<?xml version="1.0"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en">
	<id>https://fire-safety.academy/w/index.php?action=history&amp;feed=atom&amp;title=Data_centre_fire_protection</id>
	<title>Data centre fire protection - Revision history</title>
	<link rel="self" type="application/atom+xml" href="https://fire-safety.academy/w/index.php?action=history&amp;feed=atom&amp;title=Data_centre_fire_protection"/>
	<link rel="alternate" type="text/html" href="https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;action=history"/>
	<updated>2026-11-21T22:40:01Z</updated>
	<subtitle>Revision history for this page on the wiki</subtitle>
	<generator>MediaWiki 1.43.11</generator>
	<entry>
		<id>https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=6522&amp;oldid=prev</id>
		<title>Mark: Remove source references from image captions</title>
		<link rel="alternate" type="text/html" href="https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=6522&amp;oldid=prev"/>
		<updated>2026-11-13T07:24:29Z</updated>

		<summary type="html">&lt;p&gt;Remove source references from image captions&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;en&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Older revision&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revision as of 08:24, 13 November 2026&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l5&quot;&gt;Line 5:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 5:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The equipment is very sensitive to heat. Tests summarised in the handbook show that permanent damage can begin above 49 °C (120 °F), and that magnetic tapes and disks start to lose information at 52 °C (125 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; A sustained ambient temperature of 66 °C (150 °F) damages hard disks, and components begin to fail at 79 °C (174 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Major component failures occur between 300 and 500 °F (149 and 260 °C), and plastic tape cases degrade at 650 to 750 °F (343 to 399 °C), releasing flammable styrene monomer.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61-conv&amp;quot;&amp;gt;Rawson (2008), p. 8-61; Celsius values here converted from the printed Fahrenheit values.&amp;lt;/ref&amp;gt; These thresholds are far below the temperatures of a developed fire, so equipment can be lost without ever being touched by flame.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The equipment is very sensitive to heat. Tests summarised in the handbook show that permanent damage can begin above 49 °C (120 °F), and that magnetic tapes and disks start to lose information at 52 °C (125 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; A sustained ambient temperature of 66 °C (150 °F) damages hard disks, and components begin to fail at 79 °C (174 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Major component failures occur between 300 and 500 °F (149 and 260 °C), and plastic tape cases degrade at 650 to 750 °F (343 to 399 °C), releasing flammable styrene monomer.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61-conv&amp;quot;&amp;gt;Rawson (2008), p. 8-61; Celsius values here converted from the printed Fahrenheit values.&amp;lt;/ref&amp;gt; These thresholds are far below the temperatures of a developed fire, so equipment can be lost without ever being touched by flame.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:fse-data-centre-fire-protection-v2.svg|thumb|upright=1.5|alt=Temperature scale from 0 to 400 °C with markers where IT equipment, magnetic media, hard disks, microfilm, paper and tape cases are damaged|Temperatures at which damage to IT equipment and media begins. Even a small fire can heat a room past these values&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;. Based on Rawson (2008), p. 8-61&lt;/del&gt;.]]&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:fse-data-centre-fire-protection-v2.svg|thumb|upright=1.5|alt=Temperature scale from 0 to 400 °C with markers where IT equipment, magnetic media, hard disks, microfilm, paper and tape cases are damaged|Temperatures at which damage to IT equipment and media begins. Even a small fire can heat a room past these values.]]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Smoke causes a second kind of loss. Tewarson defines nonthermal damage as damage by smoke, toxic and corrosive products, such as corrosion, electrical malfunctions, discoloration and odours.&amp;lt;ref name=&amp;quot;Tewarson-3-133&amp;quot;&amp;gt;Tewarson (2002), p. 3-133.&amp;lt;/ref&amp;gt; Telephone central offices and computer rooms are among his examples of susceptible occupancies.&amp;lt;ref name=&amp;quot;Tewarson-3-136&amp;quot;&amp;gt;Tewarson (2002), p. 3-136.&amp;lt;/ref&amp;gt; Burning PVC wire insulation releases [[Hydrogen chloride|hydrogen chloride]],&amp;lt;ref name=&amp;quot;CusterMilke-2-76&amp;quot;&amp;gt;Custer and Milke (2008), p. 2-76.&amp;lt;/ref&amp;gt; and one handbook example sets a design criterion of no more than 5 ppm of it at sensitive equipment.&amp;lt;ref name=&amp;quot;Mowrer-3-186&amp;quot;&amp;gt;Mowrer and Rosenbaum (2008), p. 3-186.&amp;lt;/ref&amp;gt; The cable hazard itself is described under [[Cable fire]].&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Smoke causes a second kind of loss. Tewarson defines nonthermal damage as damage by smoke, toxic and corrosive products, such as corrosion, electrical malfunctions, discoloration and odours.&amp;lt;ref name=&amp;quot;Tewarson-3-133&amp;quot;&amp;gt;Tewarson (2002), p. 3-133.&amp;lt;/ref&amp;gt; Telephone central offices and computer rooms are among his examples of susceptible occupancies.&amp;lt;ref name=&amp;quot;Tewarson-3-136&amp;quot;&amp;gt;Tewarson (2002), p. 3-136.&amp;lt;/ref&amp;gt; Burning PVC wire insulation releases [[Hydrogen chloride|hydrogen chloride]],&amp;lt;ref name=&amp;quot;CusterMilke-2-76&amp;quot;&amp;gt;Custer and Milke (2008), p. 2-76.&amp;lt;/ref&amp;gt; and one handbook example sets a design criterion of no more than 5 ppm of it at sensitive equipment.&amp;lt;ref name=&amp;quot;Mowrer-3-186&amp;quot;&amp;gt;Mowrer and Rosenbaum (2008), p. 3-186.&amp;lt;/ref&amp;gt; The cable hazard itself is described under [[Cable fire]].&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l17&quot;&gt;Line 17:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 17:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;About two thirds of the fires, 626 of 944, stayed confined to the object of origin.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Even so, the average loss was about $14,600 per fire for those small fires and $29,400 for all fires, which shows how vulnerable the contents are.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Automatic suppression equipment was present in 379 of the fires; average loss per fire was about three quarters lower than in fires without it.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Smoke detection was present in three of five fires and was associated with a two-thirds lower average loss.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt; These comparisons describe the reported fires; they do not isolate the effect of either system from other differences between the buildings.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;About two thirds of the fires, 626 of 944, stayed confined to the object of origin.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Even so, the average loss was about $14,600 per fire for those small fires and $29,400 for all fires, which shows how vulnerable the contents are.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Automatic suppression equipment was present in 379 of the fires; average loss per fire was about three quarters lower than in fires without it.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Smoke detection was present in three of five fires and was associated with a two-thirds lower average loss.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt; These comparisons describe the reported fires; they do not isolate the effect of either system from other differences between the buildings.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:fse-data-centre-fire-protection-v3.svg|thumb|upright=1.5|alt=Bar chart of average loss per fire with and without automatic suppression and with and without detection in US electronic equipment room fires 1994 to 1998|Average direct loss per fire in US non-home electronic equipment room fires, 1994–1998 annual averages, with and without automatic suppression and smoke detection&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;. Based on Rawson (2008), pp. 8-62–8-63&lt;/del&gt;.]]&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:fse-data-centre-fire-protection-v3.svg|thumb|upright=1.5|alt=Bar chart of average loss per fire with and without automatic suppression and with and without detection in US electronic equipment room fires 1994 to 1998|Average direct loss per fire in US non-home electronic equipment room fires, 1994–1998 annual averages, with and without automatic suppression and smoke detection.]]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Location, construction and housekeeping ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Location, construction and housekeeping ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The handbook recommends that the IT room should ideally be in a separate, single-storey building of fire-resistive or noncombustible construction.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Otherwise it may be inside such a building if it is adequately separated, and offices, printers, paper stock and records should be separated from it by rated partitions.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; For ordinary hazard occupancies the partitions should have at least a 2-hour rating, with listed fire doors and fire dampers in openings and ducts.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; The room should not be located above, below or next to hazardous processes.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Such compartmentation protects high-value operations from a fire elsewhere in the building.&amp;lt;ref name=&amp;quot;Davoodi-18-12&amp;quot;&amp;gt;Davoodi (2008), p. 18-12.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The handbook recommends that the IT room should ideally be in a separate, single-storey building of fire-resistive or noncombustible construction.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Otherwise it may be inside such a building if it is adequately separated, and offices, printers, paper stock and records should be separated from it by rated partitions.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; For ordinary hazard occupancies the partitions should have at least a 2-hour rating, with listed fire doors and fire dampers in openings and ducts.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; The room should not be located above, below or next to hazardous processes.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Such compartmentation protects high-value operations from a fire elsewhere in the building.&amp;lt;ref name=&amp;quot;Davoodi-18-12&amp;quot;&amp;gt;Davoodi (2008), p. 18-12.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:fse-data-centre-fire-protection-v1.svg|thumb|upright=1.5|alt=Section of a data centre room with raised floor, racks, ceiling and underfloor detectors, sampling pipes, preaction sprinklers and agent nozzles|Typical fire protection layout of a data centre room: rated separation, detection at the ceiling and under the raised floor, preaction sprinklers and a total flooding gaseous agent with nozzles above and below the floor&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;. Based on Rawson (2008), pp. 8-59–8-63&lt;/del&gt;.]]