to
leave the stop
te : Time in which minibus-taxis enter
the stop
tl : Time in which minibus-taxis leave
the stop
The time spent, in hours, during accelera-
tion and deceleration was calculated using
the following equation:
For the queue-jumping lane and the single-
lane pre-signal strategy, the minibus-taxis
pick up and drop off passengers during the
red phase of the traffic cycle. The waiting
time during the red phase is therefore
given by the average delay equations
discussed above.
Finally, the travel time along the single
1-km corridor consists of the distance of
the corridor divided by the speed.
To determine the monetary value of
the user cost it is necessary to have a value
of time to apply to each of the three main
income groups: low, medium and high.
Estimates of the value of travel time savings
(VTTS) in South Africa vary; we decided to
use typical values compiled by Hayes and
Venter (2016) (Table 4). The percentage of
each income group that makes use of cars
Operator cost
The operator cost consists of all the costs
incurred whilst operating a vehicle. For
the private vehicle, the running cost and
maintenance cost were obtained from the
Automobile Association of South Africa;
these amounted to R3.74/km and R0.40/‌k m
respectively (Automobile Association
2013). For the minibus-taxi, little data is
available on operator costs. We used typi-
cal values obtained from the Department
of Transport’s minibus-taxi operating cost
model (Department of Transport 2008),
adjusted for inflation using a rate of 4.5%.
Table 5 summarises all the input vari-
ables used in calculating the operating cost
and briefly describes each.
The operator costs for minibus-taxis
consist of the fuel cost, and the vehicle-time,

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 61
Table 6 I nput variables used in calculating Table 7 Total intersection capacity (veh/hr per direction)
construction cost East-West East-West North-South
Intersection intervention
Variable (unit) Value (mixed traffic) (minibus-taxis) (mixed traffic)

Cost of way (Rm/lane-km) 1.970 Kerbside taxi stop intersection 1 420 329* 580

Land cost – CBD/Commercial Queue-jumping lane and the single-lane

1.649 1 420 347* 450
(Rm/lane-km) pre‑signal strategy intersections

Land cost – Outer section Dedicated taxi lane intersection 2 840 1 420* 580
0.434
(Rm/lane-km) * Values are a fraction of the total mixed traffic values
Minimum cost of station/stop (Rm) 0.4

Life of terminals (years) 20 Table 8 M

 onthly savings per minibus-taxi with each infrastructure form compared to the kerbside
stop (5-km route with priority intersections at 500 m spacings)

vehicle-distance, and vehicle-fleet costs, as Hourly Operating Minimum Maximum

Infrastructure operating cost monthly monthly
given by the following equation: cost savings/taxi savings/taxi savings/taxi

OCt = ⎫⎪
VHc ⎫ ⎫ VDc ⎫ ⎫VFc ⎫
∙ h⎪ + ⎪ ∙ x⎪ + ⎪ ∙ h⎪ +
Kerbside taxi stop R133 – – –

⎭ VH ⎭ ⎭ VD ⎭ ⎭ VH ⎭
Fc ∙ ⎫⎪ i + ( fct ∙ x)⎫⎪(11)
Queue-jumping lane R105 R28 R1 232 R4 928
ft
⎭ hi ⎭
Single-lane pre-signal strategy R108 R25 R1 100 R4 400

