Professional Documents
Culture Documents
Briefing Paper
www.leonardo-energy.org
February 2007
Being responsible for the structural design and technical infrastructure of buildings
with high fire-safety requirements usually means selecting cable and wiring in the
highest fire resistance classes. But what do terms like “fireproof”, “flame retardant”
and “flame resistant” actually mean? Do these products really meet public safety
requirements, or are there other, less well-known technologies that offer greater
safety and that at a lower price?
One of the great advantages of electric current is that it enables relatively large amounts of
energy to be concentrated in very small regions of space and in very short periods of time.
Within a matter of milliseconds, extremely high temperatures can be generated in a highly
constrained region. Indeed good use is made of this effect in arc welding, spark erosion
machining or in electrical ignition devices. Another example is the electric light bulb in which
an extremely high electric power density is present in a very delicate metal filament.
However, the possibility of creating such high concentrations of energy brings with it the risk
that they might arise accidentally, whether as a natural occurrence such as lightning or due
to a fault in a piece of electrical equipment. Should a fault, in particular a short circuit,
develop in a building’s low-voltage distribution system, the cables and wiring can be
subjected to a linear power density of over 20 kW of heat per metre, causing extremely rapid
overheating of the cable. It is therefore essential that appropriate protective measures are
put in place. The introduction of electrical devices as replacements for the open flames that
were once the main source of interior lighting and heating led to a massive reduction in the
fire hazards in residential and commercial buildings. Nevertheless, fire experts estimate that
electric power (excluding lightning damage) still causes around three billion euros of damage
a year – accounting for some ten percent of the total claims made on home content, fire and
building policies.
Tragedies such as the fires at Düsseldorf Airport, the London Underground and in numerous
alpine tunnels, which caused major damage to property and significant injury and loss of life,
have repeatedly drawn public attention to the importance of fire prevention measures when
designing and installing electrical systems. But just when is a fire “caused by electricity”? Is a
cable fire nothing more than a burning cable, or does the term imply a fire that arose
because the cable was electrically, and thus thermally, overloaded? Unfortunately, the term
is not precisely defined.
www.l eonar do- ener gy. or g
Figure 2: The cable and wiring technique is no more difficult than with conventional cables –
merely different (Image: Sachverständigenbüro Karl-Heinz Otto, www.sv-otto.de)
The key requirement when constructing or renovating a building must therefore be to prevent
cable fires, i.e. any form of burning cable. Perhaps surprisingly, cables able to prevent a fire
from actually starting irrespective of whether the ignition energy comes from the electrical
current within the conductor or is applied externally have been available on the market for
many years. These mineral-insulated cables (fig. 1) do not age, emit neither smoke nor
gases if exposed to fire, are completely non-combustible and constitute a minor or even non-
existent fuel load, depending on whether they are supplied uncovered or with an optional low
smoke and fume (LSF) sheath (fig. 3). Conventionally insulated cables by contrast emit high
levels of smoke when subjected to fire testing such as that described in the French standard
NSF C 32 070 in which the cable is heated in a cylindrical electrical furnace according to an
ISO-standardized heating profile. Figure 4 shows the standardized curve and the actual
temperature profile recorded. It is readily apparent that the ignition of the sheath causes the
temperature to deviate strongly from the target curve. The amount of heat generated by
combustion is so great that the furnace’s thermostatic control is unable to compensate for
the temperature rise. Perhaps for this reason alone the test cable fails prematurely. The
cable is therefore destined to fail in practice.