&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:fse-data-centre-fire-protection-v1.svg|thumb|upright=1.5|alt=Section of a data centre room with raised floor, racks, ceiling and underfloor detectors, sampling pipes, preaction sprinklers and agent nozzles|Typical fire protection layout of a data centre room: rated separation, detection at the ceiling and under the raised floor, preaction sprinklers and a total flooding gaseous agent with nozzles above and below the floor.]]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Most mainframe installations stand on a raised access floor, which carries power and data cables and distributes cooling air.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; The floor and its supports should be noncombustible, cable openings should be kept small, and carpet should be avoided.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Raised floors can become accumulation points for trash and debris.&amp;lt;ref name=&amp;quot;Mattern-11-31&amp;quot;&amp;gt;Mattern (2008b), p. 11-31.&amp;lt;/ref&amp;gt; Abandoned cables add fuel in the form of insulation and connectors and block cooling airflow, so a cable management programme should remove them.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Most mainframe installations stand on a raised access floor, which carries power and data cables and distributes cooling air.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; The floor and its supports should be noncombustible, cable openings should be kept small, and carpet should be avoided.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Raised floors can become accumulation points for trash and debris.&amp;lt;ref name=&amp;quot;Mattern-11-31&amp;quot;&amp;gt;Mattern (2008b), p. 11-31.&amp;lt;/ref&amp;gt; Abandoned cables add fuel in the form of insulation and connectors and block cooling airflow, so a cable management programme should remove them.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l51&quot;&gt;Line 51:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 51:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The SFPE handbook applies this to overheated PVC cable in a 7 m high air-conditioned room with switch equipment and a temperature difference of 10 °C.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; For a critical fire of 1,000 W, the smoke rises only about 6 m, so ceiling detectors may not respond.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; Rearranging the same relation shows that about 1,450 W would be needed to reach 7 m. For a ceiling 3.05 m high, as in a data centre example in the NFPA handbook, about 180 W is enough at the same temperature difference.&amp;lt;ref name=&amp;quot;Harrington-17-134&amp;quot;&amp;gt;Harrington (2008), p. 17-134.&amp;lt;/ref&amp;gt; In low rooms, the dilution of smoke by forced ventilation with high exchange rates is therefore the larger concern.&amp;lt;ref name=&amp;quot;Schifiliti-4-26&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The SFPE handbook applies this to overheated PVC cable in a 7 m high air-conditioned room with switch equipment and a temperature difference of 10 °C.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; For a critical fire of 1,000 W, the smoke rises only about 6 m, so ceiling detectors may not respond.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; Rearranging the same relation shows that about 1,450 W would be needed to reach 7 m. For a ceiling 3.05 m high, as in a data centre example in the NFPA handbook, about 180 W is enough at the same temperature difference.&amp;lt;ref name=&amp;quot;Harrington-17-134&amp;quot;&amp;gt;Harrington (2008), p. 17-134.&amp;lt;/ref&amp;gt; In low rooms, the dilution of smoke by forced ventilation with high exchange rates is therefore the larger concern.&amp;lt;ref name=&amp;quot;Schifiliti-4-26&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:fse-data-centre-fire-protection-v4.svg|thumb|upright=1.5|alt=Line chart of maximum smoke rise height against convective heat release rate for temperature differences of 5, 10 and 20 °C|Maximum height to which smoke from a small fire rises in a room whose ceiling is warmer than the fire location, for three temperature differences&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;. Based on Schifiliti et al. (2002), p. 4-27&lt;/del&gt;.]]&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:fse-data-centre-fire-protection-v4.svg|thumb|upright=1.5|alt=Line chart of maximum smoke rise height against convective heat release rate for temperature differences of 5, 10 and 20 °C|Maximum height to which smoke from a small fire rises in a room whose ceiling is warmer than the fire location, for three temperature differences.]]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Sprinklers and water mist ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Sprinklers and water mist ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;

&lt;!-- diff cache key firesb_pjbsiy42-wiki_:diff:1.41:old-5839:rev-6522:php=table --&gt;
&lt;/table&gt;</summary>
		<author><name>Mark</name></author>
	</entry>
	<entry>
		<id>https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=5839&amp;oldid=prev</id>
		<title>Peter: Add SVG figures fse-data-centre-fire-protection-v1.svg, fse-data-centre-fire-protection-v2.svg, fse-data-centre-fire-protection-v3.svg, fse-data-centre-fire-protection-v4.svg (list-3 visual plan)</title>
		<link rel="alternate" type="text/html" href="https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=5839&amp;oldid=prev"/>
		<updated>2026-11-12T12:13:50Z</updated>

		<summary type="html">&lt;p&gt;Add SVG figures fse-data-centre-fire-protection-v1.svg, fse-data-centre-fire-protection-v2.svg, fse-data-centre-fire-protection-v3.svg, fse-data-centre-fire-protection-v4.svg (list-3 visual plan)&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;en&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Older revision&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revision as of 13:13, 12 November 2026&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l5&quot;&gt;Line 5:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 5:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The equipment is very sensitive to heat. Tests summarised in the handbook show that permanent damage can begin above 49 °C (120 °F), and that magnetic tapes and disks start to lose information at 52 °C (125 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; A sustained ambient temperature of 66 °C (150 °F) damages hard disks, and components begin to fail at 79 °C (174 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Major component failures occur between 300 and 500 °F (149 and 260 °C), and plastic tape cases degrade at 650 to 750 °F (343 to 399 °C), releasing flammable styrene monomer.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61-conv&amp;quot;&amp;gt;Rawson (2008), p. 8-61; Celsius values here converted from the printed Fahrenheit values.&amp;lt;/ref&amp;gt; These thresholds are far below the temperatures of a developed fire, so equipment can be lost without ever being touched by flame.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The equipment is very sensitive to heat. Tests summarised in the handbook show that permanent damage can begin above 49 °C (120 °F), and that magnetic tapes and disks start to lose information at 52 °C (125 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; A sustained ambient temperature of 66 °C (150 °F) damages hard disks, and components begin to fail at 79 °C (174 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Major component failures occur between 300 and 500 °F (149 and 260 °C), and plastic tape cases degrade at 650 to 750 °F (343 to 399 °C), releasing flammable styrene monomer.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61-conv&amp;quot;&amp;gt;Rawson (2008), p. 8-61; Celsius values here converted from the printed Fahrenheit values.&amp;lt;/ref&amp;gt; These thresholds are far below the temperatures of a developed fire, so equipment can be lost without ever being touched by flame.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:fse-data-centre-fire-protection-v2.svg|thumb|upright=1.5|alt=Temperature scale from 0 to 400 °C with markers where IT equipment, magnetic media, hard disks, microfilm, paper and tape cases are damaged|Temperatures at which damage to IT equipment and media begins. Even a small fire can heat a room past these values. Based on Rawson (2008), p. 8-61.]]&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Smoke causes a second kind of loss. Tewarson defines nonthermal damage as damage by smoke, toxic and corrosive products, such as corrosion, electrical malfunctions, discoloration and odours.&amp;lt;ref name=&amp;quot;Tewarson-3-133&amp;quot;&amp;gt;Tewarson (2002), p. 3-133.&amp;lt;/ref&amp;gt; Telephone central offices and computer rooms are among his examples of susceptible occupancies.&amp;lt;ref name=&amp;quot;Tewarson-3-136&amp;quot;&amp;gt;Tewarson (2002), p. 3-136.&amp;lt;/ref&amp;gt; Burning PVC wire insulation releases [[Hydrogen chloride|hydrogen chloride]],&amp;lt;ref name=&amp;quot;CusterMilke-2-76&amp;quot;&amp;gt;Custer and Milke (2008), p. 2-76.&amp;lt;/ref&amp;gt; and one handbook example sets a design criterion of no more than 5 ppm of it at sensitive equipment.&amp;lt;ref name=&amp;quot;Mowrer-3-186&amp;quot;&amp;gt;Mowrer and Rosenbaum (2008), p. 3-186.&amp;lt;/ref&amp;gt; The cable hazard itself is described under [[Cable fire]].&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Smoke causes a second kind of loss. Tewarson defines nonthermal damage as damage by smoke, toxic and corrosive products, such as corrosion, electrical malfunctions, discoloration and odours.&amp;lt;ref name=&amp;quot;Tewarson-3-133&amp;quot;&amp;gt;Tewarson (2002), p. 3-133.&amp;lt;/ref&amp;gt; Telephone central offices and computer rooms are among his examples of susceptible occupancies.&amp;lt;ref name=&amp;quot;Tewarson-3-136&amp;quot;&amp;gt;Tewarson (2002), p. 3-136.&amp;lt;/ref&amp;gt; Burning PVC wire insulation releases [[Hydrogen chloride|hydrogen chloride]],&amp;lt;ref name=&amp;quot;CusterMilke-2-76&amp;quot;&amp;gt;Custer and Milke (2008), p. 2-76.&amp;lt;/ref&amp;gt; and one handbook example sets a design criterion of no more than 5 ppm of it at sensitive equipment.&amp;lt;ref name=&amp;quot;Mowrer-3-186&amp;quot;&amp;gt;Mowrer and Rosenbaum (2008), p. 3-186.&amp;lt;/ref&amp;gt; The cable hazard itself is described under [[Cable fire]].&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l16&quot;&gt;Line 16:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 17:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;About two thirds of the fires, 626 of 944, stayed confined to the object of origin.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Even so, the average loss was about $14,600 per fire for those small fires and $29,400 for all fires, which shows how vulnerable the contents are.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Automatic suppression equipment was present in 379 of the fires; average loss per fire was about three quarters lower than in fires without it.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Smoke detection was present in three of five fires and was associated with a two-thirds lower average loss.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt; These comparisons describe the reported fires; they do not isolate the effect of either system from other differences between the buildings.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;About two thirds of the fires, 626 of 944, stayed confined to the object of origin.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Even so, the average loss was about $14,600 per fire for those small fires and $29,400 for all fires, which shows how vulnerable the contents are.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Automatic suppression equipment was present in 379 of the fires; average loss per fire was about three quarters lower than in fires without it.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Smoke detection was present in three of five fires and was associated with a two-thirds lower average loss.