Experimental characterisation of ITER electric cables in postulated fire scenarios

Fusion Engineering and Design 88 (2013) 2650–2654

Contents lists available at ScienceDirect

Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes

Experimental characterisation of ITER electric cables in postulated fire scenarios Roberto Passalacqua a,∗, Pierre Cortes a, Neill Taylor a, David Beltran a, Pascal Zavaleta b, Stephane Charbaut b a b

ITER Organization, Route de Vinon sur Verdon, 13115 Saint Paul Lez Durance, France Institut de Radioprotection et de Sûreté Nucléaire (IRSN), Fires Experimentation Laboratory, Centre de Cadarache, 13115 St Paul-Lez-Durance, France

h i g h l i g h t s    

Power and control cables account for roughly 70% of all ITER fire loads. Industrial cables are tested to ensure ITER requirements are reflected in the procurement phase. Radiative heat flux leading to pyrolysis and/or ignition, flame propagation speed, generation of heat and mass loss are measured. Influence of cable shafts configurations on fire propagation is assessed.

a r t i c l e

Article history: Available online 31 January 2013 Keywords: Fire load Electric cables Pyrolysis Fire propagation Smoke release Fire resistance

a b s t r a c t

i n f o

Power and control cables account for roughly 70% of all ITER fire loads. Fire experiments were executed at IRSN Cadarache to test several industrial cables in order to ensure that the specificities of the ITER project could be reflected in the forthcoming procurement phase. Experimental results are here summarised: radiative heat flux leading to pyrolysis and/or ignition, flame propagation speed, generation of heat by combustion, mass loss as well as gas and smoke releases. Also the influence on fire propagation of cable trays’ geometry has been assessed: undoubtedly, vertical cable shafts promote fire propagation. Experimental data will be used to improve the current modeling of fire phenomenology and to assess propagation/impact on ventilation systems and Safety Importance Class (SIC) components with the ultimate target to avoid any potential release of radiological inventories. © 2013 Elsevier B.V. All rights reserved.

1. Introduction This paper presents the results of fire experiments carried out by IRSN in Cadarache (under contract with ITER IO). Four power cables and two control cables were tested in a small-scale facility (CARINEA). Four cables (three power and one control cables) were then selected to be further tested in a large-scale facility (SATURNE). For the purpose of these tests, all cables were categorised against EU/NF standards.

2. Small-scale fire tests (CARINEA facility) The CARINEA facility consists of a source of ignition, a carousel device on which the cable samples are located, and a 1.5 mdiameter exhaust hood (Fig. 1).

∗ Corresponding author. E-mail address: [email protected] (R. Passalacqua). 0920-3796/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fusengdes.2013.01.026

The six tested cables are shown in Fig. 2. Power cables are composed of three conductors with their own insulation sheaths enveloped by an outer sheath (Table 1), while control cables have additional wires/sheaths (Table 2). The heat flux reaching the cables’ surface is controlled and maintained constant (a radiant panel is able to deliver a maximum, adjustable heat flux of 40 kW/m2 ). Once pyrolysis starts and gas/smoke is released, an electric arc is induced to allow ignition. In some cases, there is ignition of gas and cables, while in others only a pyrolysis process is observed. In both cases the test is stopped (the radiant panel is switched off) when the measured gas/smoke release approaches zero. Fig. 3 shows, as an example, the ACEFLEX cables during pyrolysis (see the 20 × 20-cm2 carousel device in the center of Fig. 1) while Fig. 4 shows the cables’ status once completely burned (i.e. after ignition). Following the ASTM E 1623-04 protocol, 38 tests were carried out from May 10th to June 17th, 2011 (see reference [1] for details about the experimental protocol). Table 3 gives the ignition heat flux, IHF in kW/m2 . Note that for cable no. 2 the measurement of the ignition heat flux was affected by a high uncertainty: only 2

R. Passalacqua et al. / Fusion Engineering and Design 88 (2013) 2650–2654

2651

Table 1 Chemical composition of power cables. Power cables

Outer sheath

Conductors’ insulation

TITANEX H07RN-F 3G2.5 ALSECURE 0.6/1 kV 3G2.5 ACEFLEX RV-K 0.6/1 kV 3G2.5 FIREX PROTECH RZ1-K 3G2.5

PEa , PAb , CaCO3 , Talcc PE, PVA, CaCO3 , Al2 Si2 O5 (OH)4 e PVCf , phthalatesg , CaCO3 , talc PE, CaCO3 , PVA

