(PDF) The Aspen Fidelis Reliability Difference: Minimize Risk, Maximize Profitability · 2019-01-16 · aspenONE Process Explorer™ – Using data from the aspenONE Process Explorer dashboard, - DOKUMEN.TIPS (2023)

Stacey Jones, Sr. Product Manager, Asset Performance Management,Aspen Technology, Inc.

The Aspen Fidelis Reliability™ Difference: Minimize Risk,Maximize Profitability

White Paper

Executive Summary

Making the big decisions shouldn’t be left to subjectiveperceptions or over-simplified analysis. Decision-makers needquantifiable, trustworthy answers to make the most profitabledecisions possible.

In the past, management of your capital assets was done throughgut feel from experienced operators, some rudimentary spreadsheetanalysis or simplified reliability, availability andmaintainability (RAM) tools. Although these methods provide somebenefits, they lack a holistic approach to reliability, design andoperations. They lack accuracy and consistency.

In contrast, Aspen Fidelis Reliability uses a system approach toreliability. It allows you to quantify the true value of any designor improvement project, maintenance change, operations improvementor supply chain constraint. Fidelis will give you an accurate,comprehensive bad-actor list, quantified by lost revenue andproduction — not just by maintenance cost or downtime. As a result,you can more effectively perform lifecycle analyses on assetsincluding asset utilization, maintenance effectiveness, overallequipment effectiveness and much more.

As shown in the real-world examples later in this paper, errorsin accuracy using Fidelis versus simple RAM analysis varied from0.3 percent ($4.4 million USD) to 2.8 percent ($40.9 million) inlost production and revenue annually. Simple RAM tools just do notprovide the level of accuracy needed to answer the hardquestions.

With Aspen Fidelis Reliability, decision-makers can maximize ROIby going beyond the equipment level and accurately predictingfuture asset performance of the whole system. For accuracy anddecision-making, the Fidelis difference is eye-opening!

2

The Objectives of Asset Performance Management and Aspen FidelisReliabilityIn the white paper “Seeing Into the Future withPrescriptive Analytics: A New Vision for Asset PerformanceManagement,” Robert Golightly of AspenTech describes the objectivesof asset performance management as follows:

“Investor demands to maintain or improve revenue and margins aredriving the search for new technologies and application to drivedown costs, improve reliability and increase efficiencies.Equipment Failures and process disruptions are creating unplanneddowntime that is costing the process industries billions of dollarsin lost revenue and profit each year.

This is an area where we commonly see corporate initiativecropping up around asset performance management and riskmanagement. What these companies are searching for are ways toimprove the accuracy of detection and increase the notificationperiod of these events. With more warning, more options becomeavailable – and with options comes the opportunity to mitigate thenegative impact of these events.”

Aspen Fidelis Reliability is a great example of one suchadvanced technology. Fidelis takes capital management and decisionmaking beyond emotional debates and educated guesses and intoquantitative analysis with direct feedback on ROI. With this typeof elevated insight, managers can make the hard decisions.• How doI minimize my risk?

• Should we move forward with this major capital project?

• What is the minimal capital investment required to meet mytargets?

• How can I improve my plant availability or utilization?

• What will the facility throughput be for various design oroperational scenarios?

• Which improvement strategy provides the highest ROI?

With Aspen Fidelis Reliability, these questions can be answeredby going beyond the equipment level, so that decision makers canmaximize the economics of business decisions and accurately predictfuture asset performance of the whole system.

3

Traditional Approaches

Simple stand-ins for robust RAM analysis can have huge economicimpacts. Conventional approaches lack the sophistication andflexibility to provide the same level of rigor. More often thannot, risk analysis is done via intuition from experiencedoperators.

While this level of wisdom and experience should never be takenfor granted, for robust plant-wide analysis, no one person has theforesight to understand how a failure in an upstream process unitwill impact total revenue after feeding numerous subsequent processunits and tank farms. Other simplistic RAM modeling tools areavailable, but they lack the capabilities required to includenecessary inputs for accurate results.

