SAND2017-0836PE

Gaseous Hydrogen Separation

Distances

Chris LaFleur
Sandia National Laboratories
January 24, 2017

SAND2017-XXXX
Sandia National Laboratories is a multi-mission 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.
 Outline

• Historical Separation Distances

• 2010 Risk-Informed Separation Distances
• Updates to 2010 Risk-Informed Separation Distances

2
Historical Gaseous Separation Distances (pre-2010)

Before 2010, separation

distance requirements were
developed by expert
judgment and based on
total volume.
Creating Risk-Informed Separation Distances (2010
Ed. NFPA 55)

Determine
Determine hazard
separation
Identify exposures scenario for each
distance for each
exposure
exposure
1.0E-05
Increasing leak
diameter

Cummulative Risk (/yr)

1.0E-06
Cummulative frequency
of accidents requiring
this separation distance

1.0E-07

Risk Separation
Criteria Distance
1.0E-08
0.00 5.00 10.00 15.00
Separation Distance (m)

For more information, see Appendixes E and G in the 2010 Ed. of NFPA 55

4
Step 1: Identify Exposures

• Specified distances between a hazard source and a target

H2 human, equipment,
ignition sources, etc.
Storage

• Exposures can be:

– Property lines
– Exposed persons not involved in
servicing the system
– Air intakes
– Parked vehicles
– Public sidewalks
– Other hazardous materials
– Ignition sources
– Wall openings
– Utilities (overhead)
2. Determine Hazard Scenario for Each Exposure
• Each exposure identified was mapped to one or more hazard
scenarios and subsequent harm criteria
• Design scenarios with their associated performance criteria and
design scenarios were extracted from NFPA 1 for each hazard scenario

Exposure Hazard Scenario(s)

Lot Lines • Gas release and subsequent entrainment or accumulation by the
receptor
• Fire spread to or from adjacent equipment or structure
• Gas explosion hazard on site or affecting adjacent property
• Threat of injuries on site or adjacent property
• Ignition of an unignited release/vented hydrogen
Exposed persons other • Threat of injuries on site or adjacent property
than those involved in
servicing of the system
Ordinary Combustibles • Fire spread to or from adjacent equipment or structure

6
 Harm Criteria
• Harm criteria was based on exposure type and:
– Radiative heat flux
– Unignited jet concentration distances
– Visible flame length
• Using these distances based only on harm criteria is a consequence-
based approach
• The task group determined that the probability of occurrence should
also be considered in determining a reasonable level of safety
Exposure Harm Criteria Examples
Lot Lines Unignited jet concentration Radiation heat flux level
decay distance to 4% mole of 1577 W/m2
fraction H2
Exposed Persons Radiation heat flux level of 4732 W/m2
for a maximum of 3 minutes
Ordinary Heat flux level of 20,000 Visible flame length
combustibles W/m2
7
Consequence-Based Separation Distances Vary
Significantly with Leak DiameterHazard Distances for a Jet Fire:
1.6 kW/m2 Radiation Heat Flux
• Sandia Hydrogen models 70

60
Leak
Diameter

were used to generate (mm)

HarmDistance(m)
50
13.5

hazard distances for the 40

11.5
9.52

harm criteria dependent on

30 6.35
4.23
20 2.38

leak diameter 10
1.00
0.40
0.18
Hazard Distances for Different Consequence 0
0 20 40 60 80 100 120

45
Measures: 2.38 mm Leak Pressure (MPa)

40 Consequence
Parameter
35

• Frequency and sizes of

1.6 kW/m2
HarmDistance (m)

30 4.7 kW/m2
25 kW/m2
25

20
Flame Length
2% Hydrogen
leaks were evaluated using
15
4% Hydrogen
6% Hydrogen industry failure data and
8% Hydrogen
10

5
Bayesian statistics
0
0 20 40 60 80 100 120
System Pressure (MPa)

Can select leak diameter using a risk-informed approach

Sandia H2 Leak Model
• Model predicts (as 60
Luv/Lir/Lvis_ave_time_V.qpa2
6.0
P = 207.85 bar (3000 psig)

function of system
tank

50 2% m.f.
5.0 Data
4% m.f.
volume, pressure, and Simulation (L*: +10%)

Flame Length (m)

Distance (m)
4.0 40 6% m.f.

leak size):
Simulation (Nominal)
8% m.f.
3.0
Simulation (L*: -10%) 30
– Radiant heat flux from 20
hydrogen jet flames 2.0

