Knar Global LLC - Knowledge and Integration Architects
ENES
Home
Business Value
About
Our PhilosophyJorge Granada, FounderRegulatory Compliance
Architect Systems
Engineering Design & Concept OptimizationModel-Based Systems Engineering (MBSE)KIAME Framework
Optimization
Challenges
Digital Asset Value RealizationStop Unplanned DowntimeOptimize Maintenance SchedulingAssure Lifecycle Investment ValueUnlock Hidden Business ValueMake Better Capital DecisionsTransition to Cleaner Energy
Methods
Integrated Decision Support Systems (iDSS)Industrial Digital AssetsCausal Analysis & Non-RecurrenceOrganizational Competence BuildingProprietary MethodologiesPrysma CAUSALITYProbabilistic MethodsLife Cycle Cost OptimizationReliability-Profile Sparing
Services
Packetized Management CortexIntegrity Modeling ServicesRAM AnalysisIndustrial Digital AssetsIntegrated Decision Support SystemsNational & Regional Resource and Energy Decision ModelsLife Cycle Value AssuranceCausal Analysis & Non-RecurrenceOrganizational Competence BuildingStrategic Energy Transition
Case Studies
KnarLAB
Books
Blog
Contact
ENES
Skip to RAM Analysis content
  • Questions
  • Results
  • Cost scale
  • Capabilities
  • Pseudo-RAM
  • Deliverables
  • When required
  • Delivery
  • Limits
  • Your decision
  • Related
On this page
  • Questions
  • Results
  • Cost scale
  • Capabilities
  • Pseudo-RAM
  • Deliverables
  • When required
  • Delivery
  • Limits
  • Your decision
  • Related
Reliability, availability and maintainability model feeding an asset decision

Knar RAM Analysis

Know What Your Plant Will Deliver—Before You Commit the Capital

Decision-grade Reliability, Availability and Maintainability analysis for designs, investments, contracts and maintenance strategies that cannot be cheaply reversed.

A Knar RAM Analysis gives management a defensible forecast of production, downside exposure and lifecycle value for every option under consideration. It shows what to approve, what to change and what does not earn its cost—before the decision becomes steel, contract or lost production.

Book a RAM Decision ReviewSee results from real RAM Analysis

The central concept

A real RAM Analysis is a working representation of how the asset produces value. It connects equipment failures, repair times, planned maintenance, redundancy, process capacity, storage, utilities, operating rules and demand. It then tests each decision across many credible operating histories. The output is not an isolated availability percentage. It is expected production, downside exposure, lost margin, lifecycle cost and a management recommendation.

Management answers—not more data

  • How much saleable production will each option realistically deliver?
  • What is the financial downside when failures, maintenance and operating restrictions occur together?
  • Which equipment, operating rule or maintenance activity destroys the most value?
  • Does redundancy, storage, a spare or added maintenance pay for itself?
  • Which assumption could reverse the recommendation?
  • What should management approve, reject, modify or negotiate?

Knar begins with the decision. The model boundary, data effort and technical detail are selected only to answer that decision with the required confidence.

Results from REAL RAM Analysis

These anonymized results come from completed industrial RAM work. They are presented as modeled decision exposure or value protected—not as invented savings or guaranteed returns.

  • USD 3.25M

    Hidden contract exposure made visible

    A supplier reported approximately USD 0.8 million of eight-year exposure. Independent RAM verification showed USD 3.25 million of additional exposure linked mainly to simultaneous generation failures omitted from the commercial calculation. Management gained a quantified risk, the assumption creating it and a concrete basis for negotiation.

  • USD 10.15M

    Standby capacity economically defended

    Removing 4.3 MW of standby generation affected operation for only about 25 hours per year. Across eight years, those few hours represented USD 10.15 million of modeled lost-profit exposure. A utilization report could make the standby unit look unnecessary; RAM showed the value it protected.