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt; These comparisons describe the reported fires; they do not isolate the effect of either system from other differences between the buildings.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:fse-data-centre-fire-protection-v3.svg|thumb|upright=1.5|alt=Bar chart of average loss per fire with and without automatic suppression and with and without detection in US electronic equipment room fires 1994 to 1998|Average direct loss per fire in US non-home electronic equipment room fires, 1994–1998 annual averages, with and without automatic suppression and smoke detection. Based on Rawson (2008), pp. 8-62–8-63.]]&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Location, construction and housekeeping ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Location, construction and housekeeping ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The handbook recommends that the IT room should ideally be in a separate, single-storey building of fire-resistive or noncombustible construction.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Otherwise it may be inside such a building if it is adequately separated, and offices, printers, paper stock and records should be separated from it by rated partitions.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; For ordinary hazard occupancies the partitions should have at least a 2-hour rating, with listed fire doors and fire dampers in openings and ducts.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; The room should not be located above, below or next to hazardous processes.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Such compartmentation protects high-value operations from a fire elsewhere in the building.&amp;lt;ref name=&amp;quot;Davoodi-18-12&amp;quot;&amp;gt;Davoodi (2008), p. 18-12.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The handbook recommends that the IT room should ideally be in a separate, single-storey building of fire-resistive or noncombustible construction.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Otherwise it may be inside such a building if it is adequately separated, and offices, printers, paper stock and records should be separated from it by rated partitions.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; For ordinary hazard occupancies the partitions should have at least a 2-hour rating, with listed fire doors and fire dampers in openings and ducts.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; The room should not be located above, below or next to hazardous processes.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Such compartmentation protects high-value operations from a fire elsewhere in the building.&amp;lt;ref name=&amp;quot;Davoodi-18-12&amp;quot;&amp;gt;Davoodi (2008), p. 18-12.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:fse-data-centre-fire-protection-v1.svg|thumb|upright=1.5|alt=Section of a data centre room with raised floor, racks, ceiling and underfloor detectors, sampling pipes, preaction sprinklers and agent nozzles|Typical fire protection layout of a data centre room: rated separation, detection at the ceiling and under the raised floor, preaction sprinklers and a total flooding gaseous agent with nozzles above and below the floor. Based on Rawson (2008), pp. 8-59–8-63.]]&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Most mainframe installations stand on a raised access floor, which carries power and data cables and distributes cooling air.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; The floor and its supports should be noncombustible, cable openings should be kept small, and carpet should be avoided.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Raised floors can become accumulation points for trash and debris.&amp;lt;ref name=&amp;quot;Mattern-11-31&amp;quot;&amp;gt;Mattern (2008b), p. 11-31.&amp;lt;/ref&amp;gt; Abandoned cables add fuel in the form of insulation and connectors and block cooling airflow, so a cable management programme should remove them.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Most mainframe installations stand on a raised access floor, which carries power and data cables and distributes cooling air.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; The floor and its supports should be noncombustible, cable openings should be kept small, and carpet should be avoided.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Raised floors can become accumulation points for trash and debris.&amp;lt;ref name=&amp;quot;Mattern-11-31&amp;quot;&amp;gt;Mattern (2008b), p. 11-31.&amp;lt;/ref&amp;gt; Abandoned cables add fuel in the form of insulation and connectors and block cooling airflow, so a cable management programme should remove them.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l48&quot;&gt;Line 48:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 51:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The SFPE handbook applies this to overheated PVC cable in a 7 m high air-conditioned room with switch equipment and a temperature difference of 10 °C.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; For a critical fire of 1,000 W, the smoke rises only about 6 m, so ceiling detectors may not respond.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; Rearranging the same relation shows that about 1,450 W would be needed to reach 7 m. For a ceiling 3.05 m high, as in a data centre example in the NFPA handbook, about 180 W is enough at the same temperature difference.&amp;lt;ref name=&amp;quot;Harrington-17-134&amp;quot;&amp;gt;Harrington (2008), p. 17-134.&amp;lt;/ref&amp;gt; In low rooms, the dilution of smoke by forced ventilation with high exchange rates is therefore the larger concern.&amp;lt;ref name=&amp;quot;Schifiliti-4-26&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The SFPE handbook applies this to overheated PVC cable in a 7 m high air-conditioned room with switch equipment and a temperature difference of 10 °C.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; For a critical fire of 1,000 W, the smoke rises only about 6 m, so ceiling detectors may not respond.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; Rearranging the same relation shows that about 1,450 W would be needed to reach 7 m. For a ceiling 3.05 m high, as in a data centre example in the NFPA handbook, about 180 W is enough at the same temperature difference.&amp;lt;ref name=&amp;quot;Harrington-17-134&amp;quot;&amp;gt;Harrington (2008), p. 17-134.&amp;lt;/ref&amp;gt; In low rooms, the dilution of smoke by forced ventilation with high exchange rates is therefore the larger concern.&amp;lt;ref name=&amp;quot;Schifiliti-4-26&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:fse-data-centre-fire-protection-v4.svg|thumb|upright=1.5|alt=Line chart of maximum smoke rise height against convective heat release rate for temperature differences of 5, 10 and 20 °C|Maximum height to which smoke from a small fire rises in a room whose ceiling is warmer than the fire location, for three temperature differences. Based on Schifiliti et al. (2002), p. 4-27.]]&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Sprinklers and water mist ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Sprinklers and water mist ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;

&lt;!-- diff cache key firesb_pjbsiy42-wiki_:diff:1.41:old-5229:rev-5839:php=table --&gt;
&lt;/table&gt;</summary>
		<author><name>Peter</name></author>
	</entry>
	<entry>
		<id>https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=5229&amp;oldid=prev</id>
		<title>Linda: Add watermarked Fire Safety Academy illustration at the location specified in the visual plan</title>
		<link rel="alternate" type="text/html" href="https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=5229&amp;oldid=prev"/>
		<updated>2026-11-08T07:52:46Z</updated>

		<summary type="html">&lt;p&gt;Add watermarked Fire Safety Academy illustration at the location specified in the visual plan&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;en&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Older revision&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revision as of 08:52, 8 November 2026&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l21&quot;&gt;Line 21:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 21:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Most mainframe installations stand on a raised access floor, which carries power and data cables and distributes cooling air.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; The floor and its supports should be noncombustible, cable openings should be kept small, and carpet should be avoided.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Raised floors can become accumulation points for trash and debris.&amp;lt;ref name=&amp;quot;Mattern-11-31&amp;quot;&amp;gt;Mattern (2008b), p. 11-31.&amp;lt;/ref&amp;gt; Abandoned cables add fuel in the form of insulation and connectors and block cooling airflow, so a cable management programme should remove them.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Most mainframe installations stand on a raised access floor, which carries power and data cables and distributes cooling air.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; The floor and its supports should be noncombustible, cable openings should be kept small, and carpet should be avoided.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Raised floors can become accumulation points for trash and debris.&amp;lt;ref name=&amp;quot;Mattern-11-31&amp;quot;&amp;gt;Mattern (2008b), p. 11-31.&amp;lt;/ref&amp;gt; Abandoned cables add fuel in the form of insulation and connectors and block cooling airflow, so a cable management programme should remove them.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:fse-data-centre-fire-protection-v5.jpg|thumb|upright=1.5|alt=a data centre raised floor with several floor tiles lifted, showing cables and pedestals in the void below|Raised access floor with tiles lifted. The void carries power and data cables and cooling air, and needs its own detection and housekeeping.]]&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;In New Zealand, Verification Method C/VM2 (second edition, 2025) names floor plenums for IT cables as an example of concealed spaces.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot;&amp;gt;Ministry of Business, Innovation and Employment (2025b), p. 41.&amp;lt;/ref&amp;gt; Its concealed space scenario applies to buildings with rooms holding more than 50 people.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt; The scenario is deemed satisfied by separating elements, automatic detection or suppression.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;In New Zealand, Verification Method C/VM2 (second edition, 2025) names floor plenums for IT cables as an example of concealed spaces.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot;&amp;gt;Ministry of Business, Innovation and Employment (2025b), p. 41.&amp;lt;/ref&amp;gt; Its concealed space scenario applies to buildings with rooms holding more than 50 people.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt; The scenario is deemed satisfied by separating elements, automatic detection or suppression.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;

&lt;!-- diff cache key firesb_pjbsiy42-wiki_:diff:1.41:old-4251:rev-5229:php=table --&gt;
&lt;/table&gt;</summary>
		<author><name>Linda</name></author>
	</entry>
	<entry>
		<id>https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=4251&amp;oldid=prev</id>
		<title>Mark: FSE wiki: Fact-check: qualify observational loss comparisons and 2008 clean-agent exposure conditions (written by Claude Opus 5.5)</title>
		<link rel="alternate" type="text/html" href="https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=4251&amp;oldid=prev"/>
		<updated>2026-11-08T01:02:22Z</updated>

