An evaluation of the primary South African standard and guideline

for the provision of water for firefighting

CB Mac Bean1 and AA Ilemobade1*

1
School of Civil and Environmental Engineering, University of the Witwatersrand, Johannesburg, Private Bag 3, WITS 2050, South Africa

ABSTRACT
In South Africa, as is mostly the norm globally, national legislation and guidelines specify that potable water distribution
networks maintain the capacity to provide specified quantities of water for firefighting. This paper addresses the question: is
the South African standard and guideline pertaining to fire-flow provision appropriate for firefighting and do these ensure
the most efficient balance between providing sufficient fire protection and promoting sustainable water use? In answering
this question, this study: (i) reviewed national and international design standards and guidelines; and (ii) captured and
analysed 10 years of billable fire incident reports representing 3 859 fire events within the City of Johannesburg. Highlights
from the study include: inconsistencies in categories when comparing the SANS 10090 and The Red Book fire tables and
violations (in The Red Book) of stipulated Minimum Fire Flows; over the 10 year period, 75% of fire incidents within the City
of Johannesburg were extinguished using less than 6.6 kL of water – less than the capacity (6.9 kL) of the City’s conventional
pumping tanker during the period; 99.9% of fire incidents within the City were quenched using an average fire flow rate of
less than 1 200 L/min, which is the minimum hydrant flow rate for the lowest fire risk category in SANS 10090; and peak
fire occurrence did not correspond with typical peak residential water use. Recommendations are proffered in respect of the
above.

Keywords: firefighting water standard and guideline

INTRODUCTION by some, to be historically based on instinct and was strongly

characterised by what was available rather than a technical
Water conservation has become a priority for many water- analysis of what was needed (Law and Beever, 1995; Davis, 2000).
scarce countries around the world, including South Africa. The condition of infrastructure, development of firefighting
With the majority of potable water supply to the public being technologies and techniques, and the growth in fire safety
delivered via municipal water systems, it is critical that these awareness have all progressed with time and evolved
systems be designed, constructed and maintained in such dramatically since 1966 when the national codes for the
a manner that they promote the most efficient use of water. provision of water for firefighting in South Africa were first
Inseparable from the topic of efficient potable water use, is the published. Therefore, Van Zyl and Haarhoff (1997) recommend
efficient or inefficient provision of potable water for firefighting that the provision and requirements for fire flows be amended
and the impact that this provision has on water distribution to reflect present conditions and technologies.
systems (WDSs) and water conservation.
A global consensus on the ideal design philosophy for Objectives
providing water for firefighting remains to be established.
Likewise, civil infrastructure standards linked to firefighting This study addresses two objectives:
remain diverse and widely debated, as engineers face the • To present an analysis of international and South African
challenge of balancing firefighting water provision against design standards and guidelines pertaining to water
efficient water use. It is due to this complex trade-off between provision for firefighting
ensuring adequate supply of water for firefighting and • To present actual fire flow data recorded in the City of
minimizing water use that further research into this topic is Johannesburg, to compare this data with the primary
critically needed. South African standard and guideline values, and to make
In South Africa, the national standard (SANS 10090; recommendations to guide future revisions of the primary
SABS, 2003) and guideline (DHS and CSIR, 2019) recommend South African design standard and guideline for the
that potable water distribution mains have and maintain provision of water for firefighting
the capacity, both in flow (L/min) and pressure head (m), to
provide specified quantities of water for firefighting purposes. South African standards and guidelines for fire flows
As a result, a dominant design constraint on WDSs is
providing for fire flow, which is defined as the rate of flow of Guidelines are intended to assist decision-making, whereas
water required by the firefighting service for the extinguishing standards are enforceable absolute limits (Schlotfeldt, 1995
of fires (SABS, 2003). cited in CSIR, 2005). The national standard (also termed ‘code’)
Since fire flows significantly influence the sizing of a for the provision of water for firefighting in South Africa titled
reticulated network, it is important that these requirements are ‘Community protection against fire’ was first published in 1966
defined as accurately as possible. It is interesting to note that (SABS 090) and revised in 1972 (SABS, 1972). This code was
the demand for water supplies during firefighting is believed, compiled with the assistance of organizations from the UK, USA,
Canada, New Zealand and Germany (Van Zyl and Haarhoff,
*Corresponding author, email: adesola.ilemobade@wits.ac.za 1997). A notable feature of this code is the absence of minimum
Received 23 August 2018; accepted in revised form 26 September 2019 water pressures for both water provision and the pumping ability