These tools are specifically marketed towards RAM analysis, butmany have no concept of real process flow. Because they lack theflexibility to handle real-world complexities, they require grossoversimplifications. As the examples later in this document willshow, the difference in the accuracy of the answers isstaggering.

4

Fidelis takes capital management and decision-making fromemotional debates and educated guesses to quantitative analysiswith direct feedback on ROI.

New Technology and Approaches

Aspen Fidelis Reliability, a Monte Carlo-based, discrete-eventsimulation program, utilizes statistical sampling techniques topredict the future performance of a system.

The behavior of events in the system (pumps, motors, weatherimpacts, operational upsets, etc.) are represented by aprobabilistic distribution function. Fidelis samples from thiscurve and derives a time to failure or repair for each event. Aseach event fails, it influences the associated parent unit which inturn influences the flow sent through the pipes attached to thatunit both upstream and downstream. Fidelis tracks the flow throughthe pipes, actual tank levels, as well as the utilized andavailable capacities of all units.

In addition, any required custom logic (handle dynamic batching,seasonal changes, alternate flow paths, non-time-based failures,conditional logic or equipment aging) can be incorporated into themodel. If you can describe it in words, Fidelis can incorporateit.

Beyond providing significantly more accurate answers, AspenFidelis Reliability also excels in terms of ease of use andtransparency. The following features help to make the userexperience quick and easy to analyze:

• Validation: A static animation of the effects of eventfailures, high or low tank levels, etc. This allows the user toquickly and easily assess the system configuration without runningthe simulation.

• Review: For each simulation run, Fidelis records a liveanimation of all events, tank events, logic and performance. Thisallows the user an unparalleled experience in transparency. If forexample, you want to see why Unit Z showed a performance dip attime 500 days in lifecycle 30, you can select these parameters inthe review dialog and visually inspect what led up to the dip andhow the system responded. You can play this review forward orbackwards automatically or watch an event by event progression.

• The Fidelis Iterative Process: When Fidelis is utilized, theresults are never just handed to the final project team. A robustprocess of validation, iteration and transparency is used. RAMmodeling should never be “black magic” or a ”black box.” Along withresult details, validation and review features, and the Fidelismodeling process, the project team has buy-in and clearunderstanding from Day 1. Once the base case model is validated byall key stakeholders, only then is the model used to quantifyalternate cases.

• Flow Diagram Builder: Fidelis uses a toolbar drawing controlto quickly drop hundreds of objects very easily. It typicallyrequires less than 30 minutes to build the main flow diagram andpipe the feed units together. Inside each unit, you can completedetailed unit level diagrams as well with unlimited nesting.

5

Further rounding out this tool is its parent company: AspenTechis a proven market leader and standard-bearer in processoptimization. The synergies that exist between AspenTech’sEngineering and Manufacturing products and the Asset PerformanceManagement products (such as Fidelis) makes for an optimizationpackage that is impossible to match. Some of the potentialintegration synergies with Fidelis include:

Aspen HYSYS® or Aspen Plus® – Leverage existing flow diagramsand automatically populate starting RAM data from Fidelis.

Aspen Capital Cost Estimator™ – Improve cost estimates byincorporating statistical variance with uncertainty. In addition,calculate the high-level cost of any initial design.

Aspen Mtell® – Predict when a real-time failure will occur.Integrate this information into Fidelis to determine the responsethat will result in the least amount of downtime and revenueimpacts.

aspenONE Process Explorer™ – Using data from the aspenONEProcess Explorer dashboard, Fidelis can be integrated to recommendthe best course of action given current plant states (tanks,units), update Fidelis data with live reliability data from anyhistorian and update the dashboard with current threats and theFidelis criticality listing.

Aspen PIMS-AO™ – Fidelis can assist planners in understandingthe likelihood of meeting plans by adding statistical variabilityand rankings to scenarios.