		<summary type="html">&lt;p&gt;FSE wiki: Fact-check: qualify observational loss comparisons and 2008 clean-agent exposure conditions (written by Claude Opus 5.5)&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;en&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Older revision&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revision as of 02:02, 8 November 2026&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l1&quot;&gt;Line 1:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 1:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&#039;&#039;&#039;Data centre fire protection&#039;&#039;&#039; is the combination of construction, detection, suppression and emergency planning used to protect rooms and buildings that house information technology (IT) equipment, such as servers, storage systems and network switches. The 20th edition of the NFPA &#039;&#039;Fire Protection Handbook&#039;&#039; (2008) treats these spaces as IT equipment rooms, the subject of NFPA 75 in the United States.&amp;lt;ref name=&quot;Rawson-8-57&quot;&amp;gt;Rawson (2008), p. 8-57.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Harrington-17-125&quot;&amp;gt;Harrington (2008), p. 17-125.&amp;lt;/ref&amp;gt; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;The main goal is usually not &lt;/del&gt;life safety &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;but &lt;/del&gt;continuity: the stored information is often worth more than the hardware, and a short outage can harm a business and its customers.&amp;lt;ref name=&quot;Rawson-8-60&quot;&amp;gt;Rawson (2008), p. 8-60.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Harrington-17-135&quot;&amp;gt;Harrington (2008), p. 17-135.&amp;lt;/ref&amp;gt; Electronic equipment is damaged by heat at temperatures far below those that harm buildings, and also by smoke, corrosive gases and water.&amp;lt;ref name=&quot;Rawson-8-61&quot;&amp;gt;Rawson (2008), p. 8-61.&amp;lt;/ref&amp;gt; Typical protection therefore combines fire-rated separation, very early smoke detection at the ceiling and under the raised floor, sprinklers and often a total flooding gaseous agent.&amp;lt;ref name=&quot;Rawson-8-59&quot;&amp;gt;Rawson (2008), p. 8-59.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Rawson-8-63&quot;&amp;gt;Rawson (2008), p. 8-63.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Rawson-8-62&quot;&amp;gt;Rawson (2008), p. 8-62.&amp;lt;/ref&amp;gt; In US fires in electronic equipment rooms in 1994–1998, &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;automatic suppression reduced &lt;/del&gt;the average loss per fire &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;by &lt;/del&gt;about three quarters.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&#039;&#039;&#039;Data centre fire protection&#039;&#039;&#039; is the combination of construction, detection, suppression and emergency planning used to protect rooms and buildings that house information technology (IT) equipment, such as servers, storage systems and network switches. The 20th edition of the NFPA &#039;&#039;Fire Protection Handbook&#039;&#039; (2008) treats these spaces as IT equipment rooms, the subject of NFPA 75 in the United States.&amp;lt;ref name=&quot;Rawson-8-57&quot;&amp;gt;Rawson (2008), p. 8-57.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Harrington-17-125&quot;&amp;gt;Harrington (2008), p. 17-125.&amp;lt;/ref&amp;gt; &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Alongside &lt;/ins&gt;life safety&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;, business &lt;/ins&gt;continuity &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;is a central concern&lt;/ins&gt;: the stored information is often worth more than the hardware, and a short outage can harm a business and its customers.&amp;lt;ref name=&quot;Rawson-8-60&quot;&amp;gt;Rawson (2008), p. 8-60.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Harrington-17-135&quot;&amp;gt;Harrington (2008), p. 17-135.&amp;lt;/ref&amp;gt; Electronic equipment is damaged by heat at temperatures far below those that harm buildings, and also by smoke, corrosive gases and water.&amp;lt;ref name=&quot;Rawson-8-61&quot;&amp;gt;Rawson (2008), p. 8-61.&amp;lt;/ref&amp;gt; Typical protection therefore combines fire-rated separation, very early smoke detection at the ceiling and under the raised floor, sprinklers and often a total flooding gaseous agent.&amp;lt;ref name=&quot;Rawson-8-59&quot;&amp;gt;Rawson (2008), p. 8-59.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Rawson-8-63&quot;&amp;gt;Rawson (2008), p. 8-63.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Rawson-8-62&quot;&amp;gt;Rawson (2008), p. 8-62.&amp;lt;/ref&amp;gt; In US fires in electronic equipment rooms in 1994–1998, the average loss per fire &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;was &lt;/ins&gt;about three quarters &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;lower where automatic suppression equipment was present than where it was absent&lt;/ins&gt;.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Hazards and consequences ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Hazards and consequences ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l15&quot;&gt;Line 15:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 15:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;For 1994–1998 the handbook gives an annual average of 944 such fires, with no civilian deaths, 13 injuries and $27.71 million in direct damage.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; In 133 of these fires no equipment was involved in ignition, in 120 the electronic equipment itself was involved, and fuses, circuit breakers and fixed wiring each accounted for more than 90.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; Most fires that damage IT rooms start in ordinary equipment in or next to the room rather than in the IT equipment.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;For 1994–1998 the handbook gives an annual average of 944 such fires, with no civilian deaths, 13 injuries and $27.71 million in direct damage.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; In 133 of these fires no equipment was involved in ignition, in 120 the electronic equipment itself was involved, and fuses, circuit breakers and fixed wiring each accounted for more than 90.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; Most fires that damage IT rooms start in ordinary equipment in or next to the room rather than in the IT equipment.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;About two thirds of the fires, 626 of 944, stayed confined to the object of origin.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt; Even so, the average loss was about $14,600 per fire for those small fires and $29,400 for all fires, which shows how vulnerable the contents are.&amp;lt;ref name=&quot;Rawson-8-61&quot; /&amp;gt;&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt; Automatic suppression equipment was present in &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;only &lt;/del&gt;379 of the fires&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;, but where present it cut the &lt;/del&gt;average loss per fire &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;by &lt;/del&gt;about three quarters.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt; Smoke detection was present in three of five fires and was associated with a two-thirds &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;reduction in &lt;/del&gt;average loss.&amp;lt;ref name=&quot;Rawson-8-63&quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;About two thirds of the fires, 626 of 944, stayed confined to the object of origin.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt; Even so, the average loss was about $14,600 per fire for those small fires and $29,400 for all fires, which shows how vulnerable the contents are.&amp;lt;ref name=&quot;Rawson-8-61&quot; /&amp;gt;&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt; Automatic suppression equipment was present in 379 of the fires&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;; &lt;/ins&gt;average loss per fire &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;was &lt;/ins&gt;about three quarters &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;lower than in fires without it&lt;/ins&gt;.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt; Smoke detection was present in three of five fires and was associated with a two-thirds &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;lower &lt;/ins&gt;average loss.&amp;lt;ref name=&quot;Rawson-8-63&quot; /&amp;gt; &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;These comparisons describe the reported fires; they do not isolate the effect of either system from other differences between the buildings.&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Location, construction and housekeeping ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Location, construction and housekeeping ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l24&quot;&gt;Line 24:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 24:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;In New Zealand, Verification Method C/VM2 (second edition, 2025) names floor plenums for IT cables as an example of concealed spaces.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot;&amp;gt;Ministry of Business, Innovation and Employment (2025b), p. 41.&amp;lt;/ref&amp;gt; Its concealed space scenario applies to buildings with rooms holding more than 50 people.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt; The scenario is deemed satisfied by separating elements, automatic detection or suppression.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;In New Zealand, Verification Method C/VM2 (second edition, 2025) names floor plenums for IT cables as an example of concealed spaces.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot;&amp;gt;Ministry of Business, Innovation and Employment (2025b), p. 41.&amp;lt;/ref&amp;gt; Its concealed space scenario applies to buildings with rooms holding more than 50 people.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt; The scenario is deemed satisfied by separating elements, automatic detection or suppression.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Equipment and services need the same attention. IT equipment should meet UL 478 or UL 1950, and automated tape libraries may hold enough plastic to need their own detection and protection.&amp;lt;ref name=&quot;Rawson-8-61&quot; /&amp;gt; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Each &lt;/del&gt;computer room &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;should &lt;/del&gt;have its own air-conditioning system with noncombustible ducts, able to switch to 100% outside air and exhaust when smoke is detected.&amp;lt;ref name=&quot;Rawson-8-61&quot; /&amp;gt; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Large &lt;/del&gt;battery banks &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;should be &lt;/del&gt;in a separate ventilated room, and standby generators in separate machinery rooms or outdoors with at least one day of fuel.&amp;lt;ref name=&quot;Rawson-8-61&quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Equipment and services need the same attention. &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;The 2008 handbook says that &lt;/ins&gt;IT equipment should meet UL 478 or UL 1950, and &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;that &lt;/ins&gt;automated tape libraries may hold enough plastic to need their own detection and protection.&amp;lt;ref name=&quot;Rawson-8-61&quot; /&amp;gt; &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;It calls for each &lt;/ins&gt;computer room &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;to &lt;/ins&gt;have its own air-conditioning system with noncombustible ducts, able to switch to 100% outside air and exhaust when smoke is detected.&amp;lt;ref name=&quot;Rawson-8-61&quot; /&amp;gt; &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;It places large &lt;/ins&gt;battery banks in a separate ventilated room, and standby generators in separate machinery rooms or outdoors with at least one day of fuel.