Safety and reliability enhancement throughout Europe

50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain

Practical application of criticality analysis in the Spanish
Natural Gas Transport Network

Javier Serra Parajes

Asset Management Technician, Enagas
Paseo de los Omos 19
28005, Madrid, Spain

Adolfo Crespo Márquez

University of Seville, Spain

Abstract

Risk management is emerging as one of the fundamental pillars for asset

management in modern industry. The appearance of ISO 55001 as well as the recent
revision of ISO 9001 introduced this concept as the best basis to make decisions in
operation and maintenance.
These regulations do not require the application of a specific technique, so
companies are responsible for searching the methodology that best meets their
needs. In this case, companies do not only pursue that results provide a clear and
relevant information about the risk of equipment, but obtaining these results within
a reasonable time and of a large amount of equipment, because in some facilities it
could be analysed thousands of items.
The aim of the exhibition is to expose the real case of the application of a specific
methodology, criticality analysis, in a particular installation of OIL & GAS industry, a
gas transport network. It is not only about reviewing the theoretical concepts that
sustain the methodology and the benefits for mass analysis in an optimum time, but
to discover those aspects that occur when an real analysis is developed and how may
influence the result in a significant way.

Keywords: Maintenance Strategies, Asset and Maintenance Management, Decision

Support, Criticality Analysis, Risk management

1. Introduction
In this paper, it is shown a real development of a criticality analysis for maintenance
purposes as a base for different working lines of operation and maintenance. The main
target of the methodology is to prioritize the equipment depending on the consequences of
a hypothetical failure. The adaptation of the theoretical methodology to make it confluence
with the company strategy provides to the user an analysis of the relative importance of the
equipment.
The study of the consequences of a functional loss and the frequency of these failures let
get us closer to the concept of what equipment is more valuable for the company. The main

1
Safety and reliability enhancement throughout Europe
50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain

advantage of the methodology is to allow the analysis of a big amount of items into a
limited time. That is the reason why is so powerful as a base for developing a maintenance
management model in a company with a big number of assets.

2. Theoretical model adaptation

The target of the paper is not to show the theoretical model adaptation. In order to simplify
the reading, it is just going to be described the result of the adaptation. Just to remark that
this phase must be done as closer to the strategy of the company as possible. This
methodology must be a tool that allows to understand the relative importance of one item
for the company. Probably it is more usual to make partial assessments and analyse the
importance of an item for a specific area of the plant; safety area, maintenance area,
operational area…

If the analysis criteria are defined close to the global company strategy, the result will be a
global and common assessment of a specific item for all the areas of the company. The
target of criticality analysis is to prioritize assets evaluating its relative importance. The
criticality concept is defined as the product of the failure frequency of an item multiplies by
the possible consequence of a functional loss:
Criticality = Frequency failure * Consequence
(CTR = FF *C)

a. Criteria definitions
To define criteria to assess functional loss, most of theoretical models propose two main
concepts; criteria related with cost and criteria related with safety. That’s the reason why it
has been used the Asset Management Policy of the company as a base for criteria definition.
This policy is sustained in two main concepts that involve every working line that the
company is developing about operation and maintenance. The first base is “integrity”. In
this concept are included definitions as personal safety, industrial security and
environmental care. The second base is “Efficiency and Improvement” and involves
concepts as availability, quality service and maintenance costs.
To connect the criteria proposed by the methodology with the asset management policy of
the company, five analysis criteria are defined based in these two bases. Two of them are
related to the integrity and the other three are related to the efficiency. It is important to
remark that criteria related with costs don’t directly imply “spend money”, even “profit lost”
or “production lost”. They can be related to reputational lost, stakeholders repercussion or
even hypothetical penalties for service loss.
The criteria defined for consequence analysis are:
Safety Criteria:
• Industrial safety: The industrial safety factor assesses the consequences of the
functional loss of an element related to:
o Injuries to internal or third party personnel in the facility, and/or any other
person who could be involved in.
o Damage to of industrial assets, products and materials used in production or
in end products, either in its own or third party facilities