https://doi.org/10.17159/wsa/2019.v45.i4.7551
Available at https://www.watersa.net
ISSN 1816-7950 (Online) = Water SA Vol. 45 No. 4 October 2019
Published under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0) 691
of response units during a firefighting event. A summary of fire SANS 10090, however, presents two categories. The first is
flow values within the most recent edition of this standard (i.e. titled, ‘Fire Risk Category’, and the second, which is a subset
SANS 10090: SABS, 2003) is shown in Table 1. of the first, is titled, ‘Possible Fire Sizes’. The ‘Possible Fire
A separate national code, SANS 10252-1:2016 (SABS, 2016) Sizes’ category is used exclusively to determine ‘Minimum Fire
titled ‘Water supply and drainage for buildings Part 1: Water Flow’ rates. An adverse consequence of having two categories
supply installations for buildings’ addresses design pressures in the SANS 10090 is that the Minimum Fire Flow and the
and flows for fire installations. A minimum provision of 30 L/ Minimum Hydrant Flow are determined from different Fire
min per fire hose reel and 1 200 L/min per hydrant is stipulated Risk Categories, despite the fact that both are within the same
without any reference to fire risk categories. This code neither table and connected. By way of example, an affluent residential
refers to SANS 10090 nor provides as much detail as it does. area (Category C) where houses are spaced further than 30 m
However, SANS 10252-1 stipulates that a minimum pressure of apart (Category D1) would have SANS 10090 recommend two
300 kPa must be maintained in hoses and hydrants. Minimum Hydrant Flows of 2 000 L/min (Category C) and
Another industry-recognised reference that provides 1 200 L/min (Category D1) and two Minimum Fire Flows of
guidance on firefighting is the recently released Red Book 6 000 L/min (Category C) and 1 900 L/min (Category D1).
(DHS and CSIR, 2019) titled The Neighbourhood Planning In addition to the above matter, some violations arise when
and Design Guide (Red Book): Creating Sustainable Human employing Minimum Fire Flow values from The Red Book.
Settlement. The Red Book (DHS and CSIR, 2019) is an updated The Red Book, which is a guideline that is intended to assist
version of the CSIR (2005) Guidelines for Human Settlement decision-making, should not, without reasonable justification,
Planning and Design. In contrast to the CSIR (2005) firefighting violate standards (in this instance, SANS 10090), which present
guidelines, The Red Book (DHS and CSIR, 2019) references the enforceable absolute minimum limits (Schlotfeldt, 1995 cited
SABS 10090 (2003) code. The Red Book’s fire flow values are in CSIR, 2005). All fire flow values in The Red Book are less
presented in Table 2. than the values stipulated in SANS 10090 for similar fire risk
A notable distinction between SANS 10090 (SABS, categories. An example of this violation is seen in the first
2003) and The Red Book (DHS and CSIR, 2019), apart from two fire risk categories in both documents – The Red Book
the different values they stipulate, is their differing fire risk recommends a fire flow of 6 000 L/min for the ‘high risk’
categories. The Red Book presents a single set of fire risk category and 3 000 L/min for the ‘moderate risk 1’ category;
categories to which all its various recommendations are related. SANS 10090, on the other hand, stipulates 13 000 L/min and

Table 1. SANS 10090 fire flow summary (SABS, 2003)