– Visible flame length for 1.0 10

ignited jets 0.0

0
1 2 3 4 5 6 7
– Hydrogen concentrations in 0 10 20 30 40 50 60
Diameter (mm)
jets Time (sec)

• Assumes circular orifice

for leak geometry and
constant pressure -
conservative
• Model validated against
Sandia/SRI experiments
3. Determine Separation Distance for Each Exposure

• The outcome is a risk-informed approach to

creating separation distances
1.0E-05
Increasing leak
diameter
Cummulative Risk (/yr)

1.0E-06
Cummulative frequency
of accidents requiring
this separation distance

1.0E-07

Risk Separation
Criteria Distance
1.0E-08
0.00 5.00 10.00 15.00
Separation Distance (m)

10
 Selected Risk Guideline
• Individual fatality risk to most exposed person at facility
boundary selected for use in risk evaluation
• Use risk “Guideline” versus “Criteria”
– Criteria varies for different countries and organizations
– Making decisions based on comparison to hard risk criteria difficult
because of uncertainties in risk evaluations
• Comparison of mean risk to guideline is usually done
• Sensitivity studies and uncertainty analysis used to determine importance
of assumptions

NFPA 2 Working Group chose 2E-5 fatalities/yr as guideline

Basis – Comparative risk to gasoline stations, 10% of risk to society from all
other accidents, 1E-5/yr is a value used by most countries that have
established a risk criteria

1/31/2017
 Risk Results for Representative Systems
Total Risk 20.7 MPa (3000 psig) System Total Risk 103.4 MPa (15000 psig) System
5.0E-05 9.0E-05
Cumulative Frequency of Fatality (/yr)

4.5E-05

4.0E-05 Lot Line

7.0E-05

Frequency of Fatality (/yr)

Separation Lot Line
3.5E-05 Distance f or Separation
1%A to 10%A Sum Distance f or Sum
3.0E-05
Leak Sizes Flame Length
5.0E-05
1%A to 10%A Flame Length
2.5E-05 Leak Size
4% H2 4% Hydrogen

2.0E-05
3.0E-05
1.5E-05

1.0E-05

5.0E-06 1.0E-05
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Separation Distance (m) Separation Distance (m)

J. LaChance et al., “Analyses to Support Development of Risk-Informed Separation Distances for Hydrogen
Codes and Standards”, SANDIA REPORT, SAND2009-0874, Printed March 2009

• Risk close to the “guideline” of 2E-5 fatalities/yr selected by NFPA

Task Group 6
• Risk from leaks greater than 3% of flow area were deemed
acceptable
1/31/2017
 Resulting Revisions to Gaseous Separation Distances
• Separation distances based on Pressure and Pipe size
• Typical values provided in table in code
• Actual risk-based equation also in code to calculate for other pipe
diameters.

13
 Updates to Separation Distance Tables - 2016
• Three key decisions were made by the task group for the 2010 Edition
of NFPA 55 that were reevaluated for the latest edition of NFPA 55:
– Changed the internal pipe diameter leak size from 3% to 1% to remove
excess conservatism
• This accounts for 95% of leakage frequency from the example systems
– Changed the ‘no harm’ criteria of 1.6 kW/m2 to 4.7 kW/m2
• The 1.6kW/m2 assumed that exposed persons will not take protective
actions, such as relocating from the fire scene
– Hydrogen concentration threshold changed from 4% to 8% based on
work performed at Sandia
• Because of the removal of the excess conservatism, the task group
decided to add a safety factor of 1.5 to the safety distance

14
Updates to Separation Distance Tables

• Key decisions were made by the task group for the 2010 Edition of
NFPA 55 that were reevaluated for the latest edition of NFPA 55:
• Unignited jet concentration from 4% Separation Distance
>51.71
to 8% >0.10 >1.72 to >20.68 to
Code to 1.72 20.68 to 51.71 103.43
• Heat Flux level from 1.6 kW/m2 to 4.7 Exposures Version MPa MPa MPa MPa
kW/m2 2016 12 m 14 m 9m 10 m
Group 1
• Leak area from 3% to 1% Exposures 2019 5m 6m 4m 5m

