Gas Detectors & Detectability

Copyright 2010 ISA. All Rights Reserved.

Gas Detectors and
Detectability
Simon J. O’Connor
Detector Electronics (UK) Ltd
Standards
Certification
Education & Training
Publishing
Conferences & Exhibits

Distributed with permission of author(s) by ISA 2010 Presented at ISA Automation Week 2010; http://www.isa.org


Presenter: Simon J. O’Connor

• ~30 years as a scientist working for Shell Research. Projects
included research into:
– Spark and compression ignition of gases; Open path FTIR; Flare
Spectroscopy; Open path Air Quality monitoring; Differential Absorption
LIDAR

• Last eight years as a Senior Consultant in the Fire and Gas
Detection group of Shell Global Solutions
– Over 50 projects on terminals; offshore facilities (manned and un-
manned); refineries; gas plants; chemical sites; FPSO and FLNG. Shell
and non-Shell customers.

• Joined Det-Tronics April 2010.

2


Introduction

• In this presentation we discuss the “detectabilty” of gas
releases.

• A brief introduction to the concept of “volumetric
coverage” and “Gas Detection Mapping” will be
presented.

• Finally, results from simple CFD analysis of releases are
presented and the correlation between these and
Mapping results are shown.

3


How good are existing systems?

UK Offshore hydrocarbon release database 2001-2008*
Method of detection
Human Smoke2 Heat2 Flame2 Gas Det. Process3 Hand- Total
observation held4
Major1 9 0 0 0 23 4 0 36
Significant1 34 1 1 1 18 2 1 58

1. See reference (below) for definitions
2. Original database includes smoke, heat and flame detectors for ignited events
3. “Process” – release noticed from changes in process monitors.
4. “Hand-held” – releases noticed during hand-held sniffing of components

* “Offshore hydrocarbon releases 2001 to 2008”; Alison McGillivray, John Hare, UK Health and Safety Laboratory
Research Report RR672; UK Health and Safety Executive, 2008.
4


How good are existing systems?

• Approximately 2/3 of “Major” and 1/3 of “Significant”
releases were detected by the fixed gas detection
system:

– Is this acceptable?
– Given the significance of human observation (25% of Major and
60% of Significant), will general reductions in staffing levels lead
to reduced detection?
– Can we identify best practice for fixed gas detection systems?

5


The nature of hazardous gas releases

• Liquid spills (including flashing liquids or liquefied gases):
Generally low level, heavy
Gas clouds and wind
dispersed plumes

• Pressurised gas releases:

Release direction dominates wind. Rapid air
entrainment improves dispersion

• Combinations of the above in two phase releases.

6


Tools of the trade: Gas detectors

• Point Gas Detectors:
– Measure the concentration of gas at the location of the detector.
May be flammable or toxic gases (Catalytic / Pellister detectors,
Electrochemical cells, Infra Red (IR) point detectors).
• Line of Sight (LOS) Detectors:
– Measure the integrated concentration of gas along a (generally)
IR beam.
• Acoustic Detectors:
– Listen for the characteristic noise associated with high pressure
gas releases.

7


Tools of the trade: QRA? Exceedence?

• Release frequency (failure rates) combined with 3D
dispersion and ignition models provides important
information about the risk associated with hazardous
gases.

• However:
– Resolution limited
– Link to “tolerance” unclear
– Can assess incorrect
parameters (average
detectability x average
mitigation efficiency)

• Accept tolerable cloud?

8


Consider a release from a site component

Site fence line
Hazardous zone
Point gas detector

9



Release in the “westerly”
direction interacts with site
structures. The jet slows,
this aided by additional
turbulence, forms an
accumulation cloud that is
detected.

10



Release in the “easterly”
exits the site undetected.

11



Larger releases do not
necessarily become easier
to detect; they form
accumulations further
from the source.

LOS ?