2
 Safety and reliability enhancement throughout Europe
50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain
• Environmental: The environmental factor assesses the environmental consequences
of the functional loss of an element, including recovery costs, penalties,
compensation, etc.
Cost criteria:
• Quality service: The quality service factor assesses the impact of the functional loss of
an element on the gas reception, delivery service conditions, and any other services
that Enagas offers to its clients.
• Availability: The availability factor assesses the impact on the installation’s nominal
capacity of the functional loss of an element It matches with the design capacity
(emergency or reserve equipment not included).
• Maintenance costs: The maintenance cost factor assesses the impact of the functional
loss of an element on the corrective maintenance costs, including costs associated
with the recovery of the equipment and other equipment that may have been
damaged.
Every criterion has a specific weigh in order to change subjective opinions of technicians into
a numeric mark. When a consequence is analysed for each criterion, four severity levels are
defined, also with its weigh depending of the theoretical consequence.
The value table defined for the assessment is shown in table 1.
Industrial Safety Environmental Quality Service Availability Maintenance Costs
(35%) (15%) (25%) (20%) (5%)
Catastrophic 100 High 100 High 25 Very High 20 Very High 5
Critical 35 Medium 15 Medium 15 High 10 High 4
Moderate 20 Low 5 Low 5 Medium 5 Medium 3
Slight 0 No Impact 0 No Impact 0 Low 0 Low 1

Table 1: Criterion and severity weighs

b. Frequency failure definitions

Most extended models define four levels (low, medium, high and very high) for frequency
failure definition. In this case, it has been defined assuring that this definition follows the
real management developed in the company.
Specialist of the company agrees that they use these four definitions to classify failures
because of its frequency:
• Possible failures; an average value lower than one failure every two years
• Acceptable failures; an average value of one failure between two years and one year
• Repetitive failures; an average value between one and two failures per year
• Non acceptable failures; an average value higher than two failures per year

3. Methodology application scheme

To reach a correct development of the methodology is important to be organized and follow
the right steps of the technique as strictly as possible. For this reason, it is going to be
explained a practical example of a criticality analysis in an item of a gas facility.

3
Safety and reliability enhancement throughout Europe
50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain

It has been defined a diagram with the main steps of the technique in order to simplify the
methodology application. In the diagram are described key rules to develop a properly
analysis. It is very important to analyze the item into its physical and operational context.
The scheme proposed is:

a) To define the facility and its functionality: the aim is to get a whole vision of the
facility in which the item analysed is working. Therefore, some analysis criteria are
related to the whole plant. For example availability or quality service are not
analysed for each item, but for the whole facility. It is very important to have
defined the plant properly to make a correct assessment of the criticality.
b) To define the system function: Focussing in the system, it must be analysed the
concrete function of the item into the facility (emergency system, control
system…). It is important to know the real importance of the item for the correct
operation of the plant.
c) To define the item function and its operational context: The final target for
developing the criticality analysis is to analyse the consequence of a functional
loss. That is the reason why this function must be defined: to pump, to close, to
compress…
d) To define functional loss: Having the item function defined is easy to suppose the
functional loss. The key point in this step is to differentiate a fault to a functional
loss. Every item has multiple possible faults, but just a single functional loss for
every function of the item. If the function of an item is “to pump”, its functional
loss will be “pumping absence”.
e) To analise failure consequence: The consequences of a hypothetical functional
loss will be assessed in this step. Using the table defined in the previous chapter,
it must be selected the consequences of a hypothetical failure in every
consequence criteria.
f) To get the severity of a functional loss: Every criteria consequence has a numeric
value that reflects the importance of the consequences for the company. In this
step, these numeric values are collected in order to get a final severity value.
g) To calculate the frequency failure: Depending on the definition made in the
previous chapter, the frequency failure of every item must be calculated in order
to get the second parameter needed for the severity.
h) To get the criticality level: We can obtain the criticality level using the frequency
failure and the severity level calculated in the previous steps.

4. Practice application in a gas facility

It is going to be described a simple example of the methodology application. The previous
application scheme is going to be followed step by step trying to highlight the key aspects of
the analysis.

a) To define the facility and its functionality

The facility that is going to be analyzed is a valve point that provides gas to a “Measurement
and Regulation Station”. In the natural gas transport network, the NG is transported at high
pressure (72 bars) across the country, but in order to deliver NG to clients, the pressure of
the gas is reduced to 16 bars in MRS.

4
 Safety and reliability enhancement throughout Europe
50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain
The functions of the valve point are:
- To section the pipeline in case of leaking gas
- To deliver NG from high pressure network to the MRS.