Fire risk category
A B C D E
Central business Limited central business Residential areas Rural areas of limited Special risk areas. Individual
districts and extensive districts, smaller commercial of conventional buildings and remote from areas requiring a pre-
commercial and or industrial areas normally construction. urban areas. determined attendance over
industrial areas associated with small towns D1: Houses > 30 m apart and above the predominant
normally found in cities and decentralized areas D2: Houses 10.1 m to 30 m risk category in an area.
and large towns (areas of cities and large towns apart Includes large shopping/
where the risk to life (areas where the risk to life D3: Houses 3 m to 10 m entertainment centres,
and property due to fire and property due to fire apart informal settlements,
occurrence and spread occurrence and spread is D4: Houses < 3 m apart harbours, hospitals, prisons,
is likely to be high). likely to be moderate). large airport buildings and
petrochemical plants.
Provision of water for firefighting
Weight of response at fires Minimum fire flow Fire hydrants
Max.
Fire risk Minimum Minimum Minimum
Minimum distance
manning pumping Possible Flow In hydrant Maintenance
category number of level per capacity of each fire sizes (L/min) flow
between
Intervals
pumping units hydrants
appliance unit (L/min) (L/min)
(m)
A 2 5 3 850 Non-residential buildings 13 000 2 000 85 Annually
with divisions not greater
than 5 000 m²
B 2 4 3 850 Non-residential buildings 9 000 2 000 120 Biennially
having divisions not
greater than 2 500 m²
C 1 4 2 250 Non- residential premises 6 000 2 000 200 Triennially
not greater than 1 250 m²
D1 Houses > 30 m apart 1 900 1 200 300
D2 Houses 10.1 m to 30 m 2 850 1 200 200
1 4 2 250 apart Triennially
D3 Houses 3–10 m apart 3 800 1 400 200
D4 Houses < 3 m apart 5 700 2 000 200
E * As determined by individual risk assessment Annually

https://doi.org/10.17159/wsa/2019.v45.i4.7551
Available at https://www.watersa.net
ISSN 1816-7950 (Online) = Water SA Vol. 45 No. 4 October 2019
Published under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0) 692
Table 2. The Red Book (DHS and CSIR, 2019) fire flow values
Duration of Minimum flow Minimum Minimum
Risk Maximum Total fire flow
design fire at one hydrant pressure at pressure at rest
classification hydrant spacing
flow (h) (L/s) (L/min) (L/s) (L/min) fire node (m) of the system (m)
Red Book: page Red Book: Red Book: Table J.17
J.58, Section (iv) Table J.18
High risk: CBD and high 6 100 6 000 25 1 500 15 5
risk industrial
Moderate risk 1: 4 50 3 000 25 1 500 15 5
Industrial, business, 200 m (or as
high-rise flats ≥ 4 storeys otherwise
Moderate risk 2: Cluster required by 2 25 1 500 25 1 500 10 5
& low-income housing, the local fire
high rise flats ≤ 3 storeys department)

Low risk: Single 1 15 900 15 900 10 5

residential housing

9 000 L/min, respectively, for Categories A and B. This paper (a) Infrastructure/building planning: These methods are
recommends a uniform category for both documents and necessarily predictive in nature, are more complicated and
Minimum Fire Flows that are consistent with analysed data. involve several steps and multiple calculations. Typical
variables accounted for include: building construction,
International standards and guidelines for fire flows occupancy, fire size, heat release and sprinkler contribution.
The inclusion of a variety of variables enables adjustments
Across the world, many methods have been developed to calculate to be made to the building type or protection features (e.g.
fire flows. These methods generally form the basis on which fire adding a sprinkler system) in order to reduce the fire flow.
protection codes, such as those discussed above, are established. (b) On-scene firefighting: These methods, by comparison, are
These methods regulate the design of various WDS features such as: much simpler, allowing fire fighters on the scene to assess
• Spacing of fire hydrants whether they need more hose lines or apparatus to fight
• Minimum size of reticulation pipes the fire. They typically consist of one equation with one
• Minimum flow rates and pressures independent variable – either the volume or area of the fire.
• Storage requirements and flow durations The 16 fire flow calculation methods were simulated
In a report conducted by The Fire Protection Research for two differently sized non-residential buildings and two
Foundation, titled ’Evaluation of fire flow methodologies’, 16 differently sized single-family residential buildings. Their study
fire flow calculation methods were evaluated. The methods included both sprinklered and non-sprinklered calculations.
identified were from the USA, UK, France, Germany, the Figure 1 shows the fire flow requirements for a non-sprinklered,
Netherlands, New Zealand, and Canada. Eleven of the non-residential building of 50 000 ft 2 (4 645 m2).
methods address pre-incident infrastructure/building planning To compare the primary South African fire flow standard
(see (a) below) and five are best suited for on-scene firefighting (SANS 10090:2003, SABS, 2003) and guideline (The Red Book,
(see (b) below) (Benfer and Scheffey, 2014): 2019) with the 16 shown on Fig. 1, the Minimum Fire Flow