&amp;lt;ref name=&quot;Rawson-8-61&quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Detection ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Detection ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l55&quot;&gt;Line 55:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 55:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Gaseous agents ==&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;== Gaseous agents ==&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;A [[Gaseous extinguishing system|gaseous extinguishing system]] is used where the value of the equipment or the risk of interruption justifies it, especially to protect data in process and allow a rapid return to operation.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt; From the 1960s, Halon 1301 total flooding became the usual agent for computer rooms.&amp;lt;ref name=&quot;DiNenno-17-94&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-94.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Harrington-17-123&quot;&amp;gt;Harrington (2008), p. 17-123.&amp;lt;/ref&amp;gt; Since the Montreal Protocol it is no longer made in most industrialised countries, and the US Clean Air Act banned new production after 1 January 1994.&amp;lt;ref name=&quot;Harrington-17-128&quot;&amp;gt;Harrington (2008), p. 17-128.&amp;lt;/ref&amp;gt; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;NFPA 2001 &lt;/del&gt;lists nine [[Halocarbon clean agent|halocarbon clean agents]] and four [[Inert gas agent|inert gas agents]] as replacements.&amp;lt;ref name=&quot;Harrington-17-123&quot; /&amp;gt; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;Carbon &lt;/del&gt;dioxide &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;to &lt;/del&gt;NFPA 12 &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;remains an option&lt;/del&gt;, but at about 9% most people become unconscious and at 25 to 30% breathing stops.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt;&amp;lt;ref name=&quot;Harrington-17-134&quot; /&amp;gt; Existing Halon 1301 systems should be tested and maintained to NFPA 12A.&amp;lt;ref name=&quot;Rawson-8-63&quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;A [[Gaseous extinguishing system|gaseous extinguishing system]] is used where the value of the equipment or the risk of interruption justifies it, especially to protect data in process and allow a rapid return to operation.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt; From the 1960s, Halon 1301 total flooding became the usual agent for computer rooms.&amp;lt;ref name=&quot;DiNenno-17-94&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-94.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Harrington-17-123&quot;&amp;gt;Harrington (2008), p. 17-123.&amp;lt;/ref&amp;gt; Since the Montreal Protocol it is no longer made in most industrialised countries, and the US Clean Air Act banned new production after 1 January 1994.&amp;lt;ref name=&quot;Harrington-17-128&quot;&amp;gt;Harrington (2008), p. 17-128.&amp;lt;/ref&amp;gt; &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;The 2008 handbook &lt;/ins&gt;lists nine [[Halocarbon clean agent|halocarbon clean agents]] and four [[Inert gas agent|inert gas agents]] &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;under NFPA 2001 &lt;/ins&gt;as replacements.&amp;lt;ref name=&quot;Harrington-17-123&quot; /&amp;gt; &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;It also discusses carbon &lt;/ins&gt;dioxide &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;systems under &lt;/ins&gt;NFPA 12, but at about 9% most people become unconscious and at 25 to 30% breathing stops.&amp;lt;ref name=&quot;Rawson-8-62&quot; /&amp;gt;&amp;lt;ref name=&quot;Harrington-17-134&quot; /&amp;gt; Existing Halon 1301 systems should be tested and maintained to NFPA 12A.&amp;lt;ref name=&quot;Rawson-8-63&quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Clean agents are electrically nonconductive, leave no residue and can reach fuel inside closed cabinets.&amp;lt;ref name=&amp;quot;Harrington-17-128&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot; /&amp;gt; If power is cut before discharge, a fire in cable insulation or circuit boards becomes a Class A surface fire, which is relatively easy to extinguish.&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot; /&amp;gt; If power stays on, it remains a Class C fire that is harder to extinguish and can rekindle.&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot; /&amp;gt; The handbook therefore recommends disconnecting power on discharge unless a risk assessment shows less damage with power maintained.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Clean agents are electrically nonconductive, leave no residue and can reach fuel inside closed cabinets.&amp;lt;ref name=&amp;quot;Harrington-17-128&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot; /&amp;gt; If power is cut before discharge, a fire in cable insulation or circuit boards becomes a Class A surface fire, which is relatively easy to extinguish.&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot; /&amp;gt; If power stays on, it remains a Class C fire that is harder to extinguish and can rekindle.&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot; /&amp;gt; The handbook therefore recommends disconnecting power on discharge unless a risk assessment shows less damage with power maintained.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l83&quot;&gt;Line 83:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 83:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;|}&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;|}&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;As a worked example, take the data centre room of the handbook&#039;s performance-based example: 1,858 m² with a ceiling 3.05 m high.&amp;lt;ref name=&quot;Harrington-17-134&quot; /&amp;gt; Ignoring the floor void and equipment, the net volume is about 5,660 m³. For HFC-227ea at 20 °C, &#039;&#039;S&#039;&#039; = 0.1268 + 0.0005133 × 20 = 0.137 m³/kg.&amp;lt;ref name=&quot;DiNenno-17-114&quot; /&amp;gt; At the ISO Class A design concentration of 7.9%, the mass is 5,660 / 0.137 × 7.9 / 92.1 ≈ 3,550 kg.&amp;lt;ref name=&quot;DiNenno-17-98&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-98&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;.&amp;lt;/ref&amp;gt; This concentration lies below the agent&#039;s no observed adverse effect level of 9.0%, so the space may stay normally occupied for up to 5 minutes; see [[NOAEL and LOAEL]].&amp;lt;ref name=&quot;Harrington-17-132&quot;&amp;gt;Harrington (2008), p. 17-132.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&quot;Harrington-17-133&quot;&amp;gt;Harrington (2008), p. 17-133.&amp;lt;/ref&amp;gt; With IG-541 at its minimum Class A design concentration of 37.5%, about 2,680 m³ of gas at 21 °C is needed instead.&amp;lt;ref name=&quot;DiNenno-17-112&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-112&lt;/del&gt;.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;As a worked example, take the data centre room of the handbook&#039;s performance-based example: 1,858 m² with a ceiling 3.05 m high.&amp;lt;ref name=&quot;Harrington-17-134&quot; /&amp;gt; Ignoring the floor void and equipment, the net volume is about 5,660 m³. For HFC-227ea at 20 °C, &#039;&#039;S&#039;&#039; = 0.1268 + 0.0005133 × 20 = 0.137 m³/kg.&amp;lt;ref name=&quot;DiNenno-17-114&quot; /&amp;gt; At the ISO Class A design concentration of 7.9%, the mass is 5,660 / 0.137 × 7.9 / 92.1 ≈ 3,550 kg.&amp;lt;ref name=&quot;DiNenno-17-98&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-98.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Halocarbon agents must discharge within 10 seconds, mainly to limit the formation of decomposition products.&amp;lt;ref name=&quot;DiNenno-17-114&quot; /&amp;gt; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;The &lt;/del&gt;replacement agents &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;form &lt;/del&gt;three to eight times more [[Hydrogen fluoride|hydrogen fluoride]] than Halon 1301, &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;and the main variable is the &lt;/del&gt;fire size at discharge.&amp;lt;ref name=&quot;DiNenno-17-114&quot; /&amp;gt; With automatic detection and actuation, full-scale tests found decomposition levels acceptable in typical computer and electronics spaces.&amp;lt;ref name=&quot;DiNenno-17-116&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-116.&amp;lt;/ref&amp;gt; Systems traditionally had to hold their concentration for 10 to 20 minutes.&amp;lt;ref name=&quot;DiNenno-17-119&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-119.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;This is below the agent&#039;s 9.0% no observed adverse effect level.&amp;lt;ref name=&quot;Harrington-17-132&quot;&amp;gt;Harrington (2008), p. 17-132.&amp;lt;/ref&amp;gt; The handbook&#039;s 2008 table permits a system in a normally occupied space if accidental agent exposure is limited to five minutes; see [[NOAEL and LOAEL]].&amp;lt;ref name=&quot;Harrington-17-133&quot;&amp;gt;Harrington (2008), p. 17-133.&amp;lt;/ref&amp;gt; With IG-541 at its minimum Class A design concentration of 37.5%, about 2,680 m³ of gas at 21 °C is needed instead.&amp;lt;ref name=&quot;DiNenno-17-112&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-112.&amp;lt;/ref&amp;gt;&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt; &lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-deleted&quot;&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Halocarbon agents must discharge within 10 seconds, mainly to limit the formation of decomposition products.&amp;lt;ref name=&quot;DiNenno-17-114&quot; /&amp;gt; &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;In the handbook&#039;s n-heptane liquid-fire tests, the &lt;/ins&gt;replacement agents &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;formed &lt;/ins&gt;three to eight times more [[Hydrogen fluoride|hydrogen fluoride]] than Halon 1301, &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;with &lt;/ins&gt;fire size at discharge &lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;the main variable&lt;/ins&gt;.&amp;lt;ref name=&quot;DiNenno-17-114&quot; /&amp;gt; With automatic detection and actuation, full-scale tests found decomposition levels acceptable in typical computer and electronics spaces.&amp;lt;ref name=&quot;DiNenno-17-116&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-116.&amp;lt;/ref&amp;gt; Systems traditionally had to hold their concentration for 10 to 20 minutes.&amp;lt;ref name=&quot;DiNenno-17-119&quot;&amp;gt;DiNenno and Taylor (2008), p. 17-119.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The enclosure must also survive the discharge. Halocarbon agents first lower and then raise the room pressure, with peaks of about −383 to 383 Pa measured for HFC-227ea.&amp;lt;ref name=&amp;quot;Harrington-17-126&amp;quot;&amp;gt;Harrington (2008), p. 17-126.&amp;lt;/ref&amp;gt; Inert gases and carbon dioxide create a positive pressure that normally requires relief vents.&amp;lt;ref name=&amp;quot;Harrington-17-126&amp;quot; /&amp;gt; Typical metal stud walls resist over 1,197 Pa, but very tight rooms and rooms over 3.7 m high need care; see [[Enclosure integrity test]].&amp;lt;ref name=&amp;quot;Harrington-17-127&amp;quot;&amp;gt;Harrington (2008), p. 17-127.&amp;lt;/ref&amp;gt; Reliability also matters: for a halon system in a computer facility, the mean probability of failing to protect against a waste-paper fire was estimated at 0.13.&amp;lt;ref name=&amp;quot;Steciak-asCited&amp;quot;&amp;gt;Steciak and Zalosh (1992), as cited in National Research Council of Canada et al. (2005), p. 3.7-3.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The enclosure must also survive the discharge. Halocarbon agents first lower and then raise the room pressure, with peaks of about −383 to 383 Pa measured for HFC-227ea.&amp;lt;ref name=&amp;quot;Harrington-17-126&amp;quot;&amp;gt;Harrington (2008), p. 17-126.&amp;lt;/ref&amp;gt; Inert gases and carbon dioxide create a positive pressure that normally requires relief vents.&amp;lt;ref name=&amp;quot;Harrington-17-126&amp;quot; /&amp;gt; Typical metal stud walls resist over 1,197 Pa, but very tight rooms and rooms over 3.7 m high need care; see [[Enclosure integrity test]].&amp;lt;ref name=&amp;quot;Harrington-17-127&amp;quot;&amp;gt;Harrington (2008), p. 17-127.&amp;lt;/ref&amp;gt; Reliability also matters: for a halon system in a computer facility, the mean probability of failing to protect against a waste-paper fire was estimated at 0.13.&amp;lt;ref name=&amp;quot;Steciak-asCited&amp;quot;&amp;gt;Steciak and Zalosh (1992), as cited in National Research Council of Canada et al. (2005), p. 3.7-3.&amp;lt;/ref&amp;gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;

