Predictive Analytics - Tools to Predict Risk | AIChE

Predictive Analytics - Tools to Predict Risk

Type

Conference Presentation

Conference Type

AIChE Spring Meeting and Global Congress on Process Safety

Presentation Date

April 21, 2021

Duration

20 minutes

Skill Level

Intermediate

PDHs

0.50

Predictive
Analytics – Tools to Predict Risk

When you conclude
a Process Hazards Analysis (PHA) session you feel good. You have put your best
effort into it, worked really hard and contributed to
identify hazards, safeguards and recommendations to make your facility safer.
But did you achieve your goal?

Analyzing PHA data
is a relatively new field with only a few pioneers actively working on it but
the initial results have been impressive. Findings in the area of Risk
Discovery, Safeguard Profiles and Recommendation Value have driven insights
into process safety that had only been guessed at previously. The finding that
the average PHA team only discovers approximately 33% of the risk scenarios in
their area of study is surprising, and the inconsistencies in risk ranking
similar scenarios equally so.

Through the
analysis of several thousands, (4,000+) of data sets Risk Alive’s data science
team has been able to uncover patterns found in PHA data. Extrapolating from
those patterns predictive models have been built that investigate different
criteria such as Risk Discovery and can predict with a +/- 15% degree of
accuracy.

Model 1 – Risk
Discovery

The Risk Discovery
model can identify the PHA files most likely to have the largest gaps and
greatest numbers of missed scenarios. The magnitude of potentially missed risks
is derived from a mass comparison of similar units in our database. This
information can help an organization know which of their units have the most
uncovered risks and require additional attention.

Model 2 –
Safeguard Profile

The Safeguard
Profile model reviews the applications of different safeguard types including,
mechanical devices (PSV, PVRV, etc.), BPCS alarms, SIS, etc. and evaluates how
different strategies can be used across facilities. It highlights the potential
over protection of a facility, along with the investments associated with
installing and maintaining those safeguards. This model can also help
organization understand how their safeguard strategy compares to other similar
processing units in the database.

Model 3 -
Recommendation Return On Investment (ROI)

The Recommendation
Return on Investment model can uncover and visualize how much investment is
required for a set of recommendations and where money is being misspent on actions
that are not adding reasonable value. There can be significant savings in
understanding what recommendations are providing the most risk reduction and emphasizing
the importance on those.

Using advanced
data analytics and data science, predictive modelling can help you see where
you are most likely to find problem areas with uncovered risks and allow you to
focus your efforts on where to investigate with more detailed analytics to make
for a safer tomorrow.

Presenter(s) 

Once the content has been viewed and you have attested to it, you will be able to download and print a certificate for PDH credits. If you have already viewed this content, please click here to login.

Language 

Checkout

Checkout

Do you already own this?

Pricing

Individuals

AIChE Member Credits 0.5
AIChE Pro Members $19.00
AIChE Graduate Student Members Free
Safety and Health Division Members Free
AIChE Undergraduate Student Members Free
AIChE Explorer Members $29.00
Non-Members $29.00