Çalışmalarında Yapılan Tipik Hatalar · Tasarım Patlama Yükü –Aşılma asınç Aşılma Eğrisi (Exceedance urve ) Kişisel Risk - Yaralanabirlik (Vulnerability Analysis)

Patlama Risk Analizleri ve Etkilerini Azaltma Çalışmalarında Yapılan Tipik Hatalar

Doç. Dr. Ali Sarı

İstanbul Teknik Üniversitesi

E-mail: [email protected]

14-15 MAYIS 2018

Sunumun İçeriği

Giriş - Motivasyon

Frekans Analizi için Tarihsel Veri Tabanlarının kullanılması

Bina Patlama Yükü ve Yapı Hasarı

Tasarım Patlama Yükü – Aşılma Basınç Aşılma Eğrisi (Exceedance Curve)

Kişisel Risk - Yaralanabirlik (Vulnerability Analysis)

Domino Etkisi ve Proses Harici Tehlikelerle Risk Analizi

Patlama Etkisini Azaltma Çalışması (Mitigation Study)

API 753 Zone Tanımı ve ATEX Zone

Giriş - Motivasyon

Giriş - Motivasyon

Giriş - Motivasyon

March 2005 April 2016

Kantitatif Risk Analizine Genel Bakış

Risk = Frequency X Consequence

Likelihood of hazardous scenario

occurring

Impact to Personnel, equipment,

infrastructure

Frekans Analizi

• Define Isolatable Sections

• Conduct Parts Count

• Apply Failure Frequency Data to Parts Count

• Calculate Overall Ignition Probability

• Assign Immediate/Delayed Ignition Probability

• Apply Directional Probability

Frekans Analizi – Sızıntı Frekansı (Leak Frequency)

Issues with historical data

No transparency of causes of the events (e.g. dropped object, corrosion, etc.)

Single value is generated for the freq (uncertainty should be applied) Leak frequency is conducted as an isolated activity (doesn’t rely on

frequencies established during HAZOP or reliability assessments such as fault tree)

Offshore databases are applied to onshore processes, when in reality there are differences in:

• Types of equipment• Materials of construction• Hazards the equipment is exposed to • Safety management systems in place


Ro

of

Ro

of

Ro

of

Ro

of

Blast Load Measured by Pressure Impulse/ Duration (assumes a shape)

Reflected v Side-on Incident Angle Rise Time Negative Phase Pressure Clearing

Ref: Design of Blast Resistant Buildings in Petrochemical Facilities, ASCE


11



12


Explosion Source

High Speed Camera

Free-field Pressure Sensors

Side-onPressure Sensors (Rear of Building)

Side-onPressure Sensors (Side of Building)

Reflected Pressure Sensors

(Front of Building)

Side-on Pressure Sensors (Roof of Building)


Crash Test Dummy

Interior Furniture and Office Equipment

Pressure Sensor

Pressure Sensor


1 psi = 6.9 kPa

Sunumun İçeriği

Before the test After the test


• Any design basis considers Non-Structural Members?


Control Bldg • Looking at the southwest corner of the control room.

• Notice the corner column still standing.

• The building was ~ 20 foot (6.1 mt) tall with reinforced

• Originally built to "so called" explosion proof requirements.


19

Tasarım Patlama Yükü – Aşılma Basınç Aşılma Eğrisi

1.E-10

1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

0.1 1 10 100

Eve

nt

Fre

qu

ency

(p

er y

ear)

Pressure (psig)

Many Scenarios! What is the design Load?


Scenario 2 - Propane Release

0

2

4

6

8

10

0 100 200 300 400 500

Distance from Congested Area Edge, R' (ft)

Fre

e-F

ield

Pre

ssur

e (p

si)

TNO MEM

BST Model


0

50

100

150

200

250

300

0 100 200 300 400 500

Distance from Congested Area Edge, R' (ft)

Impu

lse

(psi

-ms)

TNO MEM

BST Model


0

10

20

30

40

50

60

70

80

90

100

16 33 66 98 131 197 328 492

Distance from Congested Area Edge (ft)

Ear

drum

Rup

ture

(%

)

TNO MEM

BST Model


0

10

20

30

40

50

60

70

80

90

100

16 33 66 98 131 197 328 492

Distance from Congested Area Edge (ft)

Str

uctu

ral D

amag

e (

%)

TNO MEM

BST Model

21


The graphical representation of the frequency ofoccurrence or exceedence and the representativeoverpressure or impulse derived for the relevantscenarios at a particular location.