&lt;!-- diff cache key firesb_pjbsiy42-wiki_:diff:1.41:old-3417:rev-4251:php=table --&gt;
&lt;/table&gt;</summary>
		<author><name>Mark</name></author>
	</entry>
	<entry>
		<id>https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=3417&amp;oldid=prev</id>
		<title>Mark: New FSE wiki article: Data centre fire protection (written by Claude Opus 5.5)</title>
		<link rel="alternate" type="text/html" href="https://fire-safety.academy/w/index.php?title=Data_centre_fire_protection&amp;diff=3417&amp;oldid=prev"/>
		<updated>2026-11-07T21:25:17Z</updated>

		<summary type="html">&lt;p&gt;New FSE wiki article: Data centre fire protection (written by Claude Opus 5.5)&lt;/p&gt;
&lt;p&gt;&lt;b&gt;New page&lt;/b&gt;&lt;/p&gt;&lt;div&gt;&amp;#039;&amp;#039;&amp;#039;Data centre fire protection&amp;#039;&amp;#039;&amp;#039; is the combination of construction, detection, suppression and emergency planning used to protect rooms and buildings that house information technology (IT) equipment, such as servers, storage systems and network switches. The 20th edition of the NFPA &amp;#039;&amp;#039;Fire Protection Handbook&amp;#039;&amp;#039; (2008) treats these spaces as IT equipment rooms, the subject of NFPA 75 in the United States.&amp;lt;ref name=&amp;quot;Rawson-8-57&amp;quot;&amp;gt;Rawson (2008), p. 8-57.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot;&amp;gt;Harrington (2008), p. 17-125.&amp;lt;/ref&amp;gt; The main goal is usually not life safety but continuity: the stored information is often worth more than the hardware, and a short outage can harm a business and its customers.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot;&amp;gt;Rawson (2008), p. 8-60.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Harrington-17-135&amp;quot;&amp;gt;Harrington (2008), p. 17-135.&amp;lt;/ref&amp;gt; Electronic equipment is damaged by heat at temperatures far below those that harm buildings, and also by smoke, corrosive gases and water.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot;&amp;gt;Rawson (2008), p. 8-61.&amp;lt;/ref&amp;gt; Typical protection therefore combines fire-rated separation, very early smoke detection at the ceiling and under the raised floor, sprinklers and often a total flooding gaseous agent.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot;&amp;gt;Rawson (2008), p. 8-59.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot;&amp;gt;Rawson (2008), p. 8-63.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot;&amp;gt;Rawson (2008), p. 8-62.&amp;lt;/ref&amp;gt; In US fires in electronic equipment rooms in 1994–1998, automatic suppression reduced the average loss per fire by about three quarters.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== Hazards and consequences ==&lt;br /&gt;
An IT centre contains more than computers. It includes blade servers, storage devices, tape libraries, routers and switches, together with air-conditioning units, power distribution units and an uninterruptible power supply (UPS).&amp;lt;ref name=&amp;quot;Rawson-8-57&amp;quot; /&amp;gt; Larger centres add battery rooms, diesel generators, transformers and switchgear in adjacent rooms.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; The handbook notes that IT centres can also be attractive targets for arson and sabotage, because so much depends on them.&amp;lt;ref name=&amp;quot;Rawson-8-58&amp;quot;&amp;gt;Rawson (2008), p. 8-58.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
The equipment is very sensitive to heat. Tests summarised in the handbook show that permanent damage can begin above 49 °C (120 °F), and that magnetic tapes and disks start to lose information at 52 °C (125 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; A sustained ambient temperature of 66 °C (150 °F) damages hard disks, and components begin to fail at 79 °C (174 °F).&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Major component failures occur between 300 and 500 °F (149 and 260 °C), and plastic tape cases degrade at 650 to 750 °F (343 to 399 °C), releasing flammable styrene monomer.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61-conv&amp;quot;&amp;gt;Rawson (2008), p. 8-61; Celsius values here converted from the printed Fahrenheit values.&amp;lt;/ref&amp;gt; These thresholds are far below the temperatures of a developed fire, so equipment can be lost without ever being touched by flame.&lt;br /&gt;
&lt;br /&gt;
Smoke causes a second kind of loss. Tewarson defines nonthermal damage as damage by smoke, toxic and corrosive products, such as corrosion, electrical malfunctions, discoloration and odours.&amp;lt;ref name=&amp;quot;Tewarson-3-133&amp;quot;&amp;gt;Tewarson (2002), p. 3-133.&amp;lt;/ref&amp;gt; Telephone central offices and computer rooms are among his examples of susceptible occupancies.&amp;lt;ref name=&amp;quot;Tewarson-3-136&amp;quot;&amp;gt;Tewarson (2002), p. 3-136.&amp;lt;/ref&amp;gt; Burning PVC wire insulation releases [[Hydrogen chloride|hydrogen chloride]],&amp;lt;ref name=&amp;quot;CusterMilke-2-76&amp;quot;&amp;gt;Custer and Milke (2008), p. 2-76.&amp;lt;/ref&amp;gt; and one handbook example sets a design criterion of no more than 5 ppm of it at sensitive equipment.&amp;lt;ref name=&amp;quot;Mowrer-3-186&amp;quot;&amp;gt;Mowrer and Rosenbaum (2008), p. 3-186.&amp;lt;/ref&amp;gt; The cable hazard itself is described under [[Cable fire]].&lt;br /&gt;
&lt;br /&gt;
The consequences are mainly economic. In most cases the stored information is more valuable than the IT hardware, so backup copies kept in a safe or remote location are an essential part of protection.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; The highest level of backup is a fully redundant remote site with mirrored software and real-time data, which is not economically feasible for every owner.&amp;lt;ref name=&amp;quot;Rawson-8-58&amp;quot; /&amp;gt; Alternatives include contracts with a commercial &amp;quot;hot site&amp;quot;, a fully equipped data centre ready to receive the client&amp;#039;s software and data.&amp;lt;ref name=&amp;quot;Schmidt-1-149&amp;quot;&amp;gt;Schmidt (2008), p. 1-149.&amp;lt;/ref&amp;gt; Such planning belongs to [[Business continuity|business continuity]] management.&lt;br /&gt;
&lt;br /&gt;
== Fire record ==&lt;br /&gt;
In the United States, fires in electronic equipment rooms of non-home structures fell from about 1,400 a year in 1980–1982 to about 900 a year in the late 1990s.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; From 1999 the national fire data used a narrower code, computer room or control centre, and about 300 fires a year were reported in 2001–2003.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Part of this decline may therefore come from the change in coding rather than from fewer fires.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
For 1994–1998 the handbook gives an annual average of 944 such fires, with no civilian deaths, 13 injuries and $27.71 million in direct damage.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; In 133 of these fires no equipment was involved in ignition, in 120 the electronic equipment itself was involved, and fuses, circuit breakers and fixed wiring each accounted for more than 90.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; Most fires that damage IT rooms start in ordinary equipment in or next to the room rather than in the IT equipment.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
About two thirds of the fires, 626 of 944, stayed confined to the object of origin.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Even so, the average loss was about $14,600 per fire for those small fires and $29,400 for all fires, which shows how vulnerable the contents are.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Automatic suppression equipment was present in only 379 of the fires, but where present it cut the average loss per fire by about three quarters.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; Smoke detection was present in three of five fires and was associated with a two-thirds reduction in average loss.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== Location, construction and housekeeping ==&lt;br /&gt;
The handbook recommends that the IT room should ideally be in a separate, single-storey building of fire-resistive or noncombustible construction.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Otherwise it may be inside such a building if it is adequately separated, and offices, printers, paper stock and records should be separated from it by rated partitions.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; For ordinary hazard occupancies the partitions should have at least a 2-hour rating, with listed fire doors and fire dampers in openings and ducts.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; The room should not be located above, below or next to hazardous processes.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt; Such compartmentation protects high-value operations from a fire elsewhere in the building.&amp;lt;ref name=&amp;quot;Davoodi-18-12&amp;quot;&amp;gt;Davoodi (2008), p. 18-12.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
Most mainframe installations stand on a raised access floor, which carries power and data cables and distributes cooling air.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; The floor and its supports should be noncombustible, cable openings should be kept small, and carpet should be avoided.&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Raised floors can become accumulation points for trash and debris.&amp;lt;ref name=&amp;quot;Mattern-11-31&amp;quot;&amp;gt;Mattern (2008b), p. 11-31.&amp;lt;/ref&amp;gt; Abandoned cables add fuel in the form of insulation and connectors and block cooling airflow, so a cable management programme should remove them.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
In New Zealand, Verification Method C/VM2 (second edition, 2025) names floor plenums for IT cables as an example of concealed spaces.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot;&amp;gt;Ministry of Business, Innovation and Employment (2025b), p. 41.&amp;lt;/ref&amp;gt; Its concealed space scenario applies to buildings with rooms holding more than 50 people.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt; The scenario is deemed satisfied by separating elements, automatic detection or suppression.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
Equipment and services need the same attention. IT equipment should meet UL 478 or UL 1950, and automated tape libraries may hold enough plastic to need their own detection and protection.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Each computer room should have its own air-conditioning system with noncombustible ducts, able to switch to 100% outside air and exhaust when smoke is detected.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt; Large battery banks should be in a separate ventilated room, and standby generators in separate machinery rooms or outdoors with at least one day of fuel.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== Detection ==&lt;br /&gt;