Figure 1. Fire flow requirements for a non-sprinklered, non-residential building of 4 645m2 (Benfer and Scheffey, 2014)

https://doi.org/10.17159/wsa/2019.v45.i4.7551
Available at https://www.watersa.net
ISSN 1816-7950 (Online) = Water SA Vol. 45 No. 4 October 2019
Published under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0) 693
requirements for a simlar structure, as defined in SANS 10090 For residential buildings fitted with sprinklers, it is
and The Red Book, are superimposed on the results presented worth noting that 12 out of the 18 Benfer and Scheffey (2014)
in the figure. It is important to note that the SANS 10090 fire methodologies shown on Fig. 2 required the same fire flow as
flow values presented in Figs 1 and 2 do not explicitly deal non-sprinkler fitted buildings. However, as seen in Fig 1 and
with a single incident. However, as expressed in SANS 10090 2, Minimum Fire Flow requirements vary greatly across many
(SABS, 2003) clause 11.4.1: ‘The required fire flow should be countries.
available to the firefighting team on arrival at the fire.’ It is
thus assumed that the comparison made below is fair. Where METHODS
applicable, both the Minimum Fire Flow and Minimum
Hydrant Flow for SANS 10090 and The Red Book are shown in Fire incident reports within the City of Johannesburg
Figs 1 and 2.
Employing the SANS 10090 (2003) ‘Possible Fire Sizes’ A fire incident report (or call slip) is a physical document that is
category ‘Non-residential buildings with divisions not filled out and submitted to the central Emergency Management
greater than 5 000 m² (53 800 ft 2)’, a Minimum Fire Flow of Services (EMS) headquarters after each fire incident attended
13 000 L/min is required (Fig. 1). In the figure, the SANS 10090 to by the fire brigade. Only billable (i.e. incidents that the fire
requirement falls within the ranges of the ISO, IFC/NFPA 1, department charges the property owner for services rendered)
and the IWUIC Building Planning methods but is, for several fire incident reports are digitally captured in spreadsheets
of the other methods, several orders of magnitude lower. The by the EMS. Billable fire incident reports are best explained
Minimum Fire Flow requirement in The Red Book for the by sections 10.1 and 10.2 (Fees) of the Fire Brigade Services
‘high risk’ category (which is a similar category to the SANS Act No. 99 (RSA, 1987). These sections (listed below) outline
10090 category above) is 6 000 L/min. The corresponding the basis on which the local fire department may charge for
Minimum Hydrant Flow for SANS 10090 and The Red Book services rendered:
are 2 000 L/min and 1 500 L/min, respectively. (1) A controlling authority may, subject to any condition
It is seen from Fig. 1 that the range of possible fire flows contemplated in section 11(2)(a), determine the fees payable
is large, not only when comparing the various methods, but by a person on whose behalf the service of the controlling
also within some ranges. The FEDG and PAS 4509 methods authority is applied-
have the largest ranges. Furthermore, as can be seen in Fig. (a) for the attendance of the service;
1, the Building Planning methods tend to recommend fire (b) for the use of the service and equipment; or
flows that are higher than the on-scene methods. many (c) for any material consumed.
of the on-scene firefighting methods do not incorporate (2) A person on whose behalf, in the opinion of the chief fire
sprinkler protection systems in their calculations (Benfer officer concerned, a service of a controlling authority has been
and Scheffey, 2014). employed, may in writing be assessed by that chief fire officer
Figure 2 shows the sprinklered and non-sprinklered fire for the payment of the fees referred to in subsection (1) or any
flow values for a residential building of 3 500 ft 2 (325 m2). portion thereof.
Employing the SANS 10090 Risk Category D1 – ‘Houses It is the above billable incidents that have been consolidated
> 30 m apart’, a Minimum Fire Flow of 1 900 L/min and and presented in this paper. In these reports, details captured
a Minimum Hydrant Flow of 1 200 L/min are required. include the duration of the call out, the quantity of water used,
Employing The Red Book ‘low risk’ category, the Fire Flow and and the appliances used during the incident (see an example
the Minimum Hydrant Flow are each 900 L/min. in Table 3). On-site calculations are carried out to estimate

