calculating fire risk at sandia national laboratories · fire risk index : 0 10 20 30 40 50 60 70...
TRANSCRIPT
Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed
Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000.
Photos placed in horizontal position
with even amount of white space
between photos and header
Calculating Fire Risk at Sandia National Laboratories
Fire Protection & Assurance Department
Alice Muña
This briefing presents a methodology for evaluating risk on a building level.
Qualitative Fire Risk Index
Benefits
Improve our Emergency Services Response Time criteria (documented in BNA)
Provide risk-based prioritization model for buildings across the site:
FPA schedule
Maintenance prioritization
Opportunities prioritization
Assumptions
Risk = Probability of a Fire * Consequence
The Fire Consequence Index (FCI) is a relative risk ranking and is not a measure of absolute risk
Categories will be selected and weighted based on their perceived contribution to severity and increased consequence of a fire, given a fire occurs
Fire Consequence Categories
Replacement Value (RPV)
Building Contents Value
Square Footage of the Building
Occupancy Type
Building Population
Fire Alarm
Automatic Sprinkler Protection
Response Time for Fire Department
Special Consequences
Information from FIMS (FCI points given in parentheses)
RPV
>$50 Million (5)
<$50 Million (4)
<$5 Million (2)
<$1 Million (1)
Building Contents Value
>$5 Million (5)
<$5 Million (4)
<$1 Million (3)
<$250 Thousand (1)
Square Footage
>100,000 ft2 (5)
<100,000 ft2 (4)
<50,000 ft2 (2)
<10,000 ft2 (1)
$0
$20
$40
$60
$80
$100
$120
$140
$160
$180
1 75 149 223 297 371 445 519 593 667 741
Millio
ns
Buildings
Replacement Value
0
100
200
300
400
500
600
700
<$1 Million <$5 Million <$50 Million >$50 Million
Nu
mb
er
of
Bu
ild
ing
s
Replacement Value
From our files…
Occupancy Type: High Hazard (5)
Factory (3)
Utility (3)
Assembly (2)
Storage (2)
Business (Office & Lab) (1)
Building Population >300 (5)
<300 (4)
<100 (2)
<10 (1)
No occupants (0)
From our files…
Working Fire Alarm transmitting to a constantly attended location? (0 if yes, 5 if no)
Automatic sprinkler protection? (0 if yes, 5 if no)
Travel Time- how long will it take the FD to reach the building? Remote Sites (5)
Tech Area III (4)
Tech Area V (3)
Tech Area II, TAIV (2)
Tech Area I (1)
Special Consequences
Special Consequences Nuclear (5)
Exposure to the Public (5)
Accelerators (4)
Critical Infrastructure (4)
Clean Rooms (3)
Explosives (3)
Hazardous Material (3)
Critical Infrastructure Support (3)
𝐹𝐶𝐼 = (𝑃𝑜𝑖𝑛𝑡 𝑉𝑎𝑙𝑢𝑒𝑖 ∗ 𝑊𝑒𝑖𝑔ℎ𝑡𝑖𝑛𝑔 𝐹𝑎𝑐𝑡𝑜𝑟𝑖)
𝑛
𝑖=1
Consequence
Category
Value
Range
Weighting
Factor
Replacement Value
(RPV)
1 - 5 4
Building Contents
Value
1 - 5 3
Square Footage 1 - 5 1
Occupancy Type 1 - 5 1
Building Population 0 - 5 6
Fire Alarm 2 - 5 1
Automatic
Sprinklers
0 - 5 3
Travel Time 1 - 5 2
Special
Consequences
0 - 5 3
Fire Consequence Index-SNL/NM
0
10
20
30
40
50
60
70
0 100 200 300 400 500 600 700 800 900
Fir
e C
on
se
qu
en
ce
In
de
x V
alu
e (
Ma
x 1
00
)
Buildings
0
10
20
30
40
50
60
70
0 100 200 300 400 500 600 700 800 900
Fir
e C
on
se
qu
en
ce
In
de
x V
alu
e (
Ma
x 1
00
)
Buildings
Fire Consequence Index-SNL/NM
Highest consequence if a
fire occurred
Buildings with Highest Fire Consequence
High RPV
High Building Contents
High Building Population
Special Hazard Hazardous Materials
Exposure to the Public
Semi-conductor facilities have the
greatest fire consequence, based on
this methodology.
