Evaluation of Oxygen Reduction System (ORS) in Large-scale Fire Tests

Overview

The study asks what the LOC, the parameter that sets the ORS design concentration, is at large scale for solid fuels in a rack-storage configuration. Two-tier rack arrays of standard commodities sat in an enclosure fed with nitrogen-diluted air at 9-17% O2 under a 5 MW calorimeter and were ignited with a 33 kW premixed propane burner representing oxygen-insensitive sources such as electrical shorts, arcs and hot work. Tested materials included Class 3, cartoned unexpanded plastic (CUP), cartoned expanded plastic (CEP) and uncartoned plastics. Earlier lab work found LOCs of 10-13% for cardboard, wood, PS and PE, below the 15-16% range from standard tests in VdS 3527 and EN 16750:2017.

Results: the fitted exponential fire growth rate b fell from 0.13 s^-1 at 17.4% O2 to 0.09 at 15.0%, 0.056 at 13.6%, 0.03 at 11.4% and near zero at 9.3%, where Class 3 commodity could not be ignited. Without the igniter, fire was self-sustaining at 15.0% and 17.4% O2 but not at 13.6% or 11.4%, a result specific to the array size and about 1 min ignition duration.

Logistic regression gave LOC at 0.05 propagation probability: cartoned commodities 11.1% (hard limit) and 13.8% (soft limit); uncartoned plastics 13.0% (hard) and 14.7% (soft). Hard limits approach fundamental gas-phase LOCs (11.1-12.0% for methane), do not depend much on configuration or ignition duration, and are recommended by the authors for ORS design. They are lower than the design concentrations in VdS 3527 and EN 16750.

Limits: soft limits depend on the specific array size and ignition duration, and the tests are limited to a two-tier rack arrangement and particular standard commodities. The results give evidence for the ORS design concentration, but generalising to other storage layouts needs further testing.

Key points

1. Earlier lab work found LOCs of 10-13% for cardboard, wood, PS and PE, below the 15-16% range from standard tests in VdS 3527 and EN 16750:2017.
2. Ignition used a 33 kW premixed propane flame to represent oxygen-insensitive sources such as electrical shorts, arcs and hot work; tested O2 ranged from 9% to 17% by volume.
3. Fitted exponential fire growth rate b fell from 0.13 s^-1 at 17.4% O2 to 0.09 at 15.0%, 0.056 at 13.6%, 0.03 at 11.4% and near zero at 9.3%, where Class 3 commodity could not be ignited.
4. Without the igniter, fire was self-sustaining at 15.0% and 17.4% O2 but not at 13.6% or 11.4%, specific to the array size and about 1 min ignition duration.
5. LOC at 0.05 propagation probability: cartoned commodities 11.1% (hard limit) / 13.8% (soft limit); uncartoned plastics 13.0% (hard) / 14.7% (soft).
6. Hard limits approach fundamental gas-phase LOCs (11.1-12.0% for methane), do not depend much on configuration or ignition duration, and are recommended for ORS design.

Using it in design

• Compare ORS design concentration with the 15-16% standard-test values in VdS 3527 and EN 16750; since large-scale hard-limit LOCs (11.1% for cartoned commodities, 13.0% for uncartoned plastics) are lower, justify the design safety margin against them.
• Knowing the ignition source was a 33 kW flame representing oxygen-insensitive electrical arcs and hot work, do not set a data-centre ORS design criterion on assumptions that only a heat-sensitive source exists.
• Do not use soft limits (igniter off) as a design input because they depend on array size and ignition duration; use hard limits.

Who should read it

Fire-safety engineers designing oxygen-reduction (hypoxic) fire prevention systems and evaluating design concentrations in standards.

Document contents

1. Introduction
2. Experimental Setup
3. Results and Discussions
4. Summary

Related standards

• EN 16750
• VdS 3527
• NFPA 75

Read the document

Publication details

Authors

Xiangyang Zhou; Yibing Xin; Sergey Dorofeev

Publisher

Journal of Physics: Conference Series (IOP Publishing)

Year

2018

Volume

1107

Pages

062005

Document language

English

Source type

Scientific

Access

Open licence — downloadable

DOI

10.1088/1742-6596/1107/6/062005

Licence

CC BY 3.0 — share with attribution

Licence evidence

iopscience.iop.org

Citation