Acoustic ?
12


Pragmatic view of Gas Detection and
Mitigation

Below acceptable detection frequency
Liquid releases
Gas releases

Ineffective mitigation
Potential consequence

Effective range of application
of Fixed detectors

Event size

13


Volumetric coverage protection

• Following on from the Lord Cullen report after Piper
Alpha. Several studies were initiated to assess the
tolerable volumes of gas that could exist on offshore
facilities.
• The concept of a 5m diameter LFL equivalent sphere
was born.
• Detection target was to detect gas clouds before they
could achieve potentially damaging volumes and
escalate the situation.

14


Gas Detection Mapping

Uses a 3D model of a facility.

Releases “idealised” clouds from
all points.

Assess detector density.

Idealised sphere constructed to
allow for point or line detector

Produces a coverage “map” and
% coverage of zone to the idealised
target.

15


Assessing detectability: NIST 5 FDS*

Module 25x11x10m

Releases of 1.3kg/s
methane

*Fire Dynamics Simulator, Version 5.0, Building and Research Laboratory, National Institute of Standards
and Technology http://www.fire.nist.gov/fds
16


Flammable gas release in an enclosed,
partially enclosed and open module

T = 1s

T = 7s T = 18s
17



• 6 point gas and 2 LOS (4 LOS in open module)
simulated.
• 6 release directions for enclosed and partially enclosed
modules with one wind direction. 6 release directions
and 4 wind directions for open module. 36 releases in all.

Percentage of releases
detected by:
LOS Points
Enclosed 100 100
Partially Enclosed 83 50
Open 83 38

18



• Introducing one detector at a time, assessing coverage in
both CFD and volumetric Mapping software. Mapped
coverage shows a correlation to the detectability:

Enclosed Open Area

100% 100%

80% 80%
% D e te c te d

60%

% Detected
60%

40% 40%

20%
20%

0%
0%
0 25 50 75 100
0 25 50 75 100
Mapped Coverage Mapped Coverage

19


Toxic gas release on a process unit.

Release point Zone 60x40x10m

Releases of 10% H2S in
methane

X axis

Y axis

20



Five release rates, four wind directions; four release directions
(80 releases in total)
21



5m Grid
10m Grid

15
20

15
10

10
5
5

0 0
-20 -15 -10 -5 0 5 10 15 20 -30 -20 -10 0 10 20 30
-5
-5
-10

-10 -15

-20
-15

5m grid 10m grid

• Three grids of detectors were simulated; these were a
5m grid, a 10m grid and a 20m grid (not shown)
• The grids are unrealistic; they are simulated to test the
relationship between detector density and “detectability”

22



100
90
80 5m grid
Detectability (%) 70 10m grid
60 20m grid
50
40
30
20
10
0
0.002 0.004 0.006 0.7 1.3
Release rate (kg/s)

• Detectors initially set to alarm at 10ppm H2S.
Relationship between detectability, release rate and
detector density clearly evident.
23



39.0

38.0

Time to the first alarm (s)
37.0

36.0 5m grid

Response time
10m grid
35.0 20m grid

34.0

33.0

32.0
0 20 40 60 80 100 120
Detector alarm setting (ppm)

• Varying the detector alarm set point; the 5m grid at
100ppm alarm level, on average, responds faster than a
10m grid or 20m grid with detectors set at 10ppm.
24


Summary

• Guaranteeing high detection rates for hazardous gas
releases on industrial sites is not trivial due to the large
range of possible scenarios and release behaviour.

• A combination of detection technology generally provides
the best solution.

• Simple volumetric Mapping can be used to assess
conformity to target clouds in complex geometries.

• CFD modelling indicates that detectability is a strong
function of detector density (number of detectors) and
this correlates to volumetric Mapping.
25


Summary

• Detector sensitivity (alarm point) is secondary; if
detectors do not intercept the plume, the situation cannot
be corrected by reducing the detector setting.

• It is critically important to assess “upper limit” of
mitigation actions and not to use these events to justify
detection systems. Detection targets should align to
“lower level” tolerable clouds.

Thank you for your attention!

26

Gas Detectors & Detectability

More Related Content

Viewers also liked (8)

Similar to Gas Detectors & Detectability (20)

More from ISA Interchange (20)

Recently uploaded (20)

Gas Detectors & Detectability