• Location of leak source• direction of gas jet• flow rate of the leak• wind direction and speed• performance of barrier elements


1.E-10

1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

0.1 1 10 100

Fre

qu

ency

of

P o

r m

ore

(p

er y

ear)

Pressure (psig)

~ 0.7 psi for 1E-04

~ 3.7 psi for 1E-05

~ 28 psi for 1E-06




26


Rise Time?

Positive Duration?

Negative Phase Pressure?

Negative Phase Duration?

Will it be applied uniformly on thewhole structure?

? ?

??


Blast Wall Connections may deform plastically subject to the blast positive phase.This will reduce the capacity of the connection andThe connection may fail due to the negative peak pressure and rebounding of the wall



• 𝑃 𝐹𝐼𝐼 = the probability of an individual experiencing an Fatality or Significant Injury (FI) (individual risk)

• 𝑃(𝐴𝑖) = the probability that accident Ai occurs.• 𝑃 𝐹𝐼|𝐴𝑖 = the conditional probability that an FI occurs, given that event/accident Ai

occurs.• 𝑃 𝐼 𝑎𝑡 𝐴𝑖 = the probability that an individual is present when accident Ai occurs.

𝑃 𝐹𝐼𝐼 = 𝑖=1,𝑛

𝑃 𝐹𝐼 |𝐴𝑖 × 𝑃 𝐼 𝑎𝑡 𝐴𝑖 × 𝑃(𝐴𝑖)

The vulnerability number (VN) is the fraction of occupants with serious, fatal injuries (FI) at a certain severity of structural damage.


Reference: American Petroleum Institute (API), 2003. Management of Hazards Associated with Location of Process Plant Buildings, 2nd ed., API RP 752, First Edition, May 1995.


𝑃 𝐹𝐼 |𝐵𝑖 × 𝑃 𝐼 𝑎𝑡 𝐵𝑖 × 𝑃(𝐵𝑖)


Reference: Chemical Industries Association (CIA), 2003. Guidance for the location and design of occupied buildings on chemical manufacturing sites, 2nd ed., London: CIA, ISBN 1 85897 114 4.


𝑃 𝐹𝐼 |𝐵𝑖 × 𝑃 𝐼 𝑎𝑡 𝐵𝑖 × 𝑃(𝐵𝑖)


𝑃𝐹𝑎𝑖𝑙 𝑖 =

𝑗,𝑘

𝑃 𝐹𝑎𝑖𝑙 𝐷 × 𝑃 𝐷 𝐴𝑗𝑘𝑖

× 𝑃(𝐴𝑗𝑘𝑖)

For each type of accidental load

Probability of damaged system failure under

relevant Accidental Load

Probability of damage, D given 𝐴𝑗𝑘𝑖

Probability of accidental action at location (j) and intensity (k)

• 𝑃(𝐴𝑗𝑘𝑖)is determined by risk analysis while the other probabilities are

determined by structural reliability analysis. • 𝑃 𝐹𝑎𝑖𝑙 𝐷 is determined by due consideration of relevant action and their

correlation with the hazard causing the damage


𝑃 𝐼|𝐷𝐿𝑖 = conditional probability that an I occurs given that damage level DLi occurs.

𝑃(𝐷𝐿𝑖) = probability that damage level Di occurs.

𝑉𝑁|𝐷𝐿𝑖 = vulnerability number at a certain structural damage level DLi.

𝑂𝑃𝑃= Occupant Presence Probability

𝑃 𝐼|𝐸𝑖 = conditional probability that an I occurs given that an escalation Ei occurs.