PE, PVAd , CaCO3 , Talc PE, CaCO3 PE PE, Talc

a b c d e f g

Polyethylene: [ CH2 CH2 ]n . Polyamide: includes amide function, C( O) NH . Talc: product derived from SiO2 . Polyethylene vinyl acetate: [ CH2 CH2 ]n [ CH2 Kaolinite: Al2 Si2 O5 (OH)4 . Polyvinyl chloride: [ CH2 CHCl ]n . Phthalate: R O C( O) C6 H4 C( O) O R.

CH O C( O) CH3

]m

.

Table 2 Chemical composition of control cables. Control cables

Outer sheath

Conductors’ insulation

Additional elements

SABIX LC02162A

PE, Al(OH)3 , PVA PE, Al(OH)3 , PVA

Talc, PPa , PEPb PE, PVFc , Talc

White wires between conductors made of cellulose White internal sheath made of PE, Al(OH)3 , PVA, Talc

a b c

Polypropylene: [ CH2 CH(CH3 ) ]n . Polyethylene–propylene: [ CH2 CH2 ]n [ CH2 Polyvinylidene Fluoride: [ CH2 CF2 ]n .

CH(CH3 ) ]m

.

Fig. 1. CARINEA facility.

Table 3 Ignition heat flux.

Fig. 2. Cables tested in the CARINEA facility. 2

Cable

Reference

IHF (kW/m )

No. 1 No. 2 No. 3 No. 4 No. 5 No. 6

TITANEX29 H07RN-F 3G2.5 ALSECURE 0.6/1 kV 3G2.5 SABIX D 345 FRNC TP (4 × 2 × 0.34 mm2 ) ACEFLEX RV-K 0.6/1 kV 3G2.5 FIREX PROTECH RZ1-K 3G2.5 LC02162A (3 × 2 × 0.5 mm2 )

36 ± 1 37 ± 3a 39 ± 1 21 ± 2 40b 24 ± 1

a Best estimation of the measurement error (which might be actually higher due to the fact that the maximum achievable incident heat flux was 40 kW/m2 ). b The measurement error is undetermined (see previous note).

of the 8 tests performed (with a heat flux ranging from 33.75 to 40 kW/m2 ) led to cable ignition, therefore the measurement error is somehow undetermined since it was not possible to test the cable at higher heat fluxes (see note1 ).

1 Best estimation of the measurement error (which might be actually higher due to the fact that the maximum achievable incident heat flux was 40 kW/m2 ).

Fig. 3. ACEFLEX cables undergoing pyrolysis (20 × 20 cm2 sample).

2652

R. Passalacqua et al. / Fusion Engineering and Design 88 (2013) 2650–2654

Table 4 Power release and heat of combustion. Cable/heat flux 2

No. 1/40 kW/m No. 2/35 kW/m2 No. 3/40 kW/m2 No. 4/40 kW/m2 No. 5/40 kW/m2 No. 6/40 kW/m2

Max. HRRS kW/m2

Average HRRS kW/m2

Average HtCb MJ/kg

THRS MJ/m2

170 210 170 415 50 220

140 (over 3 min) 180 (over 3 min) 150 (over 3 min) 130 (over 2 min) 45 (over 2 min) 180 (over 3 min)

16 (over 3 min) 20 (over 3 min) 16 (over 3 min) 11 (over 2 min) 9 (over 2 min) 20 (over 3 min)

37 (over 5 min) 48 (over 5 min) 34 (over 5 min) 40 (over 5 min) 15 (over 5 min) 46 (over 5 min)

Table 5 “Smoke” released from cables. Cable

TSRS ,Tsi (adimensional) TSRS ,Tie (adimensional) (tests with ignition)

TSRS ,Tse (adimensional) (tests with ignition)

TSRS ,Tse (adimensional) (tests without ignition)