Dedicated taxi lane R82 R51 R2 244 R8 976

Where:
VHc = Vs + Vm ∙ 0.5
VDc = Vt + Vm ∙ 0.5
VFc = Vf + Va + Vc(14)
Where:
VHc : Vehicle-time cost, the total time-
dependent cost for a minibus-taxi
h : Time spent, in hours, to travel
along the corridor
VDc : Vehicle-distance cost, the total
distance-dependent cost for a
minibus-taxi
x : Length of the corridor, measured in
kilometres
VFc : Vehicle-fleet cost, the total fleet-
dependent cost for a minibus-taxi;
a fleet-size of 1 was considered as
the operation costs of a minibus-
taxi were given per single vehicle
f it : Idling component of fuel consump-
tion for minibus-taxis
fct : Idling component of fuel consump-
tion for cars
hi : Time spent, in hours, due to idling
A 50/50-split was assumed when appor-
tioning vehicle maintenance cost according
to vehicle-hours and vehicle-distance,
as both variables affect the maintenance
costs. Vehicle fuel costs were calculated in
the same manner for both types of vehicles.
Construction cost
The construction costs were used to
determine the capital costs of each form of
infrastucture. The unit values listed by Del
(12) Mistro and Aucamp (2000), and adjusted for
inflation, are summarised in Table 6. They
(13) apply to all infrastructure types except the
single-lane pre-signal strategy which has no
construction costs, as an existing section of
road would be utilised for its purpose.
MODEL OUTPUTS
Intersection capacity
The capacities of the intersection for each
intervention are summarised in Table 7.
The capacity of the main corridor (east-
west) is significantly higher than that of
the cross street (north-south), in line with
the model assumptions. The dedicated taxi
lane allows for a greater traffic flow for
both mixed traffic and minibus taxis.
To give a sense of the potential cumula-
tive benefit of the operating cost savings to
minibus-taxi operators, the savings were
estimated for a notional 5-km route with
priority intersections spaced at 500-m inter-
vals. Considering a minibus-taxi operator
working 8 hours a day for 22 days in a month
(thus 176 hours per month), an upper limit to
the savings is obtained. If it is assumed that
the benefits accrue only during the morning
and evening peak hours (thus 44 hours per
month), a lower limit is obtained (Table 8).
The estimates show that a notional
minibus-taxi operator may save between
R1 100 and R9 000 when using the priority
infrastructure on a single idealised route
over the course of a month. These translate
into potential savings of between 19% and
38% of taxi operating costs. This makes a
strong case for the implementation of these
infrastructure forms on busy corridors, as a
way of delivering cost savings to operators.
If these savings are passed on to passengers
through fare reductions, passengers would
also reap monetary benefits. An additional
benefit to operators is that of higher vehicle
productivity due to shorter cycles. These
benefits can translate into higher revenue
(assuming there is an unserved passenger
demand), or lower fleet sizes.
Total cost
The total cost takes the user costs, operating
costs and construction costs into account.
The construction cost is only applied to the
minibus-taxis. There is a reduction of up to
30% in total cost per one-way taxi trip when
the kerbside taxi stop is compared to the
priority infrastructure forms. The dedicated
taxi lane has the lowest cost per trip at
R32.78, followed by the queue-jumping lane
at R40.81. The cost per trip for a private
vehicle amounts to R7.09, which is signifi-
cantly less costly than the minibus-taxi.
Figure 15 (p 64) shows the total costs
expressed on a per-passenger basis. As
expected, due to their higher occupancy,
minibus-taxis transport passengers at
significantly lower average cost to society
than private cars. More importantly, the
overall costs for the priority infrastructure
cases are between 12% and 30% lower
than for the base case, indicating that the
estimated additional infrastructure costs of
constructing priority facilities at intersec-
tions are more than off-set by savings in

62 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
 for the queue-jumping lane and single-lane
7 pre-signal strategy (a 9% reduction in travel
6 time) and 1.8 minutes for the dedicated taxi
lane (amounting to a 32% reduction). The
Time (minutes)

5
4 time savings for the queue-jumping lane and
single-lane pre-signal strategy are attribut-
3
able to the priority green phase that reduces
2
minibus-taxi queuing time, as well as the
1
use of the red time for passenger boarding
0 and alighting whereas the dedicated taxi
Kerbside Queue-jumping Single lane Dedicated
taxi stop lane pre‑signal strategy taxi lane lane’s time savings are due to a reduction in
Flow strategy queue lengths causing less congestion.
High-flow taxis High-flow cars Low-flow taxis Low-flow cars Private vehicles do not experience an
increase in travel time when moving from
Figure 12 T ravel time comparison between minibus-taxis and private vehicles the kerbside stop to any of the public trans-
port infrastructure forms. This is due to the
priority green time afforded to the minibus
R1.80 taxis being taken from the undersaturated
R1.60 opposite travel direction (i.e. north-south).
Cost per passenger-trip