Figure 2. Fire flow requirements for a 3 500 ft2 (325 m2) single-family home; sprinklered vs. non-sprinklered (Benfer and Scheffey, 2014)

https://doi.org/10.17159/wsa/2019.v45.i4.7551
Available at https://www.watersa.net
ISSN 1816-7950 (Online) = Water SA Vol. 45 No. 4 October 2019
Published under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0) 694
the total fire flow volume released during the incident. These

12 910.92
4 278.28
3 750.28
3 750.28
1 132.08
calculations account for water obtained from fire trucks, water

(ZAR)
Total
tankers, and hydrants.
The volume or flow rate is determined by reading the
meters installed on each appliance. In this paper, the fire
incident reports discussed are for fire incidents within the City

Amount

528.00
528.00
0.00

0.00
0.00
(ZAR)
of Johannesburg only and for the past 10 years (1 January 2006
to 30 September 2017).
Table 3 shows an example fire incident report which was
used for invoicing purposes. In the table, it can be seen that

Water

kL

30
0

0
0
4 different stations responded to one emergency and a total
of 30 kL of water was used for firefighting. The Malvern unit
was on-site for the longest duration – 2 h 27 min. It is assumed

(ZAR/kL)
17.60
17.60

17.60
17.60
Rate
that fire flow rate extracted from the municipal network was
constant over the on-site firefighting duration. While this
assumption produces an average flow rate per incident (Eq. 1), it
underestimates the peak firefighting flow (data which were not,

Amount

7 716.80
2 374.40
2 374.40
2 374.40
593.60
(ZAR)
and currently are not, recorded).

(1)

Personnel
No. of

4
4
4
1
From 1 January 2006 to 30 September 2017, there were 4

Personnel
556 billable firefighting incident reports recorded in the City of
Johannesburg. Of this number, 697 were recorded as incidents
that did not require municipal water and, therefore, the

(ZAR/h)
296.80
296.80
296.80
296.80
Rate
analysis below was based on 3 859 billable water use incidents.
This dataset excludes all non-billable fire incidents including
informal settlements, veld/grass and car/motorcycle fires.

Hours
RESULTS AND DISCUSSION

2
2
2
2
The scatter plot shown in Fig. 3 shows the magnitude and
Amount

4 666.12
1 375.88
1 375.88
1 375.88
538.48
distribution of the 3 859 fire flow volumes recorded from
1 January 2006 to 30 September 2017. Figure 4 shows the
magnitude and distribution of the 3 859 fire flow rates from
1 January 2006 to 30 September 2017. To gauge the validity
of the incidents with large fire flow volumes (> 300 kL) and
(ZAR/h)
687.94

269.24
687.94
687.94
Rate

fire flow rates (> 1 000 L/min), fire event characteristics (such
as duration, number of responding stations, presence of fire
safety officials, and fire location) were individually examined.
From this exercise, the fire incident circled in Figs 3 and 4
Hours

was identified as a likely data capture error because it did not

2
2
2
2

bear the same characteristics as the other large fire volume

incidents. The largest (800 kL) fire flow volume in the dataset
Vehicles
Table 3. An example fire incident report used for invoicing

was responded to by 6 different fire stations, lasted over 15 h

Release

20:00
19:33

19:33
19:33
time

and had fire safety officials present.

