Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No.

1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

RISK ASSESSMENT IN GAS DISTRIBUTION NETWORK LEAKS

USING THE PRELIMINARY RISK ANALYSIS AND EISENBERG
VULNERABILITY METHOD
Heracleto Carlos da Ana DOS SANTOS1; Guozhong HUANG2; Xuehong GAO3
Research article

Abstract: The natural gas distribution system is made up of an extensive network of pipelines
and some equipment such as regulators, valves, filters, meters and converters that
require maintenance to maintain operation and identify any gas leaks in advance.
The risk analysis process is widely used in the gas industries to locate faults in operations
and processes that could cause accidental release of chemicals, fire or explosion and
to provide decisions to improve operational risk safety. It was concluded that for risk
analysis (PRA) accidents occur due to lack of maintenance in the facilities and equipment
set and the Eisenberg Vulnerability Method provided an assessment of the probability of
damage to infrastructure and damage to the environment for 1%, 50% and 99% lethality.
Keywords: Risk Assessment, Natural Gas, Gas Distribution Networks, Probit Equation.

Introduction The central theme of this research will be

the identification of accident risks in gas distribution
Natural gas is the result of a process of degradation networks, using the Preliminary Hazard Analysis
of organic matter in deep reservoirs underground (PHA) technique. This technique is a widely
and can only be removed through perforations, disseminated tool in companies of various sectors as
just as it is done with oil in the depths of the sea. an instrument for assessing and controlling risks to
Among fossil fuels, it generates the lowest carbon health and safety at work and to the environment.
dioxide (CO2) emission rate, contributing greatly to Generally, the procedures related to prevention are
the reduction of the greenhouse effect thus generating centered on the formation of multidisciplinary teams,
less pollution in the atmosphere, moreover, it is with the presence of professionals from various
odorless and colorless (Marques, 2017). areas who jointly seek to establish consensus on
According to (Zimmermann, 2009; Araújo, the risks involved in each activity and recommend
2014), it is necessary to make correct decisions in the adoption of measures that prevent the occurrence
terms of prioritization and allocation of resources for of unwanted situations. To this end, it was applied as
monitoring and risk reduction, so it is necessary to a complementary method to the study, calculations
sort according to their nature and relevance because of the Eisenberg Vulnerability estimates by
all risk involves a possibility of gain or a chance of the Probit equation method to measure the estimates
loss. Pipeline operators must have an understanding of probability of damage of events exposed to
of the possible consequences of an accidental gas a certain type of consequence and given the magnitude
leak, to assist the risk management, and involvement, of an adverse effect such as: damage to humanity
to develop an appropriate code of laws and standards or death, collapse of infrastructure and damage to
for pipeline operations. But for this, it is necessary the environment. From the Eisenberg model it was
to have a deeper understanding of the characteristics possible to determine the overpressure peaks that
of the pipeline and the environment around it correspond to the probability of occurrence of each
(Junqueira, 2007). effect (NSTB, 1997).

1
University of Science and Technology Beijing, School of Civil & Resource Engineering, Beijing, China,
[email protected]
2
University of Science and Technology Beijing, School of Civil & Resource Engineering, Beijing, China,
[email protected]
3
University of Science and Technology Beijing, School of Civil & Resource Engineering, Beijing, China,
[email protected]

1
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

This last method is presented through the Probit based on the previous establishment of each of these
equations that correspond to the type of damage and categories (Frequency and Severity) (Bjerketvedt,
the peak of overpressure relative to the probability 1997).
of 1%, 50% and 99% of occurrence of the damage. Thus, each scenario of an accident is classified
These probability estimates serve to calculate into a frequency category. This frequency category
the percentage of a population exposed to a certain provides a qualitative indication of the expected
type of consequence of a given magnitude of an frequency for the occurrence of each identified
adverse effect, represented by a linear function, scenario, as shown in Table 1. This frequency
related to values to be measured (Papadakis, 1999). assessment can be determined by the experience
After this analysis it will be possible to identify of the team members, or by an accident database
and promote preventive actions in order to reduce (Sousa and Jerónimo, 2014).
existing risks, these factors contribute even more For the severity categories, each accident
to the care regarding the prevention of accidents, scenario is classified according to Table 2.
especially in the unit of distribution networks under The severity category provides a qualitative
study. Thus, this research is justified, mainly by indication of the degree of severity of the
the need to evaluate the risks in the work environment consequences of each of the identified scenarios
and also by the social relevance of this study, which (Sousa and Jerónimo, 2014).
aims to verify the forms of accident prevention in
The application of the as low as reasonably
the units of gas distribution networks (Mazzola,
practicable (ALARP) principle consists of defining
1999).
two levels of risk. The first consists of an absolute
level of risk "not permissible", which cannot be
Material and methods exceeded, regardless of the cost of containment
measures, according to (De Cicco and Fantazzini,
2015).
Preliminary risk analysis
The second is a level of risk that is considered
Dangerous events have been identified, these insignificant and therefore does not require
events are capable of giving rise to accidents in mitigating measures other than existing measures.
the analyzed events. Subsequently, the causes of each The table 3 shows the Risk Matrix, resulting from
of these events and their consequences or effects the combination of the Frequency and Severity
were identified, which depend on the evolution categories (Brito, 2007).
of the accident after its occurrence. In addition,
the qualitative evaluation of the frequency of
occurrence and their consequences was analyzed,