R1.40
R1.20
R1.00
Cost outputs
R0.80 The main outputs of this study relate to
R0.60 costs, and include user cost, operating
R0.40 cost, and total cost per passenger-trip.
R0.20 Construction cost was not shown as a cost,
R0.00 because, when reduced to a passenger-trip
Kerbside Queue-jumping Single lane Dedicated
taxi stop lane pre‑signal strategy taxi lane cost, it was not very significant.
Flow strategy
High-flow taxis High-flow cars Low-flow taxis Low-flow cars User cost
The hourly user cost results are expressed
Figure 13 U
 ser cost per passenger per trip comparison between minibus-taxis and private vehicles on a per passenger-trip basis by dividing the
total hourly user cost by the number of traf-
fic arrivals per hour and the vehicle occu-
R4.00 pancy. Figure 13 illustrates these results.
R3.50 A few observations are pertinent. Firstly,
Cost per passenger-trip

R3.00 user costs rise for high-flow cases compared

R2.50 to low-flow cases, due to the extra queuing
R2.00 delay at the intersection. Secondly, only
R1.50 for the dedicated taxi lane, minibus-taxi
R1.00 user costs are lower than those of private
R0.50 vehicle users (by R0.05 per passenger-trip).
R0.00 The difference in minibus taxi and car per
Kerbside Queue-jumping Single lane Dedicated
taxi stop lane pre‑signal strategy taxi lane person-trip cost, however, becomes more
Flow strategy significant when the priority interventions
High-flow taxis High-flow cars Low-flow taxis Low-flow cars are implemented, differing by R0.26 and
R0.30 for the queue-jumping lane and
Figure 14 O
 perating cost per passenger-trip comparison between minibus-taxis and private vehicles single-lane pre-signal strategy respectively.
Thirdly, car user costs hardly change when
operating costs and travel time for taxi deceleration time by either minibus-taxi or implementing priority features for public
passengers, without significantly raising private vehicle, is illustrated in Figure 12. transport, in line with the study objectives.
costs for private vehicles. Once again, a As expected, for all treatments, travel Lastly, and most importantly, taxi user costs
dedicated taxi lane has the lowest overall times for the low-flow case are always lower decline significantly (between 11% and 32%)
cost due to its significant time savings. than for the high-flow case due to lower with the priority treatments, reflecting the
queue delay. For the base (current) case, as delay saving accruing to taxi passengers.
Travel time well as all three forms of priority interven-
Travel time is a primary component of user tions, taxis experience more delay than Operator cost
costs. The travel time for travelling along a cars due to the assumed far-side stop The operating cost per passenger-trip for
notional 1-km corridor with one intersec- after clearing the intersection. All three minibus-taxis and private vehicles is illus-
tion that includes the stopping time at the interventions, however, see a reduction in trated in Figure 14. Per-person car costs
intersection, as well as acceleration and travel time varying between 0.5 minutes are much higher than those of a taxi trip,

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 63
largely due to the lower occupancy of the