No. 1

610 483 1699 351 452 203 580 1829 424 349 1067 737

1093

3062

2050

1613

655

2034

2409

2266

773

849

1804

3551

No. 2 No. 3 No. 4 No. 5 No. 6

Table 4 shows maximum and average power released when exposing the 20 × 20-cm2 carousel device (where cables were located) at the maximum incident heat flux of 40 kW/m2 (with the exception of cable 2 tested at 35 kW/m2 see Table 4), i.e. the maximum heat release rate (per unit of cable sample surface) HHRS , the average HRRS (over 2 or 3 min after the onset of ignition), the average heat of combustion HtCb and the total heat released (per cable sample surface) THRS (averaged over 5 min from the fire onset). Note that the average heat of combustion HtCb is assessed from Eq. (1): HtCb =

HRRS MLRS

(1)

where MLRS is the mass loss rate (per unit of cable sample surface) in g/s/m2 (see Fig. 5 related to the first test of Table 4). Cable ignition generally occurred at high heat fluxes i.e. 35–40 kW/m2 with the exception of cables no. 4 and 6 (see Table 3). The total smoke released (per unit of cable surface) TSRS , estimated from measurements of the opacity of the exhaust flow (and adimensional, see Table 5), strongly depended on the ignition occurrence.

The first data column of Table 5 shows the smoke released from the start of the test up to ignition (TSRS,Tsi ) and the smoke released from ignition to the end of test (TSRS,Tie ). The second column gives the total amount of smoke released from the start to the end of test (TSRS ,Tse ). Table 5 shows that the majority of cables (no. 1, 3, 5 and 6) released more smoke when ignition did not occur (compare the last 2 columns). This is obviously explained by the fact that, since there is no ignition (i.e. cable temperatures do not reach ignition temperatures) some of released mass remain unburned. The same behavior is noted before and after ignition (i.e. for tests with ignition, see the first column): more smoke is released before ignition than after (e.g. for cables no. 1, 2, 3, 5 and 6). A few cables (no. 2 and 4) exhibited a strong pyrolysis with a production of smoke which was even higher after ignition. This is explained by the fact that these two cables, even though characterised by a lower mass loss rate (per unit of cable surface, MLRs , see Ref. [1]), generated a smoke composed by a population of aerosols of small size which increased the opacity of fire gases.

16 cable_1_40_20x20

MLRfa (g/s/m2)

12

8

4

0 500

900

1300

time (s) Fig. 4. ACEFLEX cables after ignition (20 × 20 cm2 sample).

Fig. 5. Mass loss rate per unit of surface of cable sample (MLRS in g/s/m2 for test cable No. 1 at 40 kW/m2 – note that MLRs is here indicated “MLRfa”).

R. Passalacqua et al. / Fusion Engineering and Design 88 (2013) 2650–2654

Test PST PST PST PST PST PST

IT01 IT01 IT01 IT01 IT01 IT01

V-ALSECURE V-FIREX V-FIREX-2 V-SABIX H-FIREX H-ACEFLEX + SABIX

Cable tray

Test duration/power

Vertical Vertical Vertical Vertical Horizontal Horizontal

287 s/85 kW 342 s/85 kW 224 s/28 kW 261 s/28 kW 567 s/85 kW 611 s/85 kW

Outside surface temperature (°C)

Table 6 SATURNE fire cable tests.

3. Large-scale fire tests (SATURNE facility) The cables tested in the CARINEA facility were classified according to their reaction to fire, i.e. ignition heat flux and heat and smoke release ([1]). Four cables (three power and one control cable) were further tested in the SATURNE facility to characterise their resistance to fire propagation in a vertical and horizontal configuration. Four tests were carried out with a vertical cable tray (open vertical shaft) while two tests used a horizontal cable tray. They were performed between 15th of September and 10th of November, 2011 using a source of ignition (a gas burner) at two (low and high) power levels (Table 6). Fig. 6 here above shows the vertical tray located in the SATURNE enclosure (2000 m3 − 10 m long, 10 m wide and 20 m high) before and after the test. A large number of thermocouples were located at a regular distance from the bottom to the top of the cables’ surface. Measured surface temperatures versus time are shown in Fig. 7 (the on-off trend gives the temperature at the burner location – green line). These measurements allowed an estimation of the fire propagation velocity (from 57 to 78 cm/min, see reference [2]). The vertical tray configuration, being relatively enclosed (a 3-wall tray with one open side), was the cause of a strong ventilation/recirculation which supplied the fire with fresh air at all times (chimney effect, see Fig. 8). Fig. 9 shows the estimation of the total heat release rate HRR (in kW) for the first four tests carried out in the vertical tray. Note that for the first two tests the power of the source of ignition (as well as the duration of its action) was rather high (85 kW see Table 6), while the third and fourth vertical tests were run with a source of ignition of 28 kW This explains the faster fire development of