Arrival

17:33

17:18
17:15

17:19
time
description
Vehicle

Eng 03
BP 07

SC 02
TP 02 Turffontein

Turffontein
Total (ZAR)
Kibler Park
Stations

Malvern

Figure 3. Fire flow volumes for billable incidents between 1 January

2006 and 30 September 2017

https://doi.org/10.17159/wsa/2019.v45.i4.7551
Available at https://www.watersa.net
ISSN 1816-7950 (Online) = Water SA Vol. 45 No. 4 October 2019
Published under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0) 695
 Figure 4 also includes the SANS 10090 (2003) standard and Figure 6 presents the cumulative probability plot of average
The Red Book (DHS and CSIR, 2019) guideline values for Fire flow rates, with the SANS 10090 and The Red Book values
Flow as well as the Minimum Hydrant Flow for the different superimposed. The figure shows that 99.9% of all fire incidents
fire risk categories. An assumption made in the below analysis within the City during the designated period resulted in an
is that the recorded firefighting flows extracted from the average fire flow rate less than 1 200 L/min, which equals the
municipal network or fire equipment were what was required lowest of the Minimum Hydrant Flow rates for SANS 10090
to fight the fires. None of the 3 859 billable fire incident reports (i.e. Category D). Likewise, 99.7% of all fire incidents resulted
indicate otherwise. in an average fire flow rate less than 900 L/min, which equals
Figure 4 reveals that over the 10-year period, not a single the Minimuim Hydrant Flow rate for The Red Book’s lowest fire
fire incident in the City of Johannesburg recorded an average risk category (i.e. low risk). These findings suggest that there
flow rate greater than 6 000 L/min. This implies that over the is scope to reduce the current Minimum Fire Flows especially
10-year period, no incident can be classified as a SANS 10090 in low risk categories whilst maintaining adequate levels of
Category A, B or C nor The Red Book ‘high risk’ category fire. safety. Because of the need to fight low probability but high
During the 10-year period, only 2 incidents recorded average consequence fires in moderate- to high-risk fire category areas,
flow rates greater than 2 000 L/min. Three incidents recorded the authors caution on the application of the above statement to
average flow rates greater than 1 500 L/min. The vast majority these areas.
of average flow rates fell below both the SANS 10090 Minimum To better understand intra-day and intra-year firefighting
Hydrant Flow for Categories A, B, C and D and The Red Book trends, Fig. 7 shows, over an average month, the average
Minimum Hydrant Flow for ‘high risk’ and ‘moderate risk’ volume of water used to extinguish fires in relation to the
categories.
Figure 5 shows that 75% of fire incidents were extinguished
using less than 6.6 kL of water –this volume is less than the
capacity (6.9 kL) of a conventional pumping tanker within the
City of Johannesburg’s fleet purchased in 2003. This means that
over the study’s 10-year period, 75% of fire incidents in the City
of Johannesburg could have been extinguished without the
use of municipal fire hydrants if a pumping tanker with a full
tank of water was dispatched. The below quote from Myburgh
and Jacobs (2014 p.11) confirms similar results obtained for
3 municipal areas in the Western Cape: ‘only 8.6% of all fires
were extinguished using water from the WDS by connecting
firefighting equipment to a fire hydrant at the time of the fire.
Most fires were extinguished by means of water ejected from
a pre-filled tanker vehicle disconnected from the WDS at the
time of fighting the fire.’
Figure 5. Cumulative probability plot for all fire flow volumes

Figure 4. Average extracted fire flow rates for billable incidents between 1 January 2006 and 30 September 2017

https://doi.org/10.17159/wsa/2019.v45.i4.7551
Available at https://www.watersa.net
ISSN 1816-7950 (Online) = Water SA Vol. 45 No. 4 October 2019
Published under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0) 696
 Figure 6. Cumulative probability plot for extracted fire flow rates within the City of Johannesburg (1 January 2006 to 30 September 2017)

frequency of fire occurrence while Fig. 8 shows the frequency

of occurrence of fires and residential water use over a typical
day. In Fig. 7, the green bar chart shows the average number
of fire incidents occurring each month while the blue bar
chart shows the average fire flow volume per incident for
each month over the period 1 January 2006 to 30 September
2017. An expected seasonal trend is observed with regard to
frequency of fire occurrence, with a notable rise in incidents
from June to October, which are typically dry and low-rainfall
months in Johannesburg. While average fire flow volumes
range between 7 to 12 kL per incident, there is no observable
seasonal trend. These trends imply that, while the frequency
of fire occurrence is strongly related to climatic conditions,
the volume of water used to quench fires, and by implication,
the size of the fires, is not a function of climatic conditions
within the City of Johannesburg. As a consequence, seasonal
peak factors for fire flows may not be necessary when
incorporating the provision for water for firefighting in the
design of municipal mains within the City of Johannesburg or Figure 7. Average fire flow volume and number of incidents per
other metropolitan municipalities with similar fire flow and month between 1 January 2006 and 30 September 2017
climatic conditions.
Figure 8 displays the occurrence of incidents throughout
the course of a day, averaged over the period 1 January 2006
to 30 September 2017. The green graph shows the percentage
distribution of fire incident start times. In Fig. 8, three peaks (at
01:00, 15:00 and 20:00) are observed. The highest of the three
was at 01:00 – this represents 230 (5.9%) fire incidents. The blue
graph shows a typical diurnal residential water use pattern
published by Van Zyl (1996) (cited in Scheepers, 2012). The
water use pattern shows the primary peak residential demand
occurring at 06:00 while the secondary peak demand occurs
between 16:00 and 17:00. When compared to the start times
of fires within the City of Johannesburg, it is observed that the
start times of peak fires do not correspond with peak residential
water demand periods. The inverse is the case – the lowest
observed start times of fires were during peak demand periods.
This finding may therefore provide motivation to further Figure 8. Occurrence of fires versus a typical residential peak water
investigate the recommendation to cater for both instantaneous demand pattern