Cost per one-way passenger-trip

private car.
The minibus-taxi operating cost sees
a 29% decrease when the kerbside stop is
compared to the dedicated taxi lane, and a
15% and 14% decrease when it is compared to
the queue-jumping lane and single-lane pre-
signal strategy, respectively. This is largely
driven by the reduction in travel time. The
per person-trip costs of cars remain the same
due to the travel times that do not vary.
Sensitivity analysis
A sensitivity analysis was carried out to
check the robustness of the analysis against
variations in key input variables. These
variables included the length of the cor-
ridor, the ratio of minibus-taxi occupancy
to private vehicle occupancy, passenger
handling time for minibus-taxis, percentage
of minibus-taxis stopping to pick up or drop
off passengers, and the minibus-taxi vehicle
hours travelled in a month. The results from
the analysis are summarised in Table 9.
The values in the table indicate the change
when the base input value is compared to
the upper limit value using total cost per
passenger-trip as the value being compared.
Corridor length (while keeping the num-
ber of priority intersections constant), as
well as the percentage of minibus taxis stop-
ping to pick up or drop off passengers, had
the largest impact on the output. In the case
of corridor length, it implies a longer travel
distance between priority intersections.
Longer corridors reduced the comparative
advantage of the queue-jumping lane and
single-lane pre-signal strategy, but most
significantly, the dedicated taxi lane, as their
time savings become less significant relative
to total operating costs. In the case of per-
centage minibus taxis stopping: the greater
the percentage minibus taxis required to
stop, the less beneficial the priority intersec-
tions become, and a lower net benefit is
provided to the operators. The results are
thus consistent with the outputs delivered
by the model and do not cause the relative
ranking of the treatments to change.
R6.00
R5.00
R4.00
R3.00
R2.00
R1.00
R0.00
Kerbside
taxi stop
High-flow taxis
Figure 15 T otal cost per passenger per trip comparison between minibus-taxis and private vehicles
scarce road space may do much to promote
public transport, especially the minibus-
taxi which is the primary form of public
transport in many developing countries.
Judicious investment in priority infrastruc-
ture may be a very cost-effective way to
raise the overall efficiency of the transport
system, without significantly deteriorating
conditions for other road users such as pri-
vate car drivers and freight operators. Such
investments will provide what amounts to
an implicit operational subsidy to the para-
transit sector since they reduce operator
costs and delay. If some of these cost sav-
ings are passed on to passengers, they will
benefit not only through faster and more
reliable travel times, but also lower fares.
We developed simple analytical models
to estimate the net economic impacts on taxi
operators, passengers and private car users of
three alternative priority measures ‒ queue
jumping lanes, single-lane pre-signal strate-
gies, and dedicated taxi lanes. The models
take typical driving and traffic conditions
into account. The analysis was limited to
undersaturated corridors with medium traf-
fic volumes, on the assumption that higher
Table 9 Sensitivity analysis outputs
Variable (varied)
Kerbside
taxi stop
Queue-jumping Single lane Dedicated
lane pre‑signal strategy taxi lane
Flow strategy
High-flow cars Low-flow taxis Low-flow cars
volume corridors may be more suitable for
larger interventions involving bus lanes and
larger vehicles. The models may be useful to
examine priority interventions in real-world
cases, before detailed design or microsimula-
tion efforts are undertaken.
The models were applied to a notional
corridor to estimate the quantum of impacts
under low and medium volume scenarios.
The results showed that, compared to the
kerbside taxi stop, priority interventions
can reduce travel time by between 9% and
32%, and overall user cost (including fuel
consumption) by between 11% and 32%.
Over the course of a month, operators may
save between R1 100 and R9 000 in direct
operating costs, depending on fluctuations
in delay conditions over the course of the
day. Faster travel times may also reduce
route cycle times and raise the vehicle
productivity of minibus-taxi fleet operators.
Taking construction costs into account, the
net benefits are still positive, with the dedi-
cated taxi lane outperforming the other two
priority interventions due to shorter queues
at intersections leading to shorter waiting
times. Comparing the queue-jumping lane
Infrastructure form
Queue- Single-lane
Dedicated
jumping pre-signal
taxi lane
lane strategy

Corridor length (1 – 9 km) 4.4 4.8 4.7 6.0

CONCLUSIONS
Ratio of minibus-taxi to private
The kerbside stop is favoured by local 2.6 2.4 2.5 1.9
vehicle occupancy (2:5 to 18:1)
authorities in South Africa as a first step
Passenger handling time
towards regularising taxi operations and 2.0 1.9 2.0 1.5
(2 – 12 sec)
reducing delay to other vehicles. However,
Percentage of minibus-taxis
the net benefits can be substantially 3.2 7.3 6.8 1.9
stopping (0 – 100%)
increased by modest additions of dedicated
infrastructure at busy intersections. The Minibus-taxi vehicle hours
0.43 0.42 0.42 0.44
(40 – 360 hours)
paper contends that such repurposing of