Table 1 Frequency categories of PHA scenarios

Category Denomination Description

A Extremely remote Extremely unlikely to occur during the life of the installation
B Remote Should not occur during the life of the installation
C Improbable Unlikely to occur during the lifetime of the installation
D Probable Expected to occur at least once during the life of the installation

Table 2 Severity categories of PRA scenarios

Category Denomination Description

I Despicable No injuries or deaths occur
The most that can happen are cases of minor first aid
II Marginal Minor injuries
III Criticism Moderate-severity lesions with a likelihood of death Requires
immediate corrective action
IV Catastrophic Causes death or serious injury in several people

2
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

Table 3 Risk Matrix

Severity
I (Despicable) II (Marginal) III (Criticism) IV (Catastrophic)
A (Extremely Remote) MR HR VHR VHR
Frequency

B (Remote) LR MR HR VHR
C (Improbable) VLR LR MR HR
D (Probable) VLR VLR LR MR

VHR Very High Risk

HR High Risk
MR Medium Risk
LR Low Risk
VLR Very Low Risk

Eisenberg vulnerability model Where:

Po is the peak of highest pressure (N.m-2).
Several effects resulting from a gas cloud
I is the radiation intensity (W.m-2) and t is the
explosion have been studied for many years
exposure time (s).
and can be classified for studies of estimates of
the consequences of toxic exposures, thermal Table below shows the Probit equations
radiation to fire and explosion overpressures. Probit collected from the literature for this study and, next
equations corresponding to the type of damage and to each one, the respective damage in which it will
the peak of overpressure relative to the probability be used. With the level of overpressure, the value of
of 1%, 50% and 99% of damage occurrence are the Probit variable is found and with it the probability
presented (AIChE, 2000). of occurrence of the damage (CCPS, 2014).
The effects caused by the passage of a shock
wave can be estimated by Eisenberg's vulnerability Table 4 Probit equations with their respective uses
model, a method that makes use of the Probit
equations for probability estimates that are used to Probit Equations Effects
calculate the percentage of a population exposed to y = -15,6 + 1,93ln(Po) Infrastructure collapse
a certain type of consequence of a given magnitude y = -77,1 + 6,91ln(Po) Damage to humanity (Death)
of an adverse effect. In mathematical terms, Probit is y = -12,6 + 1,524ln(Po) Damage to the environment
the development of probabilities represented through
a linear function related to values to measure, As can be seen in the Table 4, each type of
for example human damage, damage to infrastructure damage has an equation with different parameters;
and damage to the environment (Brito and Almeida, these were determined through observations made
2008). of accidents that have occurred throughout history.
The Probit equation can be written as below
proposed by (CCPS, 2014): Calculations of the probabilities
Y = K1 + K 2 ⋅ lnV (1) of consequences
Where: Several uncertainties are present in the estimation
Y is Probit, which is related to the percentage of of scenarios and danger zones, such uncertainties
the population subject to the consequence. k1 and are not desirable for the model, because it becomes
k2 are constants that are determined from historical impossible to determine deterministically what
data and V is the magnitude of the effect. multidimensional consequences may occur due to an
accidental leakage of natural gas (Henselwood and
 4
 Phillips, 2006).
 t⋅I3 
Po= V=  4 (2)