https://doi.org/10.17159/wsa/2019.v45.i4.7551
Available at https://www.watersa.net
ISSN 1816-7950 (Online) = Water SA Vol. 45 No. 4 October 2019
Published under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0) 697
peak demand and fire demand during WDS design as using one of the City of Johannesburg’s conventional
recommended by The Red Book (DHS and CSIR, 2019: J.3.2.2) pumping tankers which have a capacity of 6.9 kL. This,
i.e.: ‘Conveyance infrastructure should have sufficient capacity by implication, means that 75% of fire incidents within
for peak demand conditions and fire-flow requirements, in the City could have been extinguished without the use of
accordance with the design guidelines in this document’ and municipal fire hydrants if a suitable tanker with a full tank
(CSIR, 2005: volume 2, Chapter 9, page 27): ‘The nominal of water was available.
capacity of the duty pump should be equivalent to the sum of o A third highlight was that, while the frequency of fire
the instantaneous peak demand and the fire demand (obtained occurrence was strongly related to climatic conditions, the
from the section on provision of water for firefighting), or volume of water used to quench the fires was not a function
the instantaneous peak demand plus an allowance of 20%, of climatic conditions
whichever is the greater.’ o A fourth highlight was that the start times of peak fires did
not correspond with peak residential water use periods
CONCLUSIONS AND Recommendations within the City of Johannesburg over the 10-year period.
The inverse was however the case – the lowest observed fire
The key results and recommendations arising from the two incidents occurred during peak demand periods
objectives addressed in this study are presented below: Based on the above findings, and the assumption that the
• Objective 1: To present an analysis of international and South results from this study can be generically applied, the following
African design standards and guidelines pertaining to water recommendations can be made:
provision for firefighting • A Minimum Hydrant Flow of 1 200 L/min is recommended
o A review of national and international standards and for all SANS 10090 and The Red Book Categories. SANS
guidelines for water provision for firefighting are presented 10252-1:2012 (SABS, 2012) stipulates the same value.
in the text. A notable distinction between the SANS 10090 • To improve the efficiency of firefighting within the City of
(SABS, 2003) standard and The Red Book (DHS and Johannesburg, especially considering the potential devastation
CSIR, 2019) guideline, apart from the different values they that could occur due to increasing instances of water cuts
recommend for Fire Flow and Minimum Hydrant Flow, is and low pressures (Kahanji et al., 2019), EMS should focus on
their differing fire risk categories. The Red Book presents a acquiring pumping appliances with sufficient capacity and
single set of fire risk categories while SANS 10090 presents volume (minimum of 6.6 kL) to extinguish fires.
two fire risk categories which, in certain instances, do Based on the findings of this study, future research may
not recommend consistent fire flow values for the same investigate:
category. • The need for seasonal peak factors when incorporating the
o In addition, The Red Book, which is a guideline, in all provision for water for firefighting in the design of municipal mains
instances, violates the Minimum Fire Flows in SANS • Catering for both instantaneous peak demand and fire
10090, which is a standard that stipulates minimum demand during WDS design as recommended by The Red
acceptable values. Book (DHS and CSIR, 2019)
o It is therefore a recommendation of this paper that the • Understanding the change in rate of water use during a fire event
SANS 10090 fire risk categories (A, B, C, D and E) be
revised. As a result of their simplicity and recent revision, AKNOWLEDGEMENTS
The Red Book (DHS and CSIR, 2019) fire risk categories
may be adopted in the recommended revision of the SANS The participation and data provided by the City of
10090 fire risk categories. Johannesburg’s Emergency Management Services personnel are
• Objective 2: To present actual fire flow data recorded in gratefully acknowledged. Also gratefully acknowledged is the
the City of Johannesburg, to compare this data with the postgraduate funding and support by Mott MacDonald.
primary South African standard and guideline values, and
to make recommendations to guide future revisions to the REFERENCES
primary South African design standard and guideline for the
provision of water for firefighting. BENFER M and SCHEFFEY J (2014) Evaluation of Fire Flow
o Fire incident reports were obtained from the City of
Methodologies. The Fire Protection Research Foundation, Quincy,
Massachusetts. https://doi.org/10.1007/978-1-4939-2889-7
Johannesburg’s EMS for the period 1 January 2006 to 30
CSIR (2005) Guidelines for Human Settlement Planning and Design.
September 2017. These reports show that the majority Red Book. CSIR Building and Construction Technology, Pretoria.
of average fire flow rates fell below both the SANS DHS and CSIR (Department of Human Settlements, South Africa
10090 (2003) and The Red Book (DHS and CSIR, 2019) and Centre for Scientific and Industrial Research (2019) The
Minimum Hydrant Flow for all its categories. The below Neighbourhood Planning and Design Guide (Red Book): Creating
highlights are evidence of this: Sustainable Human Settlement. DHS and CSIR, Pretoria. ISBN
◉ Almost all (99.9%) fire incidents recorded an average 978-0-6399283-2-6.
DAVIS SK (2000) A review of fire fighting water requirements. ME
fire flow rate less than 1 200 L/min – the lowest thesis, University of Canterbury.
Minimum Hydrant Flow rate for the SANS 10090 KAHANJI C, WALLS RS and CICIONE A (2019) Fire spread analysis
Categories for the 2017 Imizamo Yethu informal settlement conflagration in
◉ Similarly, 99.7% of all fire incidents recorded an average South Africa. Int. J. Disaster Risk Reduct. https://doi.org/10.1016/j.
extracted fire flow rate less than 900 L/min – the lowest ijdrr.2019.101146
Minimuim Hydrant Flow rate for The Red Book’s categories LAW M and BEEVER P (1995) Magic numbers and golden rules. Fire
o A second finding from the analysis of fire incident reports
Saf. Sci. 4 79–84. https://doi.org/10.3801/IAFSS.FSS.4-79
MYBURGH HM and JACOBS HE (2014), Water for firefighting in
was that 75% of fire incidents were extinguished using less five South African towns. Water SA 40 (1) 11–17. http://dx.doi.
than 6.6 kL of water and thus could have been extinguished org/10.4314/wsa.v40i1.2