64 Volume 63 Number 4 December 2021 Journal of the South African Institution of Civil Engineering
and the single-lane pre-signal strategy,
given their very similar benefits, the queue-
jumping lane appears to be more promising
as it is easier to implement and avoids many
of the potential safety and operational issues
that may arise with counter-flow pre-signal
alternatives. These issues were not studied
in detail and are perhaps worth further
investigation in cases where the space for
queue-jumping lanes does not exist.
This work was exploratory and raises
many further questions regarding the plan-
ning and design of priority facilities for
minibus taxi services in South Africa. The
analysis needs to be extended to more con-
gested and oversaturated traffic conditions
to see where the limits of their economic
feasibility lie. The assumptions regarding
low cross-traffic volumes need to be relaxed
to consider more variable delay conditions.
Further analysis is also needed to investigate
more realistic cases of traffic conditions that
fluctuate across the day, and how this might
affect the performance of the alternative
treatments. Much further work is required
on the design of priority treatments, includ-
ing signalisation of queue-jumping lanes,
geometric layout and signage, and passenger
loading/unloading facilities. Behavioural
aspects such as the stopping behaviour of
minibus taxi drivers need to be added, espe-
cially since the traffic observations presented
at the beginning of the paper showed that
drivers already habitually display adaptations
of priority lane driving behaviour, albeit
illegal; behavioural shifts and better traffic
law enforcement might need to be achieved
to make these strategies work well. Lastly, we
did not consider priority treatments involving
purely signal priority and pre-emption strate-
gies; these might also be worth investigation.
ACKNOWLEDGEMENTS
The authors would like to thank the Volvo
Research and Educational Foundation
(VREF) via the BRT Centre of Excellence
for funding and support.
REFERENCES
Automobile Association 2013. Calculating the
operating cost of a vehicle. Automobile Association
of South Africa. https://www.scribd.com/
document/128854202/Calculating-the-Operating-
Cost-of-a-Vehicle-The-AA-of-South-Africa.
Automobile Association 2019. Fuel pricing. https://
www.aa.co.za/fuel-pricing [Accessed 19 June 2019].
Ben-Dor, G, Ben-Elia, E & Benenson, I 2018. Assessing
the impacts of dedicated bus lanes on urban traffic
congestion and modal split with an agent-based
model. Procedia Computer Science, 130: 824‒829.
doi: 10.1016/j.procs.2018.04.071.
Bhattacharyya, K, Maitra, B & Boltze, M 2019.
Implementation of bus priority with queue
jump lane and pre-signal at urban intersections
with mixed traffic operations: Lessons learned?
Transportation Research Record, 2673(3): 646‒657.
Bian, B, Zhu, N, Ling, S & Ma, S 2015. Bus service
time estimation model for a curbside bus stop.
Transportation Research, Part C. Emerging
Technologies, 57: 103‒121.
Brasuell, J 2019. Dedicated bus lanes beating
congestion in Portland. https://www.planetizen.
com/news/2019/12/107482-dedicated-bus-
lanes-beating-congestion-portland [Accessed
12 February 2020].
Del Mistro, R F & Aucamp, C A 2000. Development
of a public transport cost model. Proceedings, 19th
Annual Southern African Transport Conference,
17–20 July 2000, Pretoria.
Department of Transport 2008. Taxi recapitalization
viability model. Pretoria: Department of Transport.