Next Step
More difficult to quantify
Limited data points from Sandia fires
Building missions change over time
Fire Risk = Probability of a Fire * Consequence
Probability Data
DOE Fire Loss Database Fires from all DOE sites reported
Data from 1991-2007
2,541 Events during this time period
Events Categorized: Building Fire
Brush Fire (Wildland Fire)
Vehicle Fire
Other Fire (Fires occurring outside buildings excluding brush fires)
To Calculate Probability
Probability of a fire in a year = 𝟏 − 𝒆−𝝀∗𝟑𝟔𝟓
λ= Average Number of Days between Fires
Poisson Distribution is most often used for rare events where: Occurrences are independent
Average frequency of occurrence for the time period in question is known
Event can be counted in whole numbers
DOE- Building Fire
DOE- Brush Fire
DOE- Vehicle Fire
DOE- Other Fire
$0
$2,000
$4,000
$6,000
$8,000
$10,000
$12,000
$14,000
$16,000
$18,000
$20,000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Se
ve
rity
(L
os
s/F
ire
)
Probability (Fire Frequency)
Severity vs. Probability 1991-2007
Not included: Los Alamos Fire Cost, ~200 DOE fires that have no dates associated
DOE- Building Fire
DOE- Brush Fire
DOE- Vehicle Fire
DOE- Other Fire
SNL- Building Fire
SNL- Brush Fire
SNL- Vehicle Fire
SNL- Other Fire
$0
$2,000
$4,000
$6,000
$8,000
$10,000
$12,000
$14,000
$16,000
$18,000
$20,000
0 0.2 0.4 0.6 0.8 1 1.2
Se
ve
rity
(L
os
s/F
ire
)
Probability (Fire Frequency)
Severity vs. Probability 1991-2007
Not included: Los Alamos Fire Cost, ~200 DOE fires that have no dates associated
SNL has a higher probability
of fires than DOE’s average
but has a lower average cost
per fire.
Fire Probability- By Building
Using Space Chargeback data, buildings are broken down into four types of space: General Areas
Administrative, Conference, Hall, Kitchens, Equipment Rooms, and difference in square footage between space chargeback and FIMS
Office Areas
General Office, Management Office, Secretarial Office, Computer Rooms
Lab Areas
General Lab, High Bay Lab, Shop, Special Lab
Storage Areas
Fire Probability- By Building
Using the DOE Fire Loss Database (1991-2007) and Sandia’s recent DOE Fire Loss Reports (2008-2013), we can determine the probability of fires in these four types of spaces
Probability determined using the Poisson Distribution
Space Frequency (λ) Probability
General 0.005069 0.84279
Lab 0.003707 0.74155
Office* 0.000124 0.04425
Storage 0 0
*There has only been one office fire at Sandia in this time frame, so there is
not enough data to determine an accurate frequency.
Probability of a fire in a year = 1 − 𝑒−𝜆∗365
For the entire site
Fire Probabilities
0.00000
0.10000
0.20000
0.30000
0.40000
0.50000
0.60000
0.70000
0.80000
0.90000
0 200 400 600 800 1000
Fir
e P
rob
ab
ilit
y
Buildings
>300 Buildings at 0
Frequency:
Storage Areas
0
10
20
30
40
50
60
70
0.0 0.2 0.4 0.6 0.8 1.0
FR
I V
alu
e (
Max 1
00)
Fire Probability
Fire Risk Index
Higher Risk Facilities
0
10
20
30
40
50
60
70
0.0 0.2 0.4 0.6 0.8 1.0
FR
I V
alu
e (
Max 1
00)
Fire Probability
Fire Risk Index
Fire Risk = Probability of a Fire * Consequence
0
5
10
15
20
25
30
35
40
45
50
1 101 201 301 401 501 601 701
FR
I V
alu
e (
Ma
xim
um
10
0)
Buildings
Fire Risk Index
0
10
20
30
40
50
60
70
0.0 0.2 0.4 0.6 0.8 1.0
FC
I V
alu
e (
Max 1
00)
Fire Probability
One Year Three Year Five Year
SNL Current Fire Protection Assessment (FPA) Process
Our highest-risk facilities
are on a three-year FPA
cycle.
0
10
20
30
40
50
60
70
0.0 0.2 0.4 0.6 0.8 1.0
FC
I V
alu
e (
Max 1
00)
Fire Probability
One Year Three Year Five Year
SNL Proposed FPA Process
Our highest-risk facilities
are on an annual FPA
cycle. All mission critical
facilities are still on a 3
year cycle.
What’s Next?
Difficult to quantify controls
More Data = Better Results
Implementation of FRI
Thank You!