www.advisian.com

April 2017

Data-centric Safety Critical Element ManagementKevin GanPrincipal, Asset Integrity & Digital Advisory

A Safety Critical Element (SCE) is classified as an equipment, structure or system whose failure could cause or contribute to a major accident, or whose purpose is to prevent or mitigate the effect of a major accident.

Effective management of Safety Critical Elements must be a primary focus

1. Bowtie / MAH Workshops 2. Barrier Assessments 3. Performance Standards

To identify events, threats and consequences

Typical steps to identifying SCEs include:

To categorise SCEs into groups.

An example of a barrier would be Shutdown Systems and one

of the SCE groups would be ESDVs.

Developed to maintain the integrity and ensure the functions of the SCEs are

intact.

Today, there are software packages that will visualise and integrate bowtie models and barriers to make life easier during design and operations.

Though we still face issues managing Safety Critical Elements

SCEs are missing from

the asset register or so it appears because

they’re not identified clearly

in the system

There is no methodology in place when

classifying SCEs, leading to wrong data

There are too many SCEs,

leading to high, unnecessary maintenance

costs and man-hours

There is a lack of accurate

reporting on the status of

SCE integrity - correct decisions cannot be easily

made when it comes to the

integrity of assets

The answer? Good data management.

Good data management can help answer:

What is the probability that a SCE

will fail soon?

What is its remaining useful life?

What are the causes of failure?

What maintenance

should be performed to

fix them?

• Reducing operational risk• Identifying failures before they occur• Reducing unnecessary maintenance operations• Controlling cost of maintenance• Lowering inventory costs • Discovering trends of various maintenance problems

It offers predictive maintenance through the Internet of Things

But the key questions remain

Are we considering predictive

maintenance as

part of our maintenance

strategy?

Are we implementing

it correctly, if at all?

Are we making full use of the

data available?

Validate model by pulling real-time data.Integrate model within

the organisation to start the optimisation

process.

Steps to implementing Predictive Maintenance

IdentifyOutcome

DetermineData Sources

DataIntegration

Model Development

Validate and Integrate

1 2 3 4 5

What do we want to predict?

For example, performance

data, maintenance logs, sensors, the weather

etc.

Data needs to be normalised into a single,

consistent system.

To identify meaningful

patterns using machine learning

techniques.

Once set up, maintenance needs can be anticipated and unscheduled downtime can

be avoided.

Visual Integrity ManagementVisual Integrity Management is the next level of SCE maintenance and operations management - we now have the capability to connect multiple systems and data sources into a virtual asset

Engineers can now visit remote platforms, take readings, measurements and assess maintenance requirements without having to leave the office.

We need to start exploring solutions with the tools and technologies (e.g. SCE management processes, predictive analytics, visual integrity management etc.) that are available to us today.

DISCLAIMER

This presentation has been prepared by a representative of Advisian.The presentation contains the professional and personal opinions of the presenter, which are given in good faith. As such, opinions presented herein may not always necessarily reflect the position of Advisian as a whole, its officers or executive.Any forward-looking statements included in this presentation will involve subjective judgment and analysis and are subject to uncertainties, risks and contingencies—many of which are outside the control of, and may be unknown to, Advisian. Advisian and all associated entities and representatives make no representation or warranty as to the accuracy, reliability or completeness of information in this document and do not take responsibility for updating any information or correcting any error or omission that may become apparent after this document has been issued.To the extent permitted by law, Advisian and its officers, employees, related bodies and agents disclaim all liability—direct, indirect or consequential (and whether or not arising out of the negligence, default or lack of care of Advisian and/or any of its agents)—for any loss or damage suffered by a recipient or other persons arising out of, or in connection with, any use or reliance on this presentation or information.

FOR MORE INFORMATION CONTACT

Kevin Gan | Asset Integrity & Digital Enterprise Consultant

E kevin.gan@advisian.com