 10 
 

3
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

According to (Brito and Almeida, 2008), it can Probit equations corresponding to the type of
be considered, in the context of accidents in gas damage and the peak of overpressure relative to
distribution networks that the human, structural probability in percentage terms of occurrence of
and environmental consequences have negligible the damage were presented, which are now
correlation with each other, due to the fact that represented in the table 5.
the danger rays are a few tens of meters, and The vulnerability of the consequences
the combination of these dimensions of consequences originated by these phenomena are determined by
occurs randomly and independently according to the application of Probit-type methodologies,
the specific characteristics of each stretch, so that resulting in the estimation of the probability of
probabilities can be estimated independently. occurrence of a given damage, or percentage of
From equation 3 it is possible to determine affected factors, due to the radiation doses received
the peak of overpressure for any probability of and exposure times, according to the (TNO, 2004).
occurrence of damage in previously mentioned
scenarios, from 1%, 50% and 99%, (CCPS, 2014): Case study
 y1+ k 1 
−  The project under study is gas energy, at
Po( y1) = e k2 
(3)
the Temane Thermal Power Plant, which is located in
Studies of effects of a gas explosion to cause the Temane area, in the Inhassoro district, Inhambane
harm to mankind such as death, collapse of province, in southern Mozambique (MAE, 2014).
a masonry structure and damage to the environment It has geographical boundaries, to the north with
are represented by the equations below. the district of Govuro, to the east with the Indian
Equation 4 represents the Probit Equation, for Ocean, to the south with the districts of Vilankulo,
humanity damage or death: Massinga, Funhalouro and the west with the district
of Mabote. The district of Inhassoro has an area of
 y1+ 77.1 
−  4,746 km² and a population of 48,537, according to
Po( y1) = e 6.91 
(4) the preliminary results of the 2007 Census, resulting
in a population density of 10.2 inhabitants/km².
Equation 5 represents the Probit Equation for The population registered in 2007 represents an
infrastructure collapses: increase of 11.8% compared to the 43,406 inhabitants
 y1+15.6 
− 
recorded in the 1997 Census (MAE, 2014).
Po( y 2) = e 1.93 
(5) The Inhassoro District has two islands that are
part of the Bazaruto Archipelago National Park,
Equation 6 represents the Probit Equation for the island of Santa Carolina and Bazaruto Island.
damage to the environment: These islands are approximately 30 to 35 km away
 y1+12.6  from Inhassoro, especially the Santa Carolina
− 
Po( y3) = e 1.524 
(6) Integral Reserve, which is considered a take zone
(MAE, 2014).

Table 5 Relationship between Probit variable and probability

% 0 1 2 3 4 5 6 7 8 9
0 --- 2.67 2.95 3.12 3.25 3.36 3.34 3.52 3.59 3.66
10 3.72 3.77 3.82 3.87 3.92 3.96 4.01 4.05 4.08 4.12
20 4.16 4.19 4.23 4.26 4.29 4.33 4.36 4.39 4.42 4.45
30 4.48 4.50 4.53 4.56 4.59 4.61 4.64 4.67 4.69 4.72
40 4.75 4.77 4.80 4.82 4.85 4.87 4.90 4.92 4.95 4.97
50 5.00 5.03 5.05 5.08 5.10 5.13 5.15 5.18 5.20 5.23
60 5.25 5.28 5.31 5.33 5.36 5.39 5.41 5.44 5.47 5.50
70 5.52 5.55 5.58 5.61 5.64 5.67 5.71 5.74 5.77 5.81
80 5.84 5.88 5.92 5.95 5.99 6.04 6.08 6.13 6.18 6.23
90 6.28 6.34 6.41 6.48 6.55 6.64 6.75 6.88 7.05 7.33
99 7.33 7.37 7.41 7.46 7.51 7.58 7.65 7.75 7.88 8.09