The handbook calls for smoke detection at ceiling level and under the raised floor, with underfloor devices listed for that use.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt; Underfloor plenums have high airflow, which makes many detector types unsuitable, so high-sensitivity sampling and spot detectors should be listed for high-airflow conditions.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt; A laser spot detector can be up to 100 times as sensitive as an ordinary light-emitting diode unit and is intended for computer rooms.&amp;lt;ref name=&amp;quot;Dungan-14-22&amp;quot;&amp;gt;Dungan (2008a), p. 14-22.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
[[Aspirating smoke detection]] draws air through a pipe network to a central, very sensitive detector.&amp;lt;ref name=&amp;quot;Dungan-14-22&amp;quot; /&amp;gt; These systems offer several alert and alarm levels and can release a gaseous agent in computer rooms.&amp;lt;ref name=&amp;quot;Dungan-14-22&amp;quot; /&amp;gt; Their limits are transport time, dilution of smoke over many sampling holes and the fact that airflow at an individual hole usually cannot be sensed.&amp;lt;ref name=&amp;quot;Dungan-14-22&amp;quot; /&amp;gt; The International Fire Engineering Guidelines (2005) note that such detectors are often up to 10 times as sensitive as point detectors but can respond with a considerable delay.&amp;lt;ref name=&amp;quot;IFEG-2.7-6&amp;quot;&amp;gt;National Research Council of Canada et al. (2005), p. 2.7-6.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
Air movement is the main difficulty. In the early stages of a fire the plume is weak, so supply and return flows can carry smoke away from detectors or keep it out of them.&amp;lt;ref name=&amp;quot;Schifiliti-4-26&amp;quot;&amp;gt;Schifiliti et al. (2002), p. 4-26.&amp;lt;/ref&amp;gt; No quantitative method existed in 2002 for this dilution, but [[Smoke stratification|stratification]] can be estimated.&amp;lt;ref name=&amp;quot;Schifiliti-4-26&amp;quot; /&amp;gt; Smoke rises to a height &amp;#039;&amp;#039;H&amp;#039;&amp;#039; only if the convective heat release of the fire exceeds a limit set by the temperature difference between the fire location and the ceiling:&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot;&amp;gt;Schifiliti et al. (2002), p. 4-27.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
&amp;lt;math display=&amp;quot;block&amp;quot;&amp;gt;\dot{q}_{conv} &amp;gt; 0.352\,H^{5/2}\,T_s^{3/2} \qquad \Leftrightarrow \qquad H &amp;lt; \left(\frac{\dot{q}_{conv}}{0.352\,T_s^{3/2}}\right)^{2/5}&amp;lt;/math&amp;gt;&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Symbol !! Description !! Unit&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;\dot{q}_{conv}&amp;lt;/math&amp;gt; || Convective heat release rate of the fire || W&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;H&amp;lt;/math&amp;gt; || Height from the top of the fuel to the ceiling || m&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;T_s&amp;lt;/math&amp;gt; || Temperature difference between fuel location and ceiling || °C&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
The SFPE handbook applies this to overheated PVC cable in a 7 m high air-conditioned room with switch equipment and a temperature difference of 10 °C.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; For a critical fire of 1,000 W, the smoke rises only about 6 m, so ceiling detectors may not respond.&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; Rearranging the same relation shows that about 1,450 W would be needed to reach 7 m. For a ceiling 3.05 m high, as in a data centre example in the NFPA handbook, about 180 W is enough at the same temperature difference.&amp;lt;ref name=&amp;quot;Harrington-17-134&amp;quot;&amp;gt;Harrington (2008), p. 17-134.&amp;lt;/ref&amp;gt; In low rooms, the dilution of smoke by forced ventilation with high exchange rates is therefore the larger concern.&amp;lt;ref name=&amp;quot;Schifiliti-4-26&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== Sprinklers and water mist ==&lt;br /&gt;
The handbook recommends sprinklers to NFPA 13, hydraulically designed as a wet pipe or [[Preaction sprinkler system|preaction system]].&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; In a single-interlock preaction system the pipes hold air, and water enters only after a separate detection system opens the valve.&amp;lt;ref name=&amp;quot;Huggins-16-35&amp;quot;&amp;gt;Huggins (2008), p. 16-35.&amp;lt;/ref&amp;gt; In a double-interlock system both the detection system and a sprinkler must operate, and preaction systems are typical in spaces with computer or communications equipment.&amp;lt;ref name=&amp;quot;Huggins-16-37&amp;quot;&amp;gt;Huggins (2008), p. 16-37.&amp;lt;/ref&amp;gt; Rawson advises against the double-interlock type because of its longer delay.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; The handbook&amp;#039;s own performance-based data centre example nevertheless uses a double-interlock system.&amp;lt;ref name=&amp;quot;Harrington-17-134&amp;quot; /&amp;gt; Few computer rooms meet the definition of light hazard, and rooms with large printers and paper should be treated as Ordinary Hazard Group 2.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
The handbook states that there should be little concern about shock or excessive water damage. Experience shows that properly installed and maintained sprinklers cause no measurable increase in damage to electronic equipment compared with heat, smoke and manual hose streams.&amp;lt;ref name=&amp;quot;Frank-17-41&amp;quot;&amp;gt;Frank (2008), p. 17-41.&amp;lt;/ref&amp;gt; NFPA 75, as summarised in the handbook, recommends disconnecting power to the equipment before water is applied.&amp;lt;ref name=&amp;quot;Frank-17-41&amp;quot; /&amp;gt; Portable extinguishers should be carbon dioxide or Class BC clean agent types, not dry chemical.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
A [[Water mist system|water mist system]] uses far less water, but its use as a halon replacement in computer rooms has been mixed.&amp;lt;ref name=&amp;quot;Mawhinney-16-139&amp;quot;&amp;gt;Mawhinney (2008), p. 16-139.&amp;lt;/ref&amp;gt; In total flooding mode mist cannot extinguish a fire inside an electronic cabinet, whereas a gaseous agent can penetrate the cabinet.&amp;lt;ref name=&amp;quot;Mawhinney-16-139&amp;quot; /&amp;gt; One system approved by FM Global draws subfloor smoke into 100 mm tubes, scrubs out soluble acid gases and pushes nitrogen into the cabinets above.&amp;lt;ref name=&amp;quot;Mawhinney-16-139&amp;quot; /&amp;gt; Above the floor, mist nozzles at light hazard spacing may match sprinklers on Class A fires with smaller volumes of water.&amp;lt;ref name=&amp;quot;Mawhinney-16-139&amp;quot; /&amp;gt; General claims that ceiling mist scrubs smoke are, according to Mawhinney, erroneous.&amp;lt;ref name=&amp;quot;Mawhinney-16-139&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== Gaseous agents ==&lt;br /&gt;
A [[Gaseous extinguishing system|gaseous extinguishing system]] is used where the value of the equipment or the risk of interruption justifies it, especially to protect data in process and allow a rapid return to operation.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt; From the 1960s, Halon 1301 total flooding became the usual agent for computer rooms.&amp;lt;ref name=&amp;quot;DiNenno-17-94&amp;quot;&amp;gt;DiNenno and Taylor (2008), p. 17-94.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Harrington-17-123&amp;quot;&amp;gt;Harrington (2008), p. 17-123.&amp;lt;/ref&amp;gt; Since the Montreal Protocol it is no longer made in most industrialised countries, and the US Clean Air Act banned new production after 1 January 1994.&amp;lt;ref name=&amp;quot;Harrington-17-128&amp;quot;&amp;gt;Harrington (2008), p. 17-128.&amp;lt;/ref&amp;gt; NFPA 2001 lists nine [[Halocarbon clean agent|halocarbon clean agents]] and four [[Inert gas agent|inert gas agents]] as replacements.&amp;lt;ref name=&amp;quot;Harrington-17-123&amp;quot; /&amp;gt; Carbon dioxide to NFPA 12 remains an option, but at about 9% most people become unconscious and at 25 to 30% breathing stops.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Harrington-17-134&amp;quot; /&amp;gt; Existing Halon 1301 systems should be tested and maintained to NFPA 12A.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
Clean agents are electrically nonconductive, leave no residue and can reach fuel inside closed cabinets.&amp;lt;ref name=&amp;quot;Harrington-17-128&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot; /&amp;gt; If power is cut before discharge, a fire in cable insulation or circuit boards becomes a Class A surface fire, which is relatively easy to extinguish.&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot; /&amp;gt; If power stays on, it remains a Class C fire that is harder to extinguish and can rekindle.&amp;lt;ref name=&amp;quot;Harrington-17-125&amp;quot; /&amp;gt; The handbook therefore recommends disconnecting power on discharge unless a risk assessment shows less damage with power maintained.&amp;lt;ref name=&amp;quot;Rawson-8-62&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
The agent quantity for total flooding follows from the room volume, the design concentration and the specific volume of the agent vapour. For halocarbon agents the mass is given by the first equation below; for inert gases the volume is given by the second, which assumes that leakage during discharge has a rising concentration.&amp;lt;ref name=&amp;quot;DiNenno-17-113&amp;quot;&amp;gt;DiNenno and Taylor (2008), p. 17-113.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;DiNenno-17-114&amp;quot;&amp;gt;DiNenno and Taylor (2008), p. 17-114.&amp;lt;/ref&amp;gt; The halocarbon flooding factor &amp;#039;&amp;#039;C&amp;#039;&amp;#039;/(100 − &amp;#039;&amp;#039;C&amp;#039;&amp;#039;) assumes that the agent and air lost during discharge are well mixed.&amp;lt;ref name=&amp;quot;DiNenno-17-113&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
&amp;lt;math display=&amp;quot;block&amp;quot;&amp;gt;w = \frac{V}{S}\left(\frac{C}{100 - C}\right) \qquad X = 2.303\,\frac{V}{S}\,\log_{10}\!\left(\frac{100}{100 - C}\right) V_s \qquad S = k_1 + k_2\,T&amp;lt;/math&amp;gt;&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Symbol !! Description !! Unit&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;w&amp;lt;/math&amp;gt; || Mass of halocarbon agent required || kg&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;X&amp;lt;/math&amp;gt; || Volume of inert gas required at 21 °C || m³&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;V&amp;lt;/math&amp;gt; || Net volume of the protected space || m³&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;C&amp;lt;/math&amp;gt; || Design concentration || % by volume&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;S&amp;lt;/math&amp;gt; || Specific volume of the agent at the minimum room temperature || m³/kg&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;V_s&amp;lt;/math&amp;gt; || Specific volume of the agent at 21 °C || m³/kg&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;k_1, k_2&amp;lt;/math&amp;gt; || Agent constants (HFC-227ea: 0.1268 and 0.0005133; IG-541: 0.659 and 0.00241) || m³/kg, m³/(kg·°C)&lt;br /&gt;
|-&lt;br /&gt;
| &amp;lt;math&amp;gt;T&amp;lt;/math&amp;gt; || Minimum ambient temperature of the protected space || °C&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As a worked example, take the data centre room of the handbook&amp;#039;s performance-based example: 1,858 m² with a ceiling 3.05 m high.&amp;lt;ref name=&amp;quot;Harrington-17-134&amp;quot; /&amp;gt; Ignoring the floor void and equipment, the net volume is about 5,660 m³. For HFC-227ea at 20 °C, &amp;#039;&amp;#039;S&amp;#039;&amp;#039; = 0.1268 + 0.0005133 × 20 = 0.137 m³/kg.&amp;lt;ref name=&amp;quot;DiNenno-17-114&amp;quot; /&amp;gt; At the ISO Class A design concentration of 7.9%, the mass is 5,660 / 0.137 × 7.9 / 92.1 ≈ 3,550 kg.&amp;lt;ref name=&amp;quot;DiNenno-17-98&amp;quot;&amp;gt;DiNenno and Taylor (2008), p. 17-98.&amp;lt;/ref&amp;gt; This concentration lies below the agent&amp;#039;s no observed adverse effect level of 9.0%, so the space may stay normally occupied for up to 5 minutes; see [[NOAEL and LOAEL]].&amp;lt;ref name=&amp;quot;Harrington-17-132&amp;quot;&amp;gt;Harrington (2008), p. 17-132.&amp;lt;/ref&amp;gt;&amp;lt;ref name=&amp;quot;Harrington-17-133&amp;quot;&amp;gt;Harrington (2008), p. 17-133.&amp;lt;/ref&amp;gt; With IG-541 at its minimum Class A design concentration of 37.5%, about 2,680 m³ of gas at 21 °C is needed instead.&amp;lt;ref name=&amp;quot;DiNenno-17-112&amp;quot;&amp;gt;DiNenno and Taylor (2008), p. 17-112.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