Department of Transport/Statistics South Africa 2013.
National Household Travel Survey 2013. Statistical
Release P0320. Pretoria: Statistics South Africa.
Glambrone, A & Acitelli, T 2019. Dedicated bus
lanes on H and I streets NW aim to speed up rush-
hour service, starting June 3. https://dc.curbed.
com/2019/5/3/18528013/dedicated-bus-lanes-
washington-dc [Accessed 12 February 2020].
Govender, R & Allopi, D 2007. Analysis of the
scientific aspects related to minibus taxi collisions.
Proceedings, 26th Southern African Transport
Conference, 9–12 July 2007, Pretoria.
Gu, W, Cassidy, M J & Li, Y 2012. Models of bus
queueing at curbside stops. Berkeley, CA: University
of California at Berkeley, Center for Future Urban
Transport.
Hager, K, Rauh, J & Rid, W 2015. Agent-based
modeling of traffic behavior in growing
metropolitan areas. Transportation Research
Procedia, 10: 306‒315.
Halifax Regional Municipality (Canada) 2018. Transit
priority measures. https://www.halifax.ca/
transportation/halifax-transit/moving-forward-
together/part-7-transit-priority-measures
[Accessed 30 October 2019].
Hayes, G & Venter, C 2016. Trip utility and the value
of travel time savings (VTTS) for commuter trips:
Critique and recent advances. Proceedings, 35th
Southern African Transport Conference, 4–7 July
2016, Pretoria.
Ilgin Guler, S, Gayah, V V & Menendez, M 2015.
Providing bus priority at signalized intersections
with single-lane approaches. Transportation
Research Procedia, 9: 225‒245.
Jennings, G & Behrens, R 2017. The case for investing
in paratransits. Cape Town: Volvo Research &
Educational Foundations.
Neumann, A, Röder, D & Joubert, J W 2015. Toward a
simulation of minibuses in South Africa. Journal of
Transport and Land Use, 8(1): 137‒154.
Salazar Ferro, P, Behrens, R & Wilkinson, P 2012.
Hybrid urban transport systems in developing
countries: Portents and prospects. Research in
Transport Economics, 39: 121‒132.
Schalekamp, H & Behrens, R 2010. Engaging paratransit
on public transport reform initiatives in South
Africa: A critique of policy and an investigation of
appropriate engagement approaches. Research in
Transportation Economics, 29(1): 371‒378.
Schalekamp, H & Behrens, R 2013. Engaging the
paratransit sector in Cape Town on public transport
reform: Progress, process and risks. Research in
Transportation Economics, 39(1): 185‒190.
Schalekamp, H & Klopp, J M 2018. Beyond BRT:
Innovations in minibus-taxi reform in South
African cities. Proceedings, 37th Southern African
Transport Conference, July 2018, Pretoria.
Stamos, I, Kitis, G, Basbas, S & Tzevelekis, I 2012.
Evaluation of a high occupancy vehicle lane in
central business district Thessaloniki. Social and
Behavioral Sciences, 48: 1088‒1096
Tirachini, A 2014. The economics and engineering
of bus stops: Spacing, design and congestion.
Transportation Research Part A, 59: 37‒57.
Transportation Research Board 2013. Transit Capacity
and Quality of Service Manual, 3rd ed. Washington,
DC: National Academy of Sciences.
Zheng, Z, Rasouli, S & Timmermans, H 2020.
Modeling taxi driver search behavior under
uncertainty. Travel Behaviour and Society,
22: 207‒2018.
Zhou, G & Gan, A 2009. Design of transit signal
priority at signalized intersections with queue
jumper lanes. Journal of Public Transportation,
12(4): 117‒132.
Zlatkovic, M, Stevanovic, A & Reza, Z 2013. Effects of
queue jumpers and transit signal priority on Bus
Rapid Transit. TRB Paper 13-0483. Washington,
DC: Transportation Research Board.