FIRE PERFORMANCE
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30 negative
0.20
0.10
0.0 PVC/PVC PVC/PVC PVC/PVC Halogen Free Halogen Free Mineral-Insulated Mineral-
Stranded Stranded Cross Linked Cross Linked Cable with LSF Insulated
Solid Polymar Outer Cover
Conductors Conductors Concentric Polymar Bare Cable
Concentric
Temperature (°C)
1,200
15 Minutes
1,000 Cable
Failed
920
902
800 842
739
600
400
200 Approximately
65 Minutes
20
0
0 15 30 45 60 75
Time (Minutes)
Figure 4: Target heating curve and actual temperature profile of a fire-resistant plastic-
coated cable (BICC, Manchester)
conditions match the actual conditions experienced in a real fire. The German, British and
French test procedures all involve exposing the test sample to a flame with the sample either
freely suspended or mounted on a rigid surface. The procedure set out in the NSF C 32 070
standard also includes exposure to mechanical stress caused by an iron rod of the same
diameter as the cable that falls every 30 seconds onto the cable. However, the rod does not
strike that part of the cable exposed to the flame, but one of its cold ends instead.
The British standard 6387 includes the following variants:
• Category A test: Flame exposure at 650 °C
• Category C test: Flame exposure at 950 °C,
• Category X test: Flame exposure at 650 °C and mechanical shock
• Category Z test: Flame exposure at 950 °C and mechanical shock
• Category W test: Flame exposure at 650 °C and water spray
The standard does not include a test that involves exposing the cable to a flame at a
temperature of 950 °C and to water simultaneously. It seems rather unrealistic to assume
that a fire causing a temperature of 950 °C then drops down to 650 °C before the sprinkler
system triggers, the fire service arrives, or water pipes burst. The mechanical shocks to
which the cable is subjected are limited to vibrating the support surface to which the sample
is mounted; the cable itself remains untouched. It seems more than unlikely that falling
debris, such as ceiling tiles, would only hit the cable tray or conduit and would fail to hit the
cable itself. A further questionable aspect is that each test is performed with a new sample
and never with the same one. This is tantamount to saying that a particular piece of cable
will only be subjected to one such event and would never experience multiple stressing
either simultaneously or sequentially.
The German standard, DIN 4102 Part 12, requires testing of a moderately curved sample
mounted on a support structure, but ignores all other influencing factors and only stipulates
that the sample be exposed to a flame. The typical insulation materials burn off completely in
the first few minutes and even LSF cables emit substantial quantities of smoke. After the
cable sheath has burnt and disintegrated, the only material holding the cable together and
separating the cores from one another is a shell of ash. The test procedure fails to examine
what the consequences would be if these cables were exposed to fire when clamp-mounted
or when bent or kinked within a cable tray or duct. By contrast, mineral-insulated cables
undergo practically no change during the flame exposure test and continue to function even
after being subjected to numerous severe hammer blows.
areas than they do at present. A cable offering functional integrity only under non-critical
conditions is hardly worthy of the name. Up until recently there was no national or
international standard that dealt with this issue. At present only the Belgian standard
RGIE/AREI, article 104, section e4, addresses this point and requires a five-fold increase in
the cross-sectional area of the conductor so that even at 950 °C, the temperature rise will be
no greater than 40 K. Note that this phenomenon will be observed in all cable types whether
mineral-insulated or plastic-insulated. No current test specification requires fire testing of
cables under significant load. If they did, all test samples would fail. The only load connected
during cable testing is a filament lamp that is included simply to indicate functioning and the
moment of failure.
A further problem encountered by the Belgian engineers when investigating this problem
was the different approaches that can be adopted to compute the temperature-dependence
of the resistance of copper. One approach is to apply a rearranged version of the formula
that is quoted in the EN 60742 standard and is used to compute the temperature rise in
wound components from measurement of the change in resistance:
RT = R20 [1 + α (T − 293K )]
RT is the resistance at temperature T, R20 the resistance at 20°C (293 K) and is the
temperature coefficient. The value of for copper is 0.0039. This formula ignores the
coefficient , which as it is multiplied by the square of the temperature rise can be neglected
when the temperature difference is small:
[
RT = R20 1 + α (T − 293K ) + β (T − 293K ) 2 ]
The third approach makes use of the Wiedemann-Franz power law:
1.16
T
RT = R20
T20
The latter method is the most accurate of the three according to the Belgian team. Figure 5
shows that for most technical applications, the deviations between the three curves are
negligible at typical operating temperatures, but significant differences are apparent at the
temperatures experienced in a fire. Despite being included in the standard, the linear model
(i.e. without the coefficient ) is clearly unsuitable for calculating the temperature
dependence of the resistance at fire temperatures.