  • USD 11.38M

    A production restriction converted into business value

    At an upstream asset, one sensitivity identified 169,912 barrels of modeled loss over eleven years—USD 11.38 million—associated with an air-cooling capacity restriction. The decision was no longer about a technical percentage; it was about where an operating constraint placed millions of dollars at risk.

  • USD 9.29M

    Reliability losses ranked in money

    The same integrated model identified 138,665 barrels—USD 9.29 million—of modeled eleven-year loss associated with producer-well reliability in another scenario. These sensitivities are alternative views and are not additive. Their value is prioritization: management can direct effort toward the constraint with the greatest business consequence.

  • USD 2.46M

    Unnecessary capital avoided

    In another eight-year case, independent verification increased the modeled loss estimate to USD 2.46 million. Even then, an additional turbine was not economically justified. A good RAM Analysis does not sell more equipment; it determines whether the equipment earns its place.

Read the numbers correctly

The figures above are modeled exposures or protected value under stated scenarios. They are not all realized savings, and results from separate projects must not be added. RAM creates value by making the consequence visible before management commits—or by proving that additional capital is unnecessary.

The cost-to-decision scale

A detailed base RAM study for a complex mining design was scoped at approximately USD 78,000 and ten weeks. That is a reference point, not a universal fee. Compared with that study-cost scale, USD 3.25 million is about 42 times larger and USD 10.15 million is about 130 times larger.

Study cost compared with decision exposure, drawn to scale
  • Detailed base RAM study, complex mining design

    USD 78,000

    Scoped at about ten weeks

  • Hidden contract exposure made visible

    USD 3.25M

    About 42 times larger

  • Standby capacity economically defended

    USD 10.15M

    About 130 times larger

The commercial rule is simple: when one credible assumption can move millions of dollars, spending tens of thousands to test the decision is proportionate. The relevant comparison is not the study fee versus zero; it is the study fee versus the financial consequence of approving the wrong option.

What makes a RAM Analysis decision-grade

A RAM Analysis combines several techniques and several ways of representing the plant, and a decision-grade RAM Analysis needs all of those listed below. Each one closes a specific decision gap. If one is missing, the RAM Analysis cannot connect equipment failures to production, cost and risk, so management cannot rely on it to commit capital. These techniques and representations are required to link the plant to its business result, not to make the RAM Analysis look sophisticated.

  • Physical production path

    Represents how feed, products, utilities and energy move through the asset. This converts equipment availability into achievable production.

  • Operating logic

    Represents duty/standby changes, load sharing, minimum-running requirements, batching, shutdown and restart rules. This determines whether an event stops production, reduces it or has no immediate effect.

  • Storage and buffers

    Represents tanks, stockpiles and intermediate storage. These can absorb a failure, delay a loss or move the bottleneck elsewhere.

  • Failure and repair behavior

    Represents relevant failure modes, time to failure, repair duration, restoration and restart—not one generic average for every machine.

  • Planned maintenance and resources

    Represents overhauls, inspections, routine work, crew or access constraints, and their interaction with equipment redundancy.

  • Simultaneous events

    Allows multiple units to be unavailable together. In reviewed commercial models, excluding joint failures produced lost-profit estimates up to ten times lower at two facilities.

  • Repeated simulation

    Tests many credible operating histories rather than one deterministic future. This produces expected and conservative outcomes such as P50, P80 or P90.

  • Business conversion

    Translates production loss into margin, unit cost and lifecycle value using owner-approved assumptions. An outage at high load is not economically equal to one at low load.

  • Verification and validation

    Checks the model against drawings, operating logic, source data, historical behavior and independent calculations. Differences and limitations are disclosed before the model supports approval or negotiation.

Why integration mattersNo technique answers the management question alone. Decision-grade RAM follows the complete chain:
  1. Event

    ↓→
  2. Operating response

    ↓→
  3. Capacity effect

    ↓→
  4. Storage response

    ↓→
  5. Production loss

    ↓→
  6. Financial consequence

Break that chain and the answer may look precise while being commercially wrong.