4
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

(a) (b)
Figure 1 (a) Geographic location of the study area; (b) Location of the Study Branch at the Mozambican
Territory Level
Percentage
Percentage
The Pande and Temane project has a steel For the elaboration of the (PRA) matrix,
pipeline with an extension of 865 km and 26" all hazardous events whose failures originate
in diameter connecting (Temane to South Africa). in the installation under analysis were focused,
The study will be directed to the Mozambican Valves Valves contemplating both the intrinsic failures of
Pipes Pipes
territory, where it has 29% 29% equipment, instruments and materials, as well
a central gas processing
36%
36% 120
facility with system capacity: Mil GJ/year with as human errors. However, the definition of
an extension of 531 km on the Mozambican side the system and its components that fail can
(INP, 2005). compromise function, safety, ergonomics, as well as
The gas distributed by this extension is being those that have the highest failure rate, knowing now
used as fuel by Mozambique Aluminum-(MOZAL) Filters Filters that the system is the gas distribution network and its
Regulators 21% components are valves, filters, regulators, and pipes.
and as a primary energyRegulators 21%
source for the production
of electricity and for domestic
14%
14% consumption for The procedure for data collection and
the districts of Vilankulo, the Village of Inhassoro the preparation of the preliminary risk analysis was
and the BazarutoValves Valves
Filters
Archipelago, Filters
this in the province Regulators
Regulators
of Pipes
based onPipes
a focus group through interaction with
Inhambane. And later, it will be used by the Maputo employees of the operation area of a gas distribution
Iron and Steel Project and the power plants that refinery. Structured and semi-structured questions
supply power to the cities of Maputo, Matola and were elaborated in the form of a questionnaire,
Chokwe (MIREME, 2014). in order to obtain information on the theme under
study where the subjects were asked to answer
Results questions such as: What are the possible failure
modes for each item? What rating does each
failure mode give in relation to severity indices,
Presentation of the results of frequency and how would it be classified? To obtain
the preliminary risk assessment in gas the improvement actions, the topics addressed
distribution networks leaks were: "What are the possible improvements
aimed at reducing the risks found in each item?
At this stage, results will be presented The questionnaire and the survey of the answers can
using risk assessment methods depending on be seen in the annex to this document.
the identification of the hazards associated In the preliminary risk analysis, the hazards,
with leaks in gas distribution networks. During causes and effects or consequences and
the case study, descriptions of the initial stages of the corresponding severity categories were identified,
the procedure for the execution of the (PRA) will as well as the observations and recommendations
be made, which are fundamental: Identification of relevant to the identified hazards, and the results
causes, classification of frequency and severity to should be presented in a standardized spreadsheet as
achieve risk classification and propose mitigation can be observed below:
measures for leaks in gas distribution networks.

5
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

Table 6 Preliminary Risk Analysis for GAS Distribution Network

Preliminary Risk Analysis

Area: Subsystems: Date:
Hazard Cause Detention Effects/ Recommendation

Classification
Frequency
Mode Consequences

Severity
Small Leak Intrinsic failure of Odor Flammable cloud A I MR Periodic monitoring of
the safety valve Sensors dispersion with the system
possibility of fire
Pressure
drop on the Explosion in case of
network contact of
with the the gas with an
information ignition source
in the control
room
Medium Failure in material Odor Flammable cloud C III MR Maintain quality standard
leakage quality inspection Visual dispersion with of the materials to be
(pipes, valves, filters possibility of fire purchased for their use
Sound
and regulators) Explosion in case of Review emergency
Sensors
Opening the fuse contact of procedures
plug the gas with an
ignition source
Large Leak Metallurgical failure Odor Flammable cloud C IV HR Reevaluate tests for
Large corrosions Visual dispersion with determining structural
possibility of fire modes
Disruption of Sound
distribution channels Explosion in case
Sensors
natural gas contact
with an ignition
source
Catastrophic Domino effect Visual Possibility of fire C IV HR Review operational
disruption occurrence Sound and explosion in procedures
case of immediate
Sensors
ignition at
the beginning of
the leak
Gas leak in Improper installation Odor Fire or explosion in B II MR Change the Valve
the valve Defective valve Sensors case of immediate Management of coating
ignition at projects, cathode
Corrosion
the beginning of protection, maintenance
the leak and inspection plan)
Atmospheric Damage Prevention Plan
dispersion of
Patrolling the entire route
flammable gas
of the network
Monitoring of valve
operating parameters
(temperature, pressure
and flow) in order
to reduce the time to
identify larger leaks
City gate in fenced area,
with access control
(gate closed access) and
risk signaling

6
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

Preliminary Risk Analysis

Area: Subsystems: Date:
Hazard Cause Detention Effects/ Recommendation
Mode Consequences

Classification
Frequency

Severity
Gas leak in Damaged plug Odor Fire or explosion in B IV VHR Change the valve or fix
valve fuse Sensors case of immediate the plug
plug ignition at City gate in fenced area,
the beginning of with access control
the leak (gate closed access) and
risk signaling
Gas leak in Damaged pipe Odor Fire and fire jet, in C IV HR Change the pipe
the pipe Thermal weakening Sensors case of immediate Management of coating
(by exposure to ignition projects, cathode
excessive thermal Atmospheric protection, maintenance
radiation) or dispersion of plan and inspectio
mechanical material flammable gas Damage Prevention Plan
Excessive pressure Provision of registration
increase in the and monitoring of works
pipe (by increasing
Patrolling the entire route
temperature in case
of the network
of prolonged system
failure without Specification of
drainage of the line) materials, qualification of
suppliers
Weakening of
welding or thread Monitoring of the
sewing operational parameters of
the duct resulting in
the reduction of the time
to identify larger leaks
Disruption Third-party action/ Odor Fire or explosion C IV HR Change the pipe
of the gas interference Sensors in case immediate Management of coating
transport Corrosion ignition in the projects, cathode
Visual
pipe beginning of the protection, maintenance
Material failure
leak plan and inspection)
Incorrect
construction Damage Prevention Plan
Geotechnical Provision of registration
movements and monitoring of works
Weather Network patrol
Mechanical impacts Specification of
materials, qualification of
Thermal weakening
suppliers
(by exposure to
excessive thermal Monitoring of
radiation) or the operational
mechanical material parameters of the duct
resulting in a reduction
Excessive pressure
in the time to identify
increase in the
larger ruptures
pipe (by increasing
temperature in case
of prolonged system
failure without
drainage of the line)