𝑃 𝐸𝑖|𝐷𝐿𝑖 = conditional probability that an escalation occurs given that damage level DLi occurs.

𝑉𝑁|𝐸𝑖 = vulnerability number at a certain escalation level Ei.

𝐼𝑅 = 𝑖=1,𝑛

𝑉𝑁|𝐷𝐿𝑖 × 𝑃 𝐼|𝐷𝐿𝑖 × 𝑃(𝐷𝐿𝑖)

𝐼𝑅 = 𝑉𝑁 × 𝑂𝑃𝑃 × (𝐶𝑎𝑙𝑐𝑢𝑙𝑎𝑡𝑒𝑑 𝐹𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦𝑖 − 𝐶𝑎𝑙𝑐𝑢𝑙𝑎𝑡𝑒𝑑 𝐹𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦𝑖−1)

𝐼𝑅 = 𝑖=1,𝑛

𝑉𝑁|𝐸𝑖 × 𝑃 𝐼|𝐸𝑖 × 𝑃 𝐸𝑖|𝐷𝐿𝑖 × 𝑃(𝐷𝐿𝑖)


0 500 1000 15000

5

10

15

20

25

30 1.199e-101.278e-101.337e-10 2.625e-10 2.799e-10 5.36e-10 5.36e-10 5.714e-10 5.714e-105.819e-10 6.203e-109.116e-109.719e-10 1.233e-091.705e-091.818e-09 1.918e-092.045e-09

9.719e-10

1.191e-099.116e-10

1.157e-092.875e-10

4.411e-11

2.43e-09

2.59e-09

2.787e-10

4.372e-106.725e-11

4.662e-101.117e-09

6.43e-11

1.117e-09

5.002e-08 4.87e-081.079e-07

2.696e-10

5.052e-10

2.875e-10

2.625e-10

3.741e-10

2.799e-10

1.023e-09

4.919e-10 1.09e-09

1.191e-09

3.989e-10 1.09e-09

Impulse (psi.ms)

Pre

ssur

e (p

si)

BDL 3-4BDL 2.5-3BDL 2-2.5BDL 4-FreqBDL 3-FreqBDL 2.5-FreqBDL 1&2-Freq


Es Sider, Libya Tank Fire in 2014 (Cause: Terrorist Attack) Response of a partially filled tank to blast within the shallow cloud (Buncefield explosion)


Collapsed tanks and piping system in tank farm due to fire after the Kocaeli Earthquake in Turkey in 1999

Tank farm fires after Kocaeli Earthquakein Turkey in 1999

Explosions at Arkema Facility after flooded during Harvey Storm in Texas, 2017


Domino system Heat Radiation


103

100

104

102101

(1)

(2)

(3)

(1) Secondary containment area fire:

• Dike fire at Tank 104

(1) Secondary event: • Pool fire at Tank 100

• Boil-over at Tank 100

(1) Secondary event: • Pool fire at Tank 103


Fire Impingement

104

100


103

100

104

102101

The

rmal

Rad

iatio

n (k

W/m

2 )

2

4

6

8

10

12

14

(1)

(2)

(3)

(1) Primary event:

• Dike fire at Tank 104

(2) Secondary event:

• Pool fire at Tank 100


(3) Secondary event:

• Pool fire at Tank 103



Without Domino Effect With Domino Effect



The Control Room Dilemma

Free-field blast overpressure contours


Blast Wall

Wall

Building

Explosive

Air

20’

1’

12’ 9’

40’

10’

15’

83’

Ground

This existing barricade was evaluated with CFD and found to offer only a 20% reduction in load on the building front wall. That wall and windows were still predicted to fail. Hence the barricade did not do its job.