Journal of the South African Institution of Civil Engineering Volume 63 Number 4 December 2021 65
Guidelines for the preparation of papers and technical notes
Authors should comply with the following guidelines when preparing papers for publication in the journal
The Journal of the South African Institution of Civil
Engineering is published quarterly in March, June,
September and December. Articles submitted for
publication are reviewed by a panel of referees under
the guidance of the SAICE Journal Editorial Panel.
The journal publishes research papers covering all the
disciplines of civil engineering (structural, geotechnical,
railway, coastal/marine, water, construction, environ-
mental, municipal, transportation) and associated topics
that are relevant to the civil engineering profession, and
that preferably have particular relevance to civil engi-
neering in southern Africa and the African continent.
When preparing articles for publication, authors
should please take note of the following and comply
with the guidelines as set out:
CLASSIFICATION OF ARTICLES
CONSIDERED FOR PUBLICATION
QQ Technical papers are well-researched, in-depth,
fully referenced technical articles not exceeding
6 000 words in length (excluding tables, illustra-
tions and the list of references). Related papers that
deal with ‘softer sciences’ (e.g. education, social
upliftment, etc) are accepted if they are of a techni-
cal nature and of particular interest to the civil
engineering profession. The latter type of paper will
be subject not only to peer-review by civil engineers,
but also to review by non-engineering specialists in
the field covered by the paper.
QQ Technical notes are short, fully referenced technical
articles that do not exceed 2 000 words. A typical
technical note will have limited scope often dealing
with a single technical issue of particular importance
to civil engineering.
QQ Review papers are considered for publication as
either technical papers or technical notes on condi-
tion that they are the original work of the author
and will assist the reader with the understanding,
interpreting or applying of the subject under review.
A review paper must contain criteria by which the
work under review was evaluated, and contribute
by synthesising the information and drawing new
conclusions from the dissemination of the previously
published work.
QQ Discussion on published articles is welcomed up to
six months after publication. The length of discus-
sion contributions is limited to 1 500 words. Where
appropriate, discussion contributions will be subject
to the normal reviewing process and will be for-
warded to the authors of the original article for reply.
POLICY REGARDING LANGUAGE AND
ORIGINALITY OF SUBMITTED ARTICLES
QQ Language: Manuscripts should preferably be
presented in English, as the journal is distributed
internationally. Articles submitted in any of the
other official South African languages should be
accompanied by an expanded abstract in English.
QQ Original work: Papers and technical notes must
be original contributions. Authors must affirm
that submitted material has not been published
previously, is not under consideration for publica-
tion elsewhere and will not be submitted elsewhere
while under consideration by the SAICE Journal
Editorial Panel. It is the responsibility of the authors
to ensure that publication of any paper in the journal
will not constitute a breach of any agreement or
the transgression of any law. The corresponding
author should confirm that all co-authors have
read and approved the manuscript and accept these
conditions. Authors are responsible for obtaining
permission to publish experimental data and other
information that may be confidential or sensitive.
Authors are also responsible for obtaining permis-
sion from copyright owners when reproducing mate-
rial that had been published elsewhere. Proof of such
permission must be supplied.
SUBMISSION PROCEDURES
AND REQUIRED FORMAT
QQ Online submission: Manuscripts must be uploaded
as PDF files (http://journal.saice.org.za). Individual
file sizes may not exceed 10 MB. Should you experi-
ence problems uploading your paper, please contact
the editor (verelene@saice.org.za).
QQ Format: Manuscripts should be prepared in
MS Word and presented in double-line spacing,
single-column layout with 25 mm wide margins.
Line numbers must be applied to the whole docu-
ment. All pages should bear the authors’ names and
be numbered at the bottom of the page. With the
exception of tables and figures (see below) the docu-
ment should be typed in Times New Roman 12 pt
font. Contributions should be accompanied by an
abstract of not more than 200 words.
QQ First page: The first page of the manuscript should
include the title of the paper, the number of words of
the main text (i.e. excluding figures, tables and the
list of references), the initials and surnames of the