7
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

Preliminary Risk Analysis

Area: Subsystems: Date:
Hazard Cause Detention Effects/ Recommendation

Classification
Mode Consequences

Frequency

Severity
Gas leakage Inadequate Odor Fire and Fire Jet, in D II VLR Change of filters and
through the installations Sensors case of immediate auxiliary components
filters Filter damaged or in ignition in the start Coating design
trouble of the leak management,
Clogging of filters Atmospheric maintenance plan and
dispersion of inspection
flammable gas Damage Prevention Plan
Acceptance tests
Clogging Inadequate facilities Odor Formation of A IV VHR Immediate filter change
of filters by Accumulation of Sensors hydrides and acids, Clogging management
impurities impurities dragged both of which will (design, coating, filter
Visual
from production attack the pipe protection, maintenance
wells and during causing corrosion plan and inspection
transport by in them
Damage Prevention Plan
the distribution pipe Reduction of gas
flow during its
circulation
Gas leak in Defect in regulator Odor Fire and fire jet, in C II LR Immediate exchange of
regulators Disruption of Sensors case of immediate regulators
due to the regulator due to ignition Specification of
Pressure
increased increased pressure of Atmospheric materials, qualification of
drop in the
pressure released gas dispersion of suppliers
distribution
Third-party action/ network with flammable gas Anti-corrosion
interference information management (design,
Corrosion checked in coating, cathodic
the control protection, maintenance
Material failure
room and inspection plan)
Monitoring of the
operational parameters
of the regulator
(temperature, flow and
pressure)

Presentation of the results of The peak of overpressure was determined from

the equations presented in Table 4 using as Probit
Eisenberg vulnerability estimates variable the value corresponding to the probability of
For the validation of the collected data, it was 1%, 50% and 99% for the effects related to damage
necessary to present calculations of Eisenberg's to humanity or death, collapse of infrastructure
Vulnerability estimates by the Probit equation and damage to the environment. The results were
method to measure the estimates of probability supported by a comparison between the result of the
of damage of events exposed to a certain type of general objective of the article, in relation to that
consequence and a given magnitude of an adverse found in the bibliography, in order to obtain reliable
effect such as: damage to humanity or death, information, avoiding harming the research result.
infrastructure and damage to the environment.
From the Eisenberg model it was also possible to
determine the overpressure peaks that correspond to
the probability of occurrence of each effect.

8
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

Application of the calculation of overpressure Application of the calculation of overpressure

necessary to have 1% probability of damage to necessary to have 50% probability of damage to
humanity or death (Equation 4): humanity or death (Equation 4):
Data Formula Data Formula

K1 = −77.1 = −77.1
 y1+ k 1   y1+ k 1 
K 2 = 6.91 Po (1% ) = e
−
 k2 
 6.91 −
Po ( 50% ) = e k2 


(Y = 2.67 table 5) ( 5.00 table 5)

Resolution Resolution
5.00 + 77.1 
Po (1%=) e  −
2.67 + 77.1 

Po ( 50%=) e  − 
 6.91   6.91 
Po (1% ) = 1, 0317.105 Pa Po ( 50% ) = 1, 44542.105 Pa

Application of the calculation of overpressure Application of the calculation of overpressure

necessary to have a 1% probability of infrastructure necessary to have a 50% probability of infrastructure
collapses (Equation 5): collapses (Equation 5):

Data Formula Data Formula

K1 = −15.6 K1 = −15.6
 y1+ k 1   y1+ k 1 
K 2 = 1.93 Po (1% ) = e
−
 k2 
 K 2 = 1.93 −
Po ( 50% ) = e k2 


(Y = 2.67 table 5) (Y = 5.00 table 5)