Results: Animation of Velocity Contours in Mid-section


Masonry Shield Wall

Ref: Sari et al, ASCE Structures, 2009


Interior Catch System

Construction Photos



Strongback Masonry Retrofit


Masonry Wall –Strongback Retrofit


50

250

500

750

1,000

1,250

1,500

1,750

2,000

2,250

2,500S

tan

do

ff D

ista

nce

Fee

t (f

rom

ed

ge

of

Co

ng

este

d V

olu

me

to e

dg

e o

f tr

aile

r)

570 ft

1,930 ft

ZONE 1

ZONE 2

ZONE 3

7,50

0

100,

000

200,

000

300,

000

400,

000

500,

000

600,

000

700,

000

800,

000

900,

000

1,00

0,00

0

Congested Volume (Cubic Feet see Appendix for selected distances)

Standoff Distances/Zones for Portable Buildings

250

500

750

1,000

1,250

1,500

1,750

2,000

2,250

2,500S

tan

do

ff D

ista

nce

Fee

t (f

rom

ed

ge

of

Co

ng

este

d V

olu

me

to e

dg

e o

f tr

aile

r)

570 ft 570 ft

1,930 ft 1,930 ft

ZONE 1

ZONE 2

ZONE 3

7,50

0

100,

000

200,

000

300,

000

400,

000

500,

000

600,

000

700,

000

800,

000

900,

000

1,00

0,00

0

Congested Volume (Cubic Feet see Appendix for selected distances)

Standoff Distances/Zones for Portable Buildings

Zone 1-Blast analysis required-No Occupied Light Wood Trailers- House only Essential Personnel

Zone 2Detailed Blast analysisrequired for all portablebuildings

Zone 3 (Simplified Method)

-No blast analysis required except window glass-Any portable building permitted-Window glass hazards to be addressed

Ref: API 753


Sorular?

Dinlediğiniz için Teşekkür Ederim.

Doç. Dr. Ali SARI

İTÜ

E-mail: [email protected]

FUAR İÇİ 41040 İZMİT/KOCAELİ

TEL: +90 262 315 80 00

FAX: +90 262 321 90 70

WEB: www.kosano.org.tr

E-MAIL: [email protected]

Frekans Analizi – Sızıntı Frekansı (Leak Frequency)

Then why use these offshore databases?

Availability Conservative Leak freq analysis can be conducted quickly Client wants to compare other assets using same methodology

Seismic Response of Storage Tanks - Site Seismic Hazard Curves

54

• Site seismic hazard curves can be developed for a specific site; or

• Can also be obtained from the 2015 National Building Code of Canada (NBCC) and modified based on local soil characteristics.

http://www.earthquakescanada.nrcan.gc.ca/hazard-alea/interpolat/index_2015-eng.php

http://www.earthquakescanada.nrcan.gc.ca/hazard-alea/interpolat/index_2015-eng.php

Fragility Curves for On-grade Steel Tanks with % Full > 50%

55* Michael O'Rourke and Pak So, 2000

DS1

DS5

DS4

DS3

DS2

Kern County Earthquake – 1.0gPlastic Strain Contours (Damage)

56

Elephant-foot buckling with major loss of contentment, severe damageDamage = DS4

Rupture size = 11×4.5 m

Containment loss = 100%

Red color indicates 5% plastic strainMax = 189%

Elephant-foot buckling

Fragility Curves Damage State Definitions

57

Damage State Description

DS1 No damage to tank or I/O pipes

DS2 Damage to roof, minor loss of contents, minor damage to piping, but no elephant-foot buckling

DS3 Elephant-foot buckling with minor loss of content

DS4 Elephant-foot buckling with major loss of content, severe damage

DS5 Total failure, tank collapse

Overall Modeling Procedure

58

The modeling procedure includes following steps:

Step 1: Carry out a coupled heat transfer-CFD analysis to quantify the temperature evolution over the fire exposed side of the tank.

Step 2: Carry out structural analysis to predict the performance of the tank at elevated temperature.

Thermal Analysis Structural Analysis

Documents

Çalışmalarında Yapılan Tipik Hatalar · Tasarım Patlama Yükü –Aşılma asınç Aşılma Eğrisi (Exceedance urve ) Kişisel Risk - Yaralanabirlik (Vulnerability Analysis)