Statistical Characterization of the Time to Reach Peak Heat Release Rate for Nuclear Power Plant Electrical Enclosure Fires

Authors: Raymond H.V. Gallucci

Since publication of NUREG/CR-6850 (EPRI 1011989), EPRI/NRC-RES Fire PRA Methodology for Nuclear Power Facilities in 2005, phenomenological modeling of fire growth to peak heat release rate (HRR) for electrical enclosure fires in nuclear power plant probabilistic risk assessment (PRA) has typically assumed an average 12-minute rise time. [1] One previous analysis using the data from NUREG/CR-6850 from which this estimate derived (Gallucci, “Statistical Characterization of Cable Electrical Failure Temperatures Due to Fire, with Simulation of Failure Probabilities”) indicated that the time to peak HRR could be represented by a gamma distribution with alpha (shape) and beta (scale) parameters of 8.66 and 1.31, respectively. [2] Completion of the test program by the US Nuclear Regulatory Commission (USNRC) for electrical enclosure heat release rates, documented in NUREG/CR-7197, Heat Release Rates of Electrical Enclosure Fires (HELEN-FIRE) in 2016, has provided substantially more data from which to characterize this growth time to peak HRR. [3] From these, the author develops probabilistic distributions that enhance the original NUREG/CR-6850 results for both qualified (Q) and unqualified cables (UQ). The mean times to peak HRR are 13.3 and 10.1 min for Q and UQ cables, respectively, with a mean of 12.4 min when all data are combined, confirming that the original NUREG/CR-6850 estimate of 12 min was quite reasonable. Via statistical-probabilistic analysis, the author shows that the time to peak HRR for Q and UQ cables can again be well represented by gamma distributions with alpha and beta parameters of 1.88 and 7.07, and 3.86 and 2.62, respectively. Working with the gamma distribution for All cables given the two cable types, the author performs simulations demonstrating that manual non-suppression probabilities, on average, are 30% and 10% higher than the use of a 12-min point estimate when the fire is assumed to be detected at its start and halfway between its start and the time it reaches its peak, respectively. This suggests that adopting a probabilistic approach enables more realistic modeling of this particular fire phenomenon (growth time).

Comments: 10 Pages.

Download: PDF

Submission history

[v1] 2017-06-12 12:58:17
[v2] 2017-08-28 09:11:45

Unique-IP document downloads: 81 times is a pre-print repository rather than a journal. Articles hosted may not yet have been verified by peer-review and should be treated as preliminary. In particular, anything that appears to include financial or legal advice or proposed medical treatments should be treated with due caution. will not be responsible for any consequences of actions that result from any form of use of any documents on this website.

Add your own feedback and questions here:
You are equally welcome to be positive or negative about any paper but please be polite. If you are being critical you must mention at least one specific error, otherwise your comment will be deleted as unhelpful.

comments powered by Disqus