Resolution Resolution
2.67 + 15.6  5.00 + 15.6 
Po (1%=) e  −  Po ( 50%=) e  − 
 1.93   1.93 
Po (1% ) = 1, 2917.104 Pa Po ( 50% ) = 4,3199.104 Pa

Application of the calculation of overpressure Application of the calculation of overpressure

necessary to have 1% probability of damage to necessary to have 50% probability of damage to
the environment (Equation 6): the environment (Equation 6):

Data Formula Data Formula

K1 = 12.6 K1 = −12.6
 y1+ k 1   y1+ k 1 
K 2 = 1.524 Po (1% ) = e
−
 k2 
 K 2 = 1.524 −
Po ( 50% ) = e k2 


(Y = 2.67 table 5) (Y = 5.00 table 5)

Resolution Resolution

2.67 + 12.6  5.00 + 12.6 

Po (1%=) e  −  Po ( 50%=) e  − 
 1.524   1.524 
Po (1% ) = 2, 2464.104 Pa Po ( 50% ) = 1, 03627.105 Pa

9
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

Application of the calculation of overpressure Discussion

in necessary to have 99% probability of damage to
humanity or death (Equation 4):
Preliminary risk analysis associated with
Data Formula
leakage in gas distribution network
After studies carried out throughout the research,
K1 = −77.1
 y1+ k 1  by carrying out the preliminary risk analysis
K 2 = 6.91 Po ( 99% ) = e
−
 k2 

in accordance with the literature consulted on
(Y = 7.33 table 5) the detection of natural gas leakage, which addresses
what are the equipment that make up the gas
Resolution distribution system as a way to guarantee the safe
distribution of the even approached by (Marques,
7.33 + 77.1 
Po ( 99%=) e  −  2017), it was found that the gas distribution process
 6.91  is composed of components such as Valves, Filters,
Po ( 99% ) = 2, 02505.105 Pa Regulators and Pipes, being part of it, since gas
distribution systems are considered the main critical
causes for the occurrence of leaks in the distribution
network. The critical causes were only possible to
Application of the calculation of overpressure identify taking into account the exact and concise
necessary to have a 99% probability of infrastructure description of the function of each component
collapses (Equation 5): provided by the questionnaire prepared for a group
of experts in the area the so-called focus group.
Data Formula The functions are very important, as they are a great
reference for identifying the dangers present in an
K1 = −15.6 installation, which can be caused by undesirable
 y1 k 1 
K 2 = 1.93   events.
Po ( 99% ) e 

(Y = 7.33 table 5) The Graph 1 shows the main cause of leaks

in gas distribution networks, taking into account
Resolution the preliminary risk analysis carried out throughout
the research, showing that pipes are responsible for
7.33 + 15.6 
Po ( 99%=) e  − 
leaks.
 1.93 
Percentage
Po ( 99% ) = 1, 44470.105 Pa

Application of the calculation of overpressure Valves

Pipes
necessary to have 99% probability of damage to 29%
36%
the environment (Equation 6):

Data Formula
Filters
Regulators 21%
K1 = −12.6 14%
 y1+ k 1 
K 2 = 1.524 Po ( 99% ) = e
−
k2 


(Y = 7.33 table 5) Valves Filters Regulators Pipes

Resolution Graph 1 Percentage of critical causes of leaks

in gas distribution networks
7.33 + 12.6 
Po ( 99%=) e  − 
 1.524 
According to (Corteletti, 2008), This is due to its
Po ( 99% ) = 4, 78029.105 Pa exposure and sensitivity along the path that comes
into contact with third parties by excavations for