Halocarbon agents must discharge within 10 seconds, mainly to limit the formation of decomposition products.&amp;lt;ref name=&amp;quot;DiNenno-17-114&amp;quot; /&amp;gt; The replacement agents form three to eight times more [[Hydrogen fluoride|hydrogen fluoride]] than Halon 1301, and the main variable is the fire size at discharge.&amp;lt;ref name=&amp;quot;DiNenno-17-114&amp;quot; /&amp;gt; With automatic detection and actuation, full-scale tests found decomposition levels acceptable in typical computer and electronics spaces.&amp;lt;ref name=&amp;quot;DiNenno-17-116&amp;quot;&amp;gt;DiNenno and Taylor (2008), p. 17-116.&amp;lt;/ref&amp;gt; Systems traditionally had to hold their concentration for 10 to 20 minutes.&amp;lt;ref name=&amp;quot;DiNenno-17-119&amp;quot;&amp;gt;DiNenno and Taylor (2008), p. 17-119.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
The enclosure must also survive the discharge. Halocarbon agents first lower and then raise the room pressure, with peaks of about −383 to 383 Pa measured for HFC-227ea.&amp;lt;ref name=&amp;quot;Harrington-17-126&amp;quot;&amp;gt;Harrington (2008), p. 17-126.&amp;lt;/ref&amp;gt; Inert gases and carbon dioxide create a positive pressure that normally requires relief vents.&amp;lt;ref name=&amp;quot;Harrington-17-126&amp;quot; /&amp;gt; Typical metal stud walls resist over 1,197 Pa, but very tight rooms and rooms over 3.7 m high need care; see [[Enclosure integrity test]].&amp;lt;ref name=&amp;quot;Harrington-17-127&amp;quot;&amp;gt;Harrington (2008), p. 17-127.&amp;lt;/ref&amp;gt; Reliability also matters: for a halon system in a computer facility, the mean probability of failing to protect against a waste-paper fire was estimated at 0.13.&amp;lt;ref name=&amp;quot;Steciak-asCited&amp;quot;&amp;gt;Steciak and Zalosh (1992), as cited in National Research Council of Canada et al. (2005), p. 3.7-3.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== Design approach and emergency planning ==&lt;br /&gt;
Protection works best as layers matched to the stages of a fire. Grant&amp;#039;s table for computer rooms links good housekeeping to the pre-ignition stage and smoke detection to initial pyrolysis.&amp;lt;ref name=&amp;quot;Grant-4-155&amp;quot;&amp;gt;Grant (2002), p. 4-155.&amp;lt;/ref&amp;gt; Portable extinguishers and an automatic gaseous system cover the incipient stage, sprinklers the pre-flashover stage, and fire walls the post-flashover stage.&amp;lt;ref name=&amp;quot;Grant-4-155&amp;quot; /&amp;gt; Each layer protects a different interest, from business operation to the surrounding structure.&amp;lt;ref name=&amp;quot;Grant-4-155&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
The handbook illustrates [[Performance-based design|performance-based design]] with a corporate data centre whose stakeholders list business continuity first, before life safety and property.&amp;lt;ref name=&amp;quot;Harrington-17-134&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Harrington-17-135&amp;quot; /&amp;gt; The example limits interruption of critical operations to 4 hours and requires the agent to extinguish a fire of no more than 2 kW.&amp;lt;ref name=&amp;quot;Harrington-17-136&amp;quot;&amp;gt;Harrington (2008), p. 17-136.&amp;lt;/ref&amp;gt; It also limits halide acids to 200 ppm, requires a hold time of at least 10 minutes and caps discharge pressures at 239 Pa (5 psf).&amp;lt;ref name=&amp;quot;Harrington-17-136&amp;quot; /&amp;gt; The design fire is sustained arcing in a power circuit that ignites cable insulation inside an equipment cabinet.&amp;lt;ref name=&amp;quot;Harrington-17-138&amp;quot;&amp;gt;Harrington (2008), p. 17-138.&amp;lt;/ref&amp;gt; The example shows a common conflict: owners and insurers may want power to stay on, while code authorities may prefer to cut it before discharge.&amp;lt;ref name=&amp;quot;Harrington-17-137&amp;quot;&amp;gt;Harrington (2008), p. 17-137.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
An emergency plan completes the protection. It should keep backup copies of software and data, arrange alternative equipment or another site, and invite the local fire service to learn the extinguishing systems.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt; An effective preplan includes a preliminary visit by a restoration company, which can restore much of the damaged hardware.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt; Evacuation procedures should be posted and updated every six months.&amp;lt;ref name=&amp;quot;Rawson-8-63&amp;quot; /&amp;gt; The handbook also advises protecting telephone switching equipment in the same way as an important computer room; see [[Telecommunications facility]].&amp;lt;ref name=&amp;quot;Transue-9-174&amp;quot;&amp;gt;Transue (2008), p. 9-174.&amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== Limits and uncertainty ==&lt;br /&gt;
The fire statistics describe US fires reported to municipal fire departments, exclude fires reported only to industrial brigades and use coding that changed in 1999.&amp;lt;ref name=&amp;quot;Rawson-8-59&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Rawson-8-60&amp;quot; /&amp;gt; They therefore show trends rather than precise rates. The damage temperatures are general values; components and media differ, and their failure points vary.&amp;lt;ref name=&amp;quot;Rawson-8-61&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
The agent equations assume a well-mixed room and a known leakage behaviour.&amp;lt;ref name=&amp;quot;DiNenno-17-113&amp;quot; /&amp;gt; Class A extinguishing concentrations differ between test standards, partly because UL and FM allow 600 seconds for extinguishment and ISO only 180 seconds.&amp;lt;ref name=&amp;quot;DiNenno-17-97&amp;quot;&amp;gt;DiNenno and Taylor (2008), p. 17-97.&amp;lt;/ref&amp;gt; Deep-seated fires and cable arrays with long preburn times may need more agent than surface fires.&amp;lt;ref name=&amp;quot;DiNenno-17-112&amp;quot; /&amp;gt; Raising the safety factor does not help if doors are left open.&amp;lt;ref name=&amp;quot;DiNenno-17-113&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
Detection design also rests on judgement. The SFPE handbook gives no quantitative method for smoke dilution by air handling and notes that smoke reaching a detector does not guarantee that it enters the sensing chamber.&amp;lt;ref name=&amp;quot;Schifiliti-4-26&amp;quot; /&amp;gt;&amp;lt;ref name=&amp;quot;Schifiliti-4-27&amp;quot; /&amp;gt; The requirements cited here are those summarised in the 2008 handbook for US practice, with the NFPA standards of that period. Designers in other countries follow their own codes, such as C/VM2 in New Zealand.&amp;lt;ref name=&amp;quot;MBIE-41&amp;quot; /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== See also ==&lt;br /&gt;
* [[Records protection]]&lt;br /&gt;
* [[Battery energy storage system]]&lt;br /&gt;
* [[Semiconductor fabrication facility]]&lt;br /&gt;
&lt;br /&gt;
== Notes ==&lt;br /&gt;
&amp;lt;references /&amp;gt;&lt;br /&gt;
&lt;br /&gt;
== References ==&lt;br /&gt;
* Custer, R. L. P., &amp;amp; Milke, J. A. (2008). Fundamentals of fire detection. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 2-75–2-78). National Fire Protection Association.&lt;br /&gt;
* Davoodi, H. (2008). Confinement of fire in buildings. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 18-3–18-21). National Fire Protection Association.&lt;br /&gt;
* DiNenno, P. J., &amp;amp; Taylor, G. M. (2008). Halon and halon replacement agents and systems. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 17-93–17-121). National Fire Protection Association.&lt;br /&gt;
* Dungan, K. W. (2008a). Automatic fire detectors. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 14-15–14-28). National Fire Protection Association.&lt;br /&gt;
* Frank, J. A. (2008). Characteristics and hazards of water and water additives for fire suppression. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 17-31–17-43). National Fire Protection Association.&lt;br /&gt;
* Grant, C. C. (2002). Halon design calculations. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., &amp;amp; J. M. Watts Jr. (Eds.), &amp;#039;&amp;#039;SFPE handbook of fire protection engineering&amp;#039;&amp;#039; (3rd ed., pp. 4-149–4-172). National Fire Protection Association; Society of Fire Protection Engineers.&lt;br /&gt;
* Harrington, J. L. (2008). Application of gaseous agents to special hazards fire protection. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 17-123–17-139). National Fire Protection Association.&lt;br /&gt;
* Huggins, R. (2008). Automatic sprinkler systems. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 16-29–16-53). National Fire Protection Association.&lt;br /&gt;
* Mattern, L. J. (2008b). Housekeeping practices. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 11-27–11-38). National Fire Protection Association.&lt;br /&gt;
* Mawhinney, J. R. (2008). Water mist fire suppression systems. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 16-135–16-180). National Fire Protection Association.&lt;br /&gt;
* Ministry of Business, Innovation and Employment. (2025b). &amp;#039;&amp;#039;Verification Method C/VM2: Framework for fire safety design&amp;#039;&amp;#039; (2nd ed.).&lt;br /&gt;
* Mowrer, F. W., &amp;amp; Rosenbaum, E. R. (2008). Overview of performance-based fire protection design. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 3-183–3-194). National Fire Protection Association.&lt;br /&gt;
* National Research Council of Canada, International Code Council, Department of Building and Housing, &amp;amp; Australian Building Codes Board. (2005). &amp;#039;&amp;#039;International fire engineering guidelines&amp;#039;&amp;#039; (2005 ed.). Australian Building Codes Board.&lt;br /&gt;
* Rawson, B. P. (2008). Protection of electronic equipment. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 8-57–8-64). National Fire Protection Association.&lt;br /&gt;
* Schifiliti, R. P., Meacham, B. J., &amp;amp; Custer, R. L. P. (2002). Design of detection systems. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., &amp;amp; J. M. Watts Jr. (Eds.), &amp;#039;&amp;#039;SFPE handbook of fire protection engineering&amp;#039;&amp;#039; (3rd ed., pp. 4-1–4-43). National Fire Protection Association; Society of Fire Protection Engineers.&lt;br /&gt;
* Schmidt, D. L. (2008). Emergency management and business continuity. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 1-137–1-156). National Fire Protection Association.&lt;br /&gt;
* Tewarson, A. (2002). Generation of heat and chemical compounds in fires. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., &amp;amp; J. M. Watts Jr. (Eds.), &amp;#039;&amp;#039;SFPE handbook of fire protection engineering&amp;#039;&amp;#039; (3rd ed., pp. 3-82–3-161). National Fire Protection Association; Society of Fire Protection Engineers.&lt;br /&gt;
* Transue, R. E. (2008). Fire protection of telecommunications facilities. In A. E. Cote, C. C. Grant, J. R. Hall Jr., &amp;amp; R. E. Solomon (Eds.), &amp;#039;&amp;#039;Fire protection handbook&amp;#039;&amp;#039; (20th ed., pp. 9-167–9-177). National Fire Protection Association.&lt;br /&gt;
&lt;br /&gt;
[[Category:Special Hazards &amp;amp; Building Types]]&lt;/div&gt;</summary>
		<author><name>Mark</name></author>
	</entry>
</feed>