https://doi.org/10.17159/wsa/2019.v45.i4.7551
Available at https://www.watersa.net
ISSN 1816-7950 (Online) = Water SA Vol. 45 No. 4 October 2019
Published under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0) 698
RSA (Republic of South Africa) (1987) Fire Brigade Services Act. Act No. water demand by means of a probability based end-use model.
99 of 1987. Government Gazette 11006, 23 October 1987, Issue 99. Master of Science in Engineering, Stellenbosch University.
SABS (South African Bureau of Standards) (1972) SABS 090:1972. SCHLOTFELDT CJ (1995) The “Red Book”: Background, purpose and
Community protection against fire. SABS, Pretoria. way forward. Paper presented at Red Book Workshops. February
SABS (South African Bureau of Standards) (2003) SANS 10090:2003 and March, Bloemfontein, Cape Town, Pretoria, Durban and Port
South African National Standard. Community protection against Elizabeth.
fire. Standards South Africa (a division of SABS), Pretoria. ISBN VAN ZYL JE (1996) Peak factors in municipal water reticulation
0-626-14666-6 networks. Proceedings of the Water Institution of South Africa
SABS (South African Bureau of Standards) (2012) SANS 10252-1:2012. Biennial Conference and Exhibition, 20–23 May 1996, Port
South African National Standard. Water supply and drainage for Elizabeth, South Africa.
buildings Part 1: Water supply installations for buildings. SABS VAN ZYL JE and HAARHOFF J (1997) South African fire water
Standards Division, Pretoria. guidelines and their impact on water supply system cost. J. S. Afr.
SCHEEPERS HM (2012) Deriving peak factors for residential indoor Inst. Civ. Eng. 39 (1) 16–22.

https://doi.org/10.17159/wsa/2019.v45.i4.7551
Available at https://www.watersa.net
ISSN 1816-7950 (Online) = Water SA Vol. 45 No. 4 October 2019
Published under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0) 699