10
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

the installation of rainwater networks, drainage Based on the results obtained for the 99%
networks of sewers on the track, electrical probability of damage occurring by estimating
installation networks, renovations or maintenance, Eisenberg's vulnerability model and the Probit
mining, design error and welding errors, lamination, equation, it can be concluded that the most impact
inflexible points are seen as reasons for their shock wave was linked to the scenario of damage
occurrence, and one also sees allies extreme factors in the environment and a slight effect was verified
outside human control such as tropical storms, for the scenario of damage to humanity or death,
cyclones, tsunamis, falling trees, soil movement or with values around 4.78029.105 Pa and 2.02505.105
erosion, floods. Pa respectively. (TNO, 2004; Brito and Almeida,
(EGIG, 2018) supporting Laws 13.303/2016 2008) argue that shock waves for gas explosions
and 8.666/1993 mentions that to avoid operational, at these levels pose risks to the environment and
mechanical and third-party failures, it is necessary humanity. Overpressure spikes can cause damage
to hire specialized and legalized companies to carry due to emitted thermal radiation, a factor that reports
out the activity providing data in forms of drawings a fatality in terms of large-scale vegetation burn,
of the tracks to be built and the type of material animal death and pollution, first and second degree
that will be adopted in the project, from the process burn checks, tympanum ruptures, and even can
of capturing gas in reservoirs to the distribution cause immediate death.
channels according to the purpose of the product.
That way, the companies in turn should evaluate Conclusion
alternative routes for gas distribution, which
This chapter presents the results of
environmental areas and population centers can be
the Preliminary Risk Analysis and the effects of gas
crossed; and what other infrastructure systems are
cloud explosions based on Eisenberg's vulnerability
along the route.
method in a gas distribution network, as well as
recommendations.
Calculation of Eisenberg vulnerability
The main objective of this work was to use risk
probabilities using the Probit equation analysis techniques to estimate the explosion levels
By estimating Eisenberg Vulnerability model of where shock in case of a possible natural gas
and Probit equation, the probability of 1% of damage leak for the Pande and Temane steel gas distribution
occurred, it was found that the shock wave of an network with an extension of 865 km and 26" in
explosion from a gas leak had a greater impact on diameter connecting (Temane to South Africa),
the scenario of damage to the environment, where specifically directed to Mozambican territory with
the impact value was approximately 2.2464.104 Pa. an extension of 531 km. The gas distributed by this
According to (TNO, 2005), the peak of overpressure extension is being used as fuel by Mozambique
at these levels offers a higher risk of damage Aluminum (MOZAL) and as a primary energy source
verified at the level of proximity and contact with for electricity production and internal consumption,
the environment, providing burns to vegetation, in order to assess whether throughout the installed
animal deaths and even reached the point of network offers risks in the facilities in order to be
emission of air pollution. And for other scenarios, likely to collapse infrastructure, damage to humanity
such as damage to humanity and the collapse of and damage to the environment.
infrastructure due to factors such as reaction time After identifying the risk and analyzing
and mobility to reduce the risk doesn’t come to be the effects of this risk, the results of this work
affected to the point of causing risk along the gas showed, most accidents in gas distribution networks,
distribution network. occur due to lack of maintenance in installations and
Therefore, for the results obtained by estimating the set of equipment that cause gas leaks in the valve
Eisenberg's Vulnerability model and Probit equation, fuse plug, gas leaks by filters, clogging of filters
the probability of 50% of damage occurrence, it by impurities, gas leakage by regulators due to
was noted that for the collapse in infrastructures increased pressure and catastrophic pipe disruption,
had a shock wave explosion impact level around failure in the preparation of risk analysis, lack of
4.3199.104 Pa. Value that according to (TNO, 2005; information, either due to lack of interest of the
Henselwood Phillips, 2006), mention in being in the employee, or lack of training, adequate supervision
values of higher risks, this due to the impact that in relation to the equipment installed in its network
reaches to affect high financial losses from damage and also regarding bottling, handling, transportation
to properties of the distribution network itself, losses and storage of gas.
of third-party properties, payment of compensation.

11
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

And through mathematical calculations and with regard to the safety of the workers who remain
with the aid of Eisenberg's vulnerability model there. It is necessary to have a maintenance plan,
and the Probit equation, it was determined that by being attentive to the fulfillment of maintenance
the probability of 1% of damage the occurrence deadlines, it is also important for the safety of
of damage from an explosion of a leak in the gas users, that the installation of the gas distribution
distribution network had greater impact on the network has evaluation of a professional and gas
scenario of damage to the environment, where the arrest equipment and automatic alarms in sufficient
impact value was approximately 2.2464.104 Pa. quantity, because the gas leak can be silent and
For the probability of 50% of damage occurrence, extremely dangerous.
it was noted that for the collapse in infrastructure For the purposes of the safety criteria in
had shock wave explosion impact level around the installation and operation of the storage bases
4.3199.104 Pa and finally for the probability of up to the gas distribution networks, it should be
99% of damage occurrence it can be concluded established that any installation involving handling,
that the shock wave of greater impact was linked to handling and storage must have intrinsic safety
the scenario of damage in the environment and measures provided for already in the design phase of
a slight effect was verified for the scenario of the plant or equipment. Some of these main measures
humanity with values around 4.78029.105 Pa and are the installations of diameter pipes suitable for
2.02505.105 Pa respectively. the working pressure, the locations of large capacity
installations and too many branches in areas
Recommendations adequately detached from urban conglomerates,
shopping centers and the use of specific electrical
It proposes that more modern constructions of equipment and installations for areas subject to
facilities should be designed in order to optimize the formation of explosive air and gas mixtures.
the control conditions of the process unit and also