In this analyzed system, items related to the process are mostly valves. There are other
functions made by other items of the facility related mostly with safety or control, but they
are going to be omitted in order to simplify the example.

Fig. 1 : scheme of the facility

We can show the facility scheme in the figure 1. The example will be focused on valve “a”.

b) To define the system function:

As it has been exposed previously, in the example we have focused the analysis in the valves
system, so in this case, the system function is the same that the facility function. It is going
to be described de valve function:

• Valve “a” is a “sectioning valve” and its function is to cut the NG flow through the
pipeline in case of leaking.

5
Safety and reliability enhancement throughout Europe
50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain

Fig. 2 : System function

c) To define the item function and its operational context

The valve is motorized. In case of valve “a” the function is usually opened to let the flow of
NG through the pipeline. So the valve has two functions, to be opened in normal operation
and to be closed and to cut the NG flow if a leak is detected. With the second function we
can section the pipeline and to limit the damage derived from a possible incident.

d) To define functional loss

Criticality analysis is a methodology that doesn’t analyze the hypothetical failure mode
(breakdown, misalignment…) otherwise it assesses the consequence of a functional loss
independently of the cause of the failure.
That is the reason why is so important to define properly the functional loss of every
maintainable item. Moreover, each item has a main function (to pump, to compress…) but
have also secondary functions (proper pressure, efficiency point…). In order to simplify the
analysis, it is important to define just the main functions. It could be studied the functional
loss over all the main and secondary functions, but one of the key points of the
methodology is to analyze a big amount of items in a short period of time. If every function
is included into the analysis, it could delay the development of the project.
So in this case, there will be only analyzed the main functions: “failed to open”, “failed to
close”.
e) To analise failure consequence:
To develop this step properly, is important to be strict with the hypothetical situation and
the correct question that let us to assess the consequences of a functional loss. The correct
question that must be done is “What would be the consequence of a valve “a” failure, in
case it would be needed?

It is important to remark that the functional loss must be assessed in the moment that its
operation is required, so the consequences are negative for the facility. For example, the

6
 Safety and reliability enhancement throughout Europe
50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain
analysis of a gas detector failure doesn’t have any sense if it is not supposed a gas leak.
Without gas presence, a gas detector failure never would have consequences. The right
question for this example would be: “What would be the consequence of a gas detector
failure, in case of a gas leak?

It is also interesting to note that concatenated failures are not analyzed in criticality analysis.
It means that in order to assess a consequence failure, it must be only supposed the failure
of the analyzed item and not hypothetical failures of other related equipment. In the end,
criticality analysis assesses the consequence of an item functional loss for every criterion
defined previously. It implies to adapt the question that must be done for every criterion:

-“What would be the industrial safety consequence of a gas detector failure, in case of a gas
leak?
-“What would be the environmental consequence of a gas detector failure, in case of a gas
leak?
-…

The hypothetical situation estimated for the analysis is a pipeline leak. There must be
analyzed the possible consequences of a functional loss of the valve in the situation
described. As it has been defined previously, the valve has two main functions: to open to
allow the natural gas flow across the pipeline, and to close to cut this flowing. When two or
more functions are analyzed, it must be taken the highest mark of every assessment. In this
case the most critical function is a “close failure” so is the function that will be analyzed.

Industrial safety; the target of the valve is to cut the pipeline section that has benn
damaged and in consequence, to limit the natural gas flow across the pipeline avoiding risks
related with the leak. If the valve could not be closed, there will be a high risk becouse
natural gas would being emitted to the atmosphere and in consequence there will be
explosion risk. The consequence of a valve functional loss on industrial safety criterion will
be “catastrophic”.

Environmental; the assessment of this criterion is related with the natural gas emitted to
the atmosphere. With the same reasons than the industrial safety analysis, if the valve could
not be closed, natural gas would be flowing to the atmosphere. In this case the consequence
is assessed as “Low” because the environment impact does not implies third parties.

Quality service; In this analysis, it must be assess the consequence of the valve function loss
related to the clients. In this case, the function of the valve point is to allow the natural gas
flow across the pipeline and to deliver natural gas to the Measurement and Regulation
Station.This is a clear example that shows the importance of the operational context of the
analysis. A valve failure would not cut the natural gas delivering. In fact, the real problem is
that the natural flow could not be cutted. But the operational context define that if a valve
can not be closed, it must be closed the nearest valve. In this case the natural gas flow
would be cutted and the delivering to the MRS woud be interrupted. According to the
severity levels definition, the consequence of this failure would be “high”.