authors, professional status (if applicable), SAICE
affiliation (Member, Fellow, Visitor, etc), telephone
numbers (landline and mobile), and e-mail and
postal addresses. The name of the corresponding
author should be underlined. Five keywords should
be suggested.
QQ Figures, tables, photos and illustrations: These
should preferably be submitted in colour, as the
journal is a full-colour publication.
QQ Their positions should be clearly marked in the
text as follows: [Insert Figure 1].
QQ Figures, tables, photos, illustrations and equa-
tions should be numbered consecutively and
should appear in the text directly after they have
been referred to for the first time.
QQ Illustrations must be accompanied by appropriate
captions. Captions for tables should appear above
the table. All other captions should appear below
the illustration (figures, graphs, photos).
QQ Only those figures and photographs essential to
the understanding of the text should be included.
All illustrations should be referred to in the text.
QQ Figures should be produced using com-
puter graphics. Hand-drafted figures will not be
­accepted. Lettering on figures should be equiva-
lent to a Times New Roman 9 pt font or slightly
larger (up to 12 pt) if desired. Lettering smaller
than 9 pt is not acceptable.
QQ Tables should be typed in Times New Roman
9 pt font. They should not duplicate information
already given in the text, nor contain material
that would be better presented graphically.
Tabular matter should be as simple as possible,
with brief column headings and a minimum
number of columns.
QQ Mathematical expressions and presentation
of symbols:
QQ Equations should be presented in a clear form
which can easily be read by non-mathematicians.
Each equation should appear on a separate line
and should be numbered consecutively.
QQ Symbols should preferably reflect those used in
Microsoft Word Equation Editor or Mathtype,
or should be typed using the Times New Roman
symbol set.
QQ Variables in equations (x, y, z, etc, as well as
lower case Greek letters) should be presented
in italics. Numbers (digits), upper case Greek
letters, symbols of metric measurement units
(m for metres, s for seconds, etc) and mathemati-
cal/trigonometrical functions (such as sin, cos
and tan) are not written in italics, but in upright
type (Roman). Variables and symbols used in the
body of the text should match the format used
in the equations, i.e. upright or italics, whichever
is applicable.
QQ Metric measurement abbreviations/units should
conform to international usage – the SI system of
units should be used.
QQ Decimal commas may be used, but decimal
points are preferred.
QQ Symbols should preferably be defined in the text,
but if this is not feasible, a list of notations may be
provided for inclusion at the end of the paper.
QQ Headings: Sections and paragraphs should not be
numbered. The following hierarchy of headings
should be followed:
HEADING OF MAIN SECTION
Heading of subsection
Heading of sub-subsection
QQ References: References should follow the Harvard
system. The format of text citations should be as
follows: “Jones (1999) discovered that …” or “recent
results (Brown & Carter 1985; Green et al 1999)
indicated that …”
 References cited in the text should be listed in
alphabetical order at the end of the paper. References
by the same author should be in chronological order.
The following are examples of a journal article, a
book and a conference paper:
QQ Donald, A W 1954. Pore pressure coefficients.
Géotechnique, 4(4):143–147.
QQ Terzaghi, K 1943. Practical soil mechanics. New
York: Wiley.
QQ Smith, A W, Black, C & White, A 1999. Factors
affecting the strength of soils. Proceedings, ASCE
Conference on Shear Strength of Soils, Colorado,
pp 503–532.
Papers published previously in the Journal of the
South African Institution of Civil Engineering should
be cited if applicable.
QQ Footnotes, trade names, acronyms, abbreviations:
These should be avoided. If acronyms are used, they
should be defined when they first appear in the text.
Do not use full stops after abbreviations or acronyms.
QQ Return of amended papers: Papers requiring
amendments will be accepted up to six months after
the referee reports had been sent to authors, after
which the paper will be withdrawn from the system.
FINAL ARTICLE
QQ Copyright: On acceptance of the paper or technical
note, copyright must be transferred by the author/s
to the South African Institution of Civil Engineering
on the form that will be provided by the Institution.
QQ Photos of authors: The final corrected version of
the paper should be accompanied by recent, high-
resolution head and shoulders colour photographs
and a profile not exceeding 100 words for each of
the authors.
QQ Proofs: First proofs of papers will be sent to authors
in PDF format for verification before publication.
No major re-writes will be allowed, only essential
minor corrections.