Aspen Petroleum Scheduler™ and/or Aspen Plant Scheduler™ – Byaccessing the generated schedules, Fidelis can integrate dynamicfailures to optimize the schedule and tankage. Fidelis can also beused to quantify the likelihood of success for a given schedule andcause of misses to increase equipment utilization (includinglogistics).

6

Analyzing the Impact of Real-World ChallengesAspen FidelisReliability was used to analyze several real-world challenges. Ineach case, simplistic RAM tools and Microsoft Excel required theuser to ignore or oversimplify these challenges. The difference inthe level of accuracy between the two methods was calculated.

The results are represented in terms of percentage of accuracyso as to extrapolate the real impact on revenue across multipleindustrial verticals. Applying this to real dollars, 1 percent inaccuracy would represent $14.6 million annually at a typical 200kbbl refinery (at $20/barrel).

Challenges that were examined include:• Flowoptimization/complex routing of flows

• Tank/buffer optimization and tank leveling

• Conditional or variable impact of failures

• Impacts from logistics and supply chain

• Equipment aging or incomplete renewal after maintenance

• Non-time-based failures

For the comprehensive model, all six of these real-worldchallenges were integrated. For each subsequent case, one challengewas systematically removed, and the simulation rerun. In this way,Aspen Fidelis Reliability was able to quantify the delta betweenrobust stochastic modeling and the simplified alternative.

In addition, a case was run with all six real-world challengesremoved. This case represents the full impact of ignoring thesecomplexities found in every typical processing and manufacturingfacility around the world.

7

Figure 1: Delta Between Fidelis Modeling and the SimplifiedAlternative

Fidelis will give you an accurate, comprehensive bad-actor list,quantified by lost revenue and production — not just bymaintenance.

* Reference refinery assumed 200 kbbls per day ($20/barrel)

E�ec

t on

Site

Ava

ilabi

lity

by R

emov

ing

Chal

leng

es

Delta in Annually

Revenue ($)*

Delta from Fidelis Comprehensive Site

Availability

Optimization / Complex Flow

Routing Removed88.9% -$23,360,000-1.6%

Tank / Buer

Optimization Removed88.4% -$30,660,000-2.1%

Conditional or Variable Impact

of Failures Removed92.4% $27,740,0001.9%

Logistics and Supply Chain

Impact Removed93.3% $40,880,0002.8%

Equipment Aging / Incomplete

Renewal from Maintenance Removed91.1% $8,760,0000.6%

Total of All Removed 88.2% -$32,120,000-2.2%

Fidelis Comprehensive Site Availability90.5%

Time Based Failures Removed 90.2% $-4,380,0000.3%Non -

8

ConclusionYou can only make the best decisions and maximizeprofitability if you have quantifiable answers you can trust. Toachieve this level of accuracy, managers need to invest in a RAMtool that can handle the real-world challenges of today’s processindustries. Aspen Fidelis Reliability is a tool designed to meetthese challenges.

On average, RAM simulations using Fidelis on capital projectsover the past 18 years have resulted in 5 percent savings incapital costs and production increases of approximately 3 percent —with total monetary benefits ranging from $1 million USD to morethan $300 million USD.

As the previous analysis shows, small over-simplifications in asimulation can add up to big margins of error. Doing a model thecorrect way the first time can save millions in design corrections,debottlenecking or overspending on capital.

9

AspenTech is a leading software supplier for optimizing assetperformance. Our products thrive in complex, industrialenvironments where it is critical to optimize the asset design,operation and maintenance lifecycle. AspenTech uniquely combinesdecades of process modeling expertise with machine learning. Ourpurpose-built software platform automates knowledge work and buildssustainable competitive advantage by delivering high returns overthe entire asset lifecycle. As a result, companies incapital-intensive industries can maximize uptime and push thelimits of performance, running their assets faster, safer, longerand greener.

www.aspentech.com

© 2019 Aspen Technology, Inc. AspenTech®, aspenONE®, the Aspenleaf logo, the aspenONE logo

and OPTIMIZE are trademarks of Aspen Technology, Inc. All rightsreserved. AT-04560