2653

1000 TCS_CAB2_0500

800

TCS_CAB3_1000 TCS_CAB2_2000

600

TCS_CAB3_2000 TCS_CAB3_2500

400

TCS_CAB2_3000 BURNER_ON

200 0

0

400

800

1200

time (s) Fig. 7. Surface temperature along the cables at various heights (PST IT01 VALSECURE cables).

Fig. 8. ALSECURE cable during the fire test.

1200 1000

V-ALSECURE

HRR (kW)

V-FIREX

800

V-FIREX-2 V-SABIX

600

BURNER_ALSECURE BURNER_FIREX

400

BURNER_FIREX-2 BURNER_SABIX

200 0 0

500

1000

1500

2000

Time (s)

Fig. 6. ALSECURE cable before and after the fire test.

Fig. 9. Heat release rate during the four vertical tests (note that the power of the burner for the FIREX-2 and SABIX cables, i.e. 28 kW is not correctly indicated).

2654

R. Passalacqua et al. / Fusion Engineering and Design 88 (2013) 2650–2654

100

25 MLR

80

Ht Comb

Ht Comb (MJ/kg)

MLR (g/s)

20

Table 6). Other data referring, for example, to cable mass burning rates, CO2 and CO releases, unburned soot releases, soot granulometry, as well as to the spatial characterisation of radiative heat flux from cables can be found in reference [2]. Experiments with vertical cable trays exhibited fast fire propagation. Horizontal tests showed instead that because of the cable fire resistance (even when using a source of ignition characterised by a long-lasting high power i.e. 85 kW for 567 s, see Table 6), fires quickly extinguish once the source of ignition is removed (Fig. 11).

15

60

10

40

5

20

4. Conclusions

0

ITER power and control cables constitute an important fire load (roughly 70% of total). Fire experiments are in progress at IRSN Cadarache to characterise cables’ performances and to test configurations which reduce risk of fire propagation. The first two campaigns of experiments (CARINEA and SATURNE) provided a large number of data as fire power release, cables’ heat of combustion and smoke (soot) characterisation. Large-scale experiments also showed that a specific cable with good fire retardant capabilities is not further burning when the source of ignition is removed from a horizontal tray, whilst it will completely burn if located in a vertical tray. The fire escalation is initially strongly depending on the power and duration of the source of ignition but in the longer term essentially depends on the geometric characteristics of the vertical tray/shaft (chimney effect). Further planned experiments at IRSN are going to investigate fires from horizontal trays and in particular to test potential fire propagation within multiple horizontal cable trays. These past and future experiments will help to further improve ITER safety level against fire risks by selecting the most appropriate cable materials and fire safety provisions.

0 100

200

300

400

500

600

700

800

time (s) Fig. 10. Mass loss rate and specific heat of combustion for the FIREX cables in the vertical test.

5. Disclaimer The views and opinions expressed herein do not necessarily reflect those of the ITER Organisation. Fig. 11. Cable tray after the fire horizontal test (FIREX cables).

Acknowledgements the two first tests (with ALSECURE and FIREX cables respectively the green and red trends) promoted by a strong air recirculation. The response of the FIREX cable to the two different sources of ignition can be shown by comparing the red and the blue trends. The SABIX control cable, with a diameter comparable to the FIREX and ALSECURE power cables (but additional shielding materials, see Table 2), exhibited a larger amount of unburned debris and a low HRR (compare the brown and the blue trends in Fig. 9). Fig. 10 shows the mass loss rate (in g/s) and the specific heat of combustion (in MJ/kg) for the second vertical test (FIREX cables, see

Authors are thankful to all NSA Division members and to IRSN colleagues involved in fire experiments for their continuous support and help in preparation of this paper. References [1] P. Zavaleta, Fire loads and behaviour from cables fire – small scale tests in the CARINEA facility, SERCI-2011-207, 2011. [2] P. Zavaleta, Fire loads and behaviour from cables fire – cables fire in SATURNE, SERCI-2011-318, 2011.