Do not accept pseudo-RAM

Excel is useful for inputs, checks and reporting. The problem begins when a static spreadsheet is presented as a model of a dynamic industrial system.

Do not accept a RAM study that:

  • Reject: Multiplies nameplate capacity by one average availability percentage.
  • Reject: Adds component availabilities without recreating system configuration and production flow.
  • Reject: Assumes failures occur one at a time.
  • Reject: Ignores partial production, degraded operation, restart time or minimum-capacity rules.
  • Reject: Ignores tanks, stockpiles, buffers, batching or downstream restrictions.
  • Reject: Uses one generic failure and repair average for materially different equipment and failure modes.
  • Reject: Excludes planned maintenance or assumes every activity has unlimited access and resources.
  • Reject: Reports one answer without a credible range.
  • Reject: Assigns the same financial loss to every outage hour.
  • Reject: Accepts vendor assumptions without testing their production and commercial effects.
  • Reject: Cannot reproduce the base case or explain differences from plant history.
  • Reject: Delivers an availability percentage without production, financial consequence or a decision recommendation.

The acceptance test

Ask the provider to trace one credible event through the model: “If this unit fails while another unit is under maintenance, at this demand level, what happens to capacity, flow, storage, production and margin?” If the model cannot complete that chain, it is not decision-grade RAM.

What management receives

Expected production
What the complete system can realistically deliver.
Downside range
The likely outcome and a conservative planning outcome—not one optimistic number.
Loss ranking
The equipment, failure modes, maintenance work and operating constraints causing the greatest value loss.
Option comparison
Production, risk and financial differences between alternatives evaluated on the same basis.
Investment test
Whether redundancy, storage, spares or maintenance changes are economically justified.
Reversal points
The demand, repair time, price or operating assumption that would change the recommendation.
Data-quality statement
What is known, what is assumed and how each limitation affects confidence.
Executive recommendation
What to approve, reject, modify, negotiate or investigate.
Reusable model
A decision asset that can answer new questions as designs, prices and operating conditions change.

When RAM should be required

Require RAM when all three conditions are present:

  1. Condition 1

    The decision affects production, capital, lifecycle cost or contractual performance.

  2. Condition 2

    Failures, repairs, maintenance, storage, utilities or operating constraints interact.

  3. Condition 3

    Reversing the decision later would be expensive, slow or impossible.

Typical decisions include plant design and expansion, redundancy, major equipment selection, supplier guarantees, debottlenecking, storage, spare-parts strategy, turnaround timing and scope, maintenance-policy changes, and recurring production loss without an agreed cause.

Do not commission RAM by default. A direct calculation may be sufficient when the system is simple, the consequence is minor or the organization will not act on the result. Knar defines the minimum credible analysis needed for the decision—not the largest possible model.

How Knar delivers it

  1. 1

    Step 1: Start with the decision

    Define the approval required, the alternatives and the consequence of being wrong.

  2. 2

    Step 2: Set the minimum credible boundary

    Include only the systems, utilities, storage, operating rules and external restrictions capable of changing the decision.

  3. 3

    Step 3: Build and challenge the base case

    Recreate current or proposed operation; record source data, owner-approved assumptions and limitations.

  4. 4

    Step 4: Test alternatives on equal terms

    Use the same operating, reliability and economic basis for every option.

  5. 5

    Step 5: Issue an auditable recommendation

    State the preferred option, business consequence, confidence range, assumptions and conditions that could reverse it.

  6. 6

    Step 6: Keep the model useful

    Update it as design, prices and operating experience change, so the work becomes decision infrastructure rather than a report stored away.

What RAM does—and does not do

RAM does not replace the plant manager, engineering authority, safety review or investment committee. It does not make weak data disappear, promise certainty or make the decision. It makes uncertainty, assumptions and consequences visible enough for accountable management judgment.