Availability; In this criterion, it must be analysed if the valve functional loss would imply a
facility function loss. The facility has two main functions (defined previously). In this
hypothetical situation, a valve functional loss would imply the loss of both functions; to cut

7
 Safety and reliability enhancement throughout Europe
50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain

the natural gas flow across the pipeline and to deliver natural gas to MRS. According to the
severity levels definitions, the consequence of the failure would be “very high”.

Maintenance costs; this is probably the most objective criterion because it must be
estimated the costs related to the hypothetical failure. The technicians agree the most
common valve’s failures and then estimate an average price for the reparation. In this case
it has been estimated that the cost would be between 600€ and 5000€ what means that the
consequence failure related to this criterion would be “medium”.

f) To get the severity of a functional loss:

The next step is to obtain the mark of the consequences assessed in the previous chapter in
function of the table defined by the technicians. The final table is the next:
Environmental Availability Maintenance Costs
Industrial Safety (35%) Service Quality (25%)
(15%) (20%) (5%)
Catastrophical 100 High 100 High 25 Very Hihg 20 Very High 5
Critical 35 Medium 15 Medium 15 High 10 High 4
Moderate 20 Low 5 Low 5 Medium 5 Medium 3
No impact 0 No impact 0 No impact 0 Low 0 Low 1

Table 2 : Assessment example

g) To calculate the frequency failure:

This is probably the easiest step of the methodology. Just one requeriment is needed: a
good failure register of the facility. With these records, the data can be obtained almost
inmediatly. If the register does not exist there are alternative processes. It can be used the
knowledge of the technicians or even public data bases as OREDA. In the example case, it
estimated that this kind of valves do not fail usually, so the frequency failure is “low”.

h) To get the criticality level:

In the end, the item must be located into the criticality matrix. It will be the graphic
representation of the criticality analysis. The mark related to the severity of the functional
loss is fixed in x-axys. In the example, the failure of the valve has the maximum puntuation
(related to industrial safety criterion). So it is fixed at the right extreme of the matrix. The
mark related with the frequency failure is fixed in y-axis. In the example the valve frequency
failure is low, so the item is fixed at the bottom extreme of the matrix.

1,6

1,2

1 a
0-9 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90-100

Table 3 : Criticality matrix example

8
 Safety and reliability enhancement throughout Europe
50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain
5. Conclusions
To have a methodology that sustains the operation and maintenance strategies is an
important requirement for the industry. Over the years, most of strategies have been
focused on availability and costs, but risk management is more important every day. Many
methodologies could be used to get a quality certification as ISO 9001, but not many
methodologies are really useful for a real company development.
Criticality analysis allows allocating with a simple methodology your items assigning them a
relative value in function of the strategy and target of the company.It is a very powerful tool
and a solid starting point for an optimising policy of OPEX in the lifecycle.

6. Acknowledgements
The members of the University of Seville would like to acknowledge the collaboration of the
MM Binladin Chair of Operations and Maintenance from the University of Taibah, in the
development of this research.

7. References
[1] J. Adams and A. K. Parlikad, “Dynamic Maintenance Based on Criticality in Electricity
Networks,” in 5th IET/IAM Asset Management Conference, 2015.

[2] A. Kelly, Maintenance strategy. Elsevier, 1997.

[3] Assetivity - Asset Management Consultants, “Equipment Criticality Analysis – is it a Waste of

Time?,” 2015. [Online]. Available: http://www.assetivity.com.au/article/reliability-
improvement/equipment-criticality-analysis-a-streamlined-approach.html. [Accessed: 15-Oct-
2015].

[4] A. Crespo Márquez, P. Moreu de León, A. Sola Rosique, and J. F. Gómez Fernández, “Criticality
Analysis for Maintenance Purposes: A Study for Complex In-service Engineering Assets,” Qual.
Reliab. Eng. Int., p. n/a–n/a, 2015.

[5] A. C. Marquez, The Maintenance Management Framework: Models and Methods for Complex
Systems Maintenance, 1st ed. Springer-verlag London, 2007.