5,5
5,0
R/R 20°C
4,5
4,0
3,5
3,0
2,5
2,0
calculated according to EN 60742
1,5 calculated including square component
calculated using exponential formula
1,0
T
0,5
0°C 200°C 400°C 600°C 800°C 1000°C
Figure 5: Temperature dependence of the resistance of copper, calculated using three
different methods
There is at least one cable manufacturer that exceeds the specifications of the standards by
testing for the, albeit unlikely, case that the conductor is overloaded. In this test, a mineral-
insulated and a plastic-insulated sample each with a conductor cross section of 2.5 mm² are
connected in series and subjected to a current of 200 A. Depending on the particular
structure of the sample, the plastic-sheathed cable either ignites suddenly along its entire
length or one of its cores melts. But even before this occurs, large quantities of smoke have
escaped from both ends of the cables, smoke which can be ignited by the smallest spark.
The MI cable, by contrast, has withstood numerous repeat tests without damage, with the
same test sample cable being used for months at a time.
suppliers in these countries. A list of suppliers is available from the German Copper Institute
(DKI).
Comparison of Dimensions
Cable diametre
13.2 mm
Cable diametre
7.2 mm
Mineral-insulated cable
2 * 1.5mm2 with copper sheath as protective earth
Cable comprises copper sheath and mineral insulant, optional outer sheath of LSF plastic
If mineral-insulated cables are installed, there is no risk that a fire could lead to
environmental contamination. Plastic-insulated cables, on the other hand, can produce
hydrochloric acid that can diffuse into concrete masonry corroding the reinforcing steel. In
some cases, relatively minor cable fires have made demolition of the building necessary.
While the newer halogen-free cables avoid this hazard, they still release smoke and carbon
monoxide when exposed to fire. According to fire safety experts, most fire victims do not die
in the flames, but lose their orientation because of the smoke and are then killed by carbon
monoxide poisoning. Magnesium oxide is completely unreactive and can be disposed of as
an environmentally neutral waste product, and the copper can of course be recovered and
recycled almost indefinitely.
Figure 8: Efforts to protect this historic monument would have been ruined had thick, bright-
orange plastic-insulated fireproof cable been installed (Image: LMI Europa Metalli, Fornaci di
Barga / Italien)
8 Summary
Current standards for “cables with improved fire performance” have been carefully
formulated with an eye to what is technically feasible so that manufacturers of plastic-
insulated cables can just about manage to comply with the requirements of the standard if
they make use of specialized polymers. Given this situation, it is not unreasonable to
consider using MI cable when designing and dimensioning the wiring for areas with
enhanced safety requirements, as this would go well beyond the specifications in present
national and international standards. A first step in this direction was taken in CentrO, a new
ii
shopping centre in Oberhausen, where 100 km of mineral-insulated cable were installed .
European and international standards organizations should also follow the example set by
the Belgian standard and as soon as possible (which experience shows is not that fast)
revise their specifications to include greater conductor cross-sections in order to counteract
the significant increase in conductor resistance when cables are exposed to fire.
However, standards are not laws and the law is clear that no matter what the standards
might say, trained and qualified electricians must act responsibly and are responsible for
their actions. This is tantamount to requiring the use of the best available technology
whenever and wherever this would improve personal safety.
i
www.eupen.com
ii
Reference: HTW Hetzel, Tor-Westen & Partner Ingenieurgesellschaft KG, www.htw-
ingenieure.de