• Added safety factor of 1.5 Group 2 2016 6m 7m 4m 5m

Exposures 2019 5m 6m 3m 4m

Group 3 2016 5m 6m 4m 4m
Exposures 2019 4m 5m 3m 4m

H2 Gas Bulk HVAC

System
2016: 10 m for 70 MPa
storage Store

Building
Opening
Updates to Separation Distance Tables

• Key decisions were made by the task group for the 2010 Edition of
NFPA 55 that were reevaluated for the latest edition of NFPA 55:
• Unignited jet concentration from 4% Separation Distance
>51.71
to 8% >0.10 >1.72 to >20.68 to
Code to 1.72 20.68 to 51.71 103.43
• Heat Flux level from 1.6 kW/m2 to 4.7 Exposures Version MPa MPa MPa MPa
kW/m2 2016 12 m 14 m 9m 10 m
Group 1
• Leak area from 3% to 1% Exposures 2019 5m 6m 4m 5m

• Added safety factor of 1.5 Group 2 2016 6m 7m 4m 5m

Exposures 2019 5m 6m 3m 4m

Group 3 2016 5m 6m 4m 4m
Exposures 2019 4m 5m 3m 4m
H2 Gas Bulk
System HVAC
2016: 10 m for 70 MPa
storage Store
2019: 5 m for 70 MPa
Store
storage
Building
Opening

Risk-based code requirements based on proposed risk threshold revisions enable

more sites to readily accept hydrogen infrastructure
Separation Distance Reductions: Group 1 Exposures
Group 1 Exposures
16
2005 Table Value for Air
14 Intakes (Highest Group 1
Exposure Value)
Separation Distance (m)

12 2005 Table Value for Ignition

Sources
10
2005 Table Value for Wall
8 Openings

6
2016 Table for Group 1
Exposures
4

2 2019 Proposed Table for

Group 1 Exposures
0
>0.10 to 1.72 >1.72 to 20.68 >20.68 to 51.71 >51.71 to 103.43
MPa MPa MPa MPa

Group 1 Exposures:
• Lot lines
• Air intakes
• Operable openings in buildings
• Ignition sources
17
 Separation Distance Reductions: Group 2 Exposures

Group 2 Exposures
8

7 2005 Table Value for

Separation Distance (m)

Public Sidewalks and

6 Parked Vehicles
5 2016 Table for Group 2
Exposures
4

3 2019 Proposed Table

for Group 2 Exposures
2

0
>0.10 to 1.72 >1.72 to 20.68 >20.68 to 51.71 >51.71 to
MPa MPa MPa 103.43 MPa

Group 2 Exposures:
• Exposed persons other than those
servicing the system
• Parked Cars
18
 Separation Distance Reductions: Group 3 Exposures
Group 3 Exposures
16

14 2005 Table for Fast-

Separation Distance (m)

Burning Solids (Highest

12 Value for Group 3
Exposures)
10
2016 Table for Group 3
8 Exposures

4 2019 Proposed Table for

Group 3 Exposures
2

0
>0.10 to 1.72 >1.72 to 20.68 >20.68 to 51.71 >51.71 to
MPa MPa MPa 103.43 MPa

Group 3 Exposures:
• Hazardous material storage systems
• Slow burning combustible solids
• Fast burning solids
• Overhead utilities
19
• Flammable gas metering, etc.
 Summary
• Task group updated bulk hydrogen gaseous separation distances
based on a risk-informed scientific approach for the 2010 Edition of
NFPA 55
• NFPA 2/55 task group reviewed judgements made in the first iteration
and removed excess conservatism from key judgements made earlier
• These numbers were approved by the larger NFPA 2/55 Committees
in the first draft meeting in Fall 2016

For more information, see:

• 2010 Ed. NFPA 55, Appendix E & Appendix G
• SAND2009-0874
20
Back-Up Slides

21
 Progress: Science-Based Prescriptive Requirement
Revisions LH2
• Goal: Use QRA tools and methods to revise bulk liquid hydrogen
system separation distances in NFPA 55/NFPA 2
• Progress:
– The NFPA 55/2 hydrogen storage task group performed a risk analysis on
a representative bulk liquefied hydrogen storage system and determined
nine release scenarios with the highest risk
• Six of the highest-risk scenarios are during liquid hydrogen transfer
operations from a tanker truck to the bulk LH2 storage tank
• Three scenarios are during normal system operations
– Determined model inputs and risk criteria for the nine scenarios

Results of the risk analysis on the bulk liquefied hydrogen storage

system will be fed into liquid hydrogen models

22
 Details of LH2 Prescriptive Code Revision Scenario
Selection and Prioritization
Event Likelihood Classification
• CGA P-28 OSHA Process Safety Level Annual Probability Probability Description
1 Frequent > 1.0 Expected to occur once per year or more
Management and EPA Risk 2 Reasonably probable 1.0 to 0.1
frequently.
Expected to occur once per 10 years.
Management Plan Guidance 3
4
Occasional 0.01 to 0.1
Remote 0.001 to 0.01
Expected to occur once per 100 years.
Expected to occur once per 1000 years.
Document for Bulk Liquid 5 Extremely remote 0.0001 to 0.001 Expected to occur once per 10,000
years.