References
AICHE, 2000. Guidelines for chemical transportation risk analysis. Center for chemical process safety of
the American institute of chemical engineers. Second edition. New York: John Wiley & Sons Ltd publiser.
Araújo, D. 2014. Risk assessment in buildings with internal distribution networks of combustible gases. Using
electre tri. St. Joseph of The Fields. Salvador-Brasil: Qualitymark publisher.
Bjerketvedt, D. 1997. Gas explosion handbook. Journal of Hazardous Materials, 52: 1-150. DOI:https://doi.
org/10.1016/S0304-3894(97)81620-2.
Brito, A. 2007. Multicriteria risk assessment in natural gas pipelines: a review classification approach with electre
tri. 2nd edition. Recife: Federal University of Pernambuco.
Brito, A. J., Almeida, A. T. 2008. Multi-attribute risk assessment for risk ranking of natural gas pipelines. Reliability
Engineering & System Safety, 94: 187-198. DOI:https://doi.org/101016/j.ress.2008.02.014.
CCPS, 2014. Guidelines for evaluating the characteristics of vapor cloud explosions, flash fire, and bleves.
American institute of chemical engineers. Second edition. New York: John Wiley & Sons Ltd publiser.
Corteletti, C. 2008. Comparative analysis of environmental impacts for the design of pipeline scans and transmission
lines, 53-64. In Porto de Galinhas (Org.): Disasters in Paraíba: risks, vulnerability and resilience. PB, João
Pessoa: Editora Universitária da UFPB. ISBN: 978-85-237-0590-9.
De Cicco, F., Fantazzini, M. 2015. Modern risk management techniques. Sao Paulo: Anais Virtuais Ltd.
Gas pipeline incidents [online]. EGIG, 2018 [cit. 2023-01-20]. Available at: www.egig.nl
Henselwood, F., Phillips, G. 2006. A matrix-based risk assessment approach for addressing linear hazards such as
pipelines. Journal of Loss Prevention in the Process Industries, 19: 433-441. DOI:10.1016/j.jlp.2005.10.005
Licensing Regulation for Oil Infrastructures and Operations [online]. INP, 2005 [cit. 2023-01-20]. Available at:
Mzbusiness.com
Junqueira, A. 2007. Democratization of liquefied petroleum gas and sustainable development. Sao Paulo. Thesis
(doctorate). São Paulo State University.
MAE, 2014. Profile of the District of Inhassoro. Mozambique. Mozambique. Plural Editor. Mozambique.
Marques, L. 2017. Detection of Natural Gas Exhausts. Rio de Janeiro: Editorial Operational Research.

12
Transactions of the VSB - Technical University of Ostrava Vol. XVIII, No. 1, 2023
Safety Engineering Series, ISSN 1805-3238 pp. 1 - 13, DOI 10.35182/tses-2023-0001

Mazzola, A. 1999. Thermal Interaction Analysis in Pipeline Systems: A Case Study. Journal of Loss Prevention
in the Process Industries, 12: 495-505. DOI:https://doi.org/10.1016/S0950-4230(99)00022-4.
Policies and legal regime of activities [online]. MIREME, 2014 [cit. 2023-01-20]. Available at: Mzbusiness.com
NSTB, 1997. Protecting Public Safety through Excavation Damage Prevention. NTSB/PA-00/01.
Washington, D.C.
Papadakis, A. 1999. Major hazard pipelines: comparative study of onshore transmission accident. Journal of Loss
Prevention in the Process Industries, 12: 91-107. DOI: https://doi.org/10.1016/S0950-4230(98)00048-5
Sousa, P., Jerónimo, M. 2014. Preliminary analysis of environmental risks for activities related to drilling of a land
oil well. 1st edit. Sao Caetano do Sul: University Center of the Mauá Institute of Technology.
TNO, 2004. Methods for the determination of possible damage to people and objects resulting from releases of
hazardous materials. The green book: The Netherlands committee for the prevention of disasters. The director-
general of labor. Apeldoorn: Committee for the Prevention of Disasters (CPR 16e).
TNO, 2005. Methods for the calculation of the physical effects of the escape of dangerous materials: liquids and
gases. 3rd edit. Apeldoorn: Netherlands organization for applied scientific research.
Zimmermann, A. 2009. Risk analysis of a natural gas leak in a pipeline. Santa Catarina. Masters dissertation.
Federal University of Santa Catarina/UFSC. Technological Center. Department of Chemical Engineering and
Food Engineering.

13