[6] M. Braglia, M. Frosolini, and R. R. Montanari, “Fuzzy TOPSIS Approach for Failure Mode,
Effects and Criticality Analysis,” Qual. Reliab. Eng. Int., vol. 19, no. 5, pp. 425–443, Sep. 2003.

[7] W. Gilchrist, “Modelling Failure Modes and Effects Analysis,” Int. J. Qual. Reliab. Manag., vol.
10, no. 5, 1993.

[8] J. Puente, R. Pino, P. Priore, D. de la Fuente, and D. D. La Fuente, “A decision support system
for applying failure mode and effects analysis,” Int. J. Qual. Reliab. Manag., vol. 19, no. 2, pp.
137–150, Mar. 2002.

[9] S. M. Seyed-Hosseini, N. Safaei, and M. J. Asgharpour, “Reprioritization of failures in a system

failure mode and effects analysis by decision making trial and evaluation laboratory
technique,” Reliab. Eng. Syst. Saf., vol. 91, no. 8, pp. 872–881, Aug. 2006.

[10] M. Braglia, “MAFMA: multi-attribute failure mode analysis,” Int. J. Qual. Reliab. Manag., vol.
17, no. 9, pp. 1017–1033, Dec. 2000.

9
Safety and reliability enhancement throughout Europe
50th ESReDA Seminar, May 18th 2016, University of Sevilla, Spain

[11] T. R. Moss and J. Woodhouse, “Criticality analysis revisited,” Qual. Reliab. Eng. Int., vol. 15,
no. 2, pp. 117–121, Mar. 1999.

[12] M. Ben-Daya, A. Raouf, M. Ben-Daya, and A. Raouf, “A revised failure mode and effects
analysis model,” Int. J. Qual. Reliab. Manag., vol. 13, no. 1, pp. 43–47, Feb. 1996.

[13] J. B. Bowles and C. E. Peláez, “Fuzzy logic prioritization of failures in a system failure mode,
effects and criticality analysis,” Reliab. Eng. Syst. Saf., vol. 50, no. 2, pp. 203–213, Jan. 1995.

[14] C. E. Pelaez and J. B. Bowles, “Using fuzzy logic for system criticality analysis,” in Proceedings
of Annual Reliability and Maintainability Symposium (RAMS), 1994, pp. 449–455.

[15] A. C. F. Guimarães and C. M. F. Lapa, “Fuzzy inference to risk assessment on nuclear

engineering systems,” Appl. Soft Comput., vol. 7, no. 1, pp. 17–28, Jan. 2007.

[16] M. Braglia and M. Bevilacqua, “Fuzzy modelling and analytical hierarchy processing as a
means of quantifying risk levels associated with failure modes in production systems,”
Technol. Law Insur., vol. 5, no. 3–4, pp. 125–134, 2000.

[17] N. R. Sankar and B. S. Prabhu, “Modified approach for prioritization of failures in a system
failure mode and effects analysis,” Int. J. Qual. Reliab. Manag., vol. 18, no. 3, pp. 324–336,
2001.

[18] C. L. Chang, P. H. Liu, and C. C. Wei, “Failure mode and effects analysis using grey theory,”
Integr. Manuf. Syst., vol. 12, no. 3, pp. 211–216, 2001.

[19] C.-L. Chang, C.-C. Wei, Y.-H. Lee, C.-L. Chang, C.-C. Wei, and Y.-H. Lee, “Failure mode and
effects analysis using fuzzy method and grey theory,” Kybernetes, vol. 28, no. 9, pp. 1072–
1080, 1999.

[20] M. Bevilacqua, M. Braglia, and R. Gabbrielli, “Monte Carlo simulation approach for a modified
FMECA in a power plant,” Qual. Reliab. Eng. Int., vol. 16, no. 4, pp. 313–324, Jul. 2000.

[21] R. Smith, “Equipment Criticality Analysis,” Analysis, no. 6, pp. 1–20.

[22] W. J. Moore and A. G. Starr, “An intelligent maintenance system for continuous cost-based
prioritisation of maintenance activities,” Comput. Ind., vol. 57, no. 6, pp. 595–606, Aug. 2006.

[23] M. Cerrada, J. Cardillo, J. Aguilar, and R. Faneite, “Agents-based design for fault management
systems in industrial processes,” Comput. Ind., vol. 58, no. 4, pp. 313–328, May 2007.

10