Hydrogen Systems was used as a 6 Improbable < 0.0001 Expected to occur less than once per
10,000 years. Extremely unlikely to
occur.
basis for typical LH2 system Hazard Severity Classification
definition and HAZOP scenario Level Description Potential Consequences
1 Catastrophic May cause fatality to non-associated members of the
identification 2 Critical
public.
May cause severe injury to non-associated members of

• Each scenario was reviewed and the public, fatality or serious injury to works of the public,
fatality or serious injury to workers of persons
conducting business at a refueling site or significant
assigned an Even Hazard and damage to equipment/facilities.
3 Marginal May cause minor injury, or minor system damage.
Hazard Severity value. 4 Negligible Will not result in injury or system damage.

• Based on these values, the Likelihood

1 2 3 4 5 6
scenario was given a risk ranking

Severity
1 1 1 1 2 3 4
2 1 1 2 3 3 4
which was used to prioritize the Risk Ranking: 3 2 2 3 3 4 4
4 4 4 4 4 4 4

scenario 1: High Risk

2: Moderate Risk
3: Low Risk
4: Routine Risk

23
 LH2 Prioritized Scenarios to be Used for Separation
Distance Revision
HAZOP Number and Description HAZOP Number and Description
Release 1.18 High flow of gaseous hydrogen from Release 4.15 Loss of containment from pipe
scenarios trailer vent stack due to venting excess scenarios leading from tank to vaporizer
during pressure after LH2 transfer during or vaporizer itself caused by
liquid normal thermal cycles or ice falling
transfer to system from vaporizers
bulk 1.19 Normal flow from trailer vent stack due operation
storage to venting excess pressure after LH2 6.15 Misdirected flow caused by
tank transfer operator error resulting in large
low level release of cold
1.6 High flow from line rupture, valve or
gaseous hydrogen through
component failure during transfer
bottom drain valve of vent stack
process
during normal tank venting
1.4 High temperature due to external fire process
causes high flow venting through tank
2.1 High pressure because of a
vent stack
leak in inner vessel allowing
1.8 Reverse flow during transfer process hydrogen into the vacuum area
caused by human error and pressure
mismanagement
1.16 Loss of containment from external
impacts, consider all causes

24
Approach: Application of QRA to Performance-Based Design

Representative Station: Performance- Real-World

H2FIRST Reference Based Design of Station:
Liquid Station Refueling Station Linde Foster City

Meets all Modifies Modifies One Key Requirement

Prescriptive Key and Incorporates Mitigating Factors
Requirements Requirements

Prepare Performance-Based
Calculate Benchmark Incorporate Design Report and Documentation
Performance Mitigating Utilizing HyRAM QRA Toolkit
Criteria Factors

Follows Real-World Permitting

Calculate Risk Process
Risk Equivalent
Performance
Criteria

Vet with H2 Code

Industry and
Stakeholders
25
Risk-Informed Approach to Select Leak Diameter

• Examined appropriate
leakage data to determine
leak size distribution
– Selected leak size

Used QRA to determine if risk

from leaks greater than
selected leak size is acceptable
for typical systems
 Required Leakage Frequencies as a Function of
Leak Size and Pressure
Very little hydrogen-specific data available:
• Not enough for traditional statistical approach
• Instead, representative values are selected from other industries (oil and gas)
1.0E-02
Problems with this approach:
• Not hydrogen specific 1.0E-03

Leakage Frequency (/m-yr)

• Parameter uncertainty distribution 1.0E-04
is uncharacterized
1.0E-05

Solution:
1.0E-06
• Use Bayesian statistics to generate
leakage frequencies 1.0E-07 Generic Mean

– Combine sources of generic data Generic Median

Hydrogen Mean
with H2 specific data 1.0E-08
Hydrogen Median

• Allows attachment of different “layers”

Published Frequencies
1.0E-09
of significance to the data 0.01% 0.10% 1.00% 10.00% 100.00%
Leakage Area (%Flow Area)

Reference: “Handbook of Parameter Estimation for Probabilistic Risk Assessment,” NUREG/CR-6823, U.S. Nuclear
Regulatory Commission, Washington, D.C. (2003).
Mean Component Leakage
Frequencies from Bayesian Analysis

Hydrogen Leakage Frequencies

1.0E-01

1.0E-02
Mean Leakag Frequency (/yr)

Compressors
1.0E-03
Cylinders
Hoses
1.0E-04
Joints
Valves
1.0E-05 Pipes

1.0E-06

1.0E-07
0.10% 1.00% 10.00% 100.00%
Leak Area (% Flow Area)
Component Leak Frequencies Used to Determine
Cumulative System Leakage Probability
Evaluated for the representative storage facilities:
1.00
1.0E+00
Cumulative Probability of System Leakage

System Leakage Frequency (/yr)

0.95 1.0E-01

20.7 MPa 20.7 MPa

103.4 MPa 103.4 MPa

0.90 1.0E-02

1.0E-03
0.85
0.01% 0.10% 1.00% 10.00% 100.00%
0.01% 0.10% 1.00% 10.00% 100.00%
Leak Size (% Flow Area) Leak Size (% Flow Area)

Expert opinion used to select 3% of system flow area:

• captures >95% percent of the leaks
• the resulting separation distances protect up to the 3% leak size
• QRA performed to determine if associated risk from leaks greater than this
is acceptable
QRA Data, Models, and Assumptions
• Used leak frequencies from Bayesian analysis that
incorporates hydrogen-specific data
• Used AVT ignition probabilities
• Used Tsao and Perry Probit function
• Currently only includes random leakage events
(common to all facilities)
• No VCEs included in analysis (high momentum jets)
• No volume effects have been incorporated
(conservative)
• Surface effects not included (non-conservative)
• Assumes circular leaks (conservative)
 Accomplishment: Benchmark Risk Value for Gaseous
Hydrogen Station
• Developed draft report which assessed the risk of an H2FIRST
reference station using QRA and consequence-only analysis
• Will be integrated into H2FIRST as an appendix (SAND2015-2660R)

Cases Safety Calculation Baseline Result

Lot Line Perform QRA on H2First The PLL for this scenario
Separation reference station to is equal to 2.18E-05
Distance determine Potential Loss fatalities/system-year.
of Life (PLL) metric at 60
ft.
Parked Perform consequence The temperature at 30 ft.
Vehicle calculation to determine is close to ambient
Separation jet flame temperature at temperature.
30 ft.
Distance
Basis is 300 kg/day gaseous station

Demonstrating the calculation of benchmark risk values can be used for alternate
methods of code compliance.
31
 Approach: Key Barrier – Prescriptive LH2 Separation
Distances
• Current bulk distance values
– Based on historical values
– Present critical limitation to hydrogen infrastructure growth
• Science-based Code Improvements - Ongoing effort by NFPA
2/55 subcommittee to revise based on risk-informed science of
LH2 release behavior
• Alternative Methods for Code Compliance - In the meantime,
this effort is exploring a path forward for short term deviation
from separation distances for LH2

32
Progress: LH2 Informing Science-based Code Revisions

• Goal: Use QRA tools and methods to

revise bulk liquid hydrogen system
separation distances in NFPA 55/NFPA 2
• Progress:
– Multi-Party CRADA signed with Bki and Fire
Protection Research Foundation to enable
industry participation in support of LH2
model validation experimentation efforts
– Providing technical leadership and hydrogen
release behavior models to incorporate
current science and technology information
to risk-inform code requirements
– Details given in SCS-010 AMR presentation

Validated LH2 release model will be used to risk-inform

the revised LH2 bulk separation
33
 Technology Transfer Activities

• Technology transfer
strategies are tied to the
accessibility of HyRAM QRA
tool kit to other users
(AHJs, Station designers,
etc.) utilizing alternative
means of code compliance
• Refer to AMR SCS-011
presentation

34
 Summary
• Benchmark Risk:
– Addresses: Reducing barriers related to lack of technical data for SCS revision
– By: Identifying research gaps and developing scientific framework for
crediting hydrogen system safety features
• Alternate Means of Code Compliance
– Addresses: Education of AHJs,
– By: Validating and demonstrating alternative methods of code compliance
• Science-based Code Improvements
– Addresses: Reducing barriers related to lack of technical data for SCS revision
– By: Providing expertise to support science-based code revisions of bulk LH2
separation distances
• ISO TC 197
– Addresses: Harmonization with international codes
– By: Active technical leadership on working groups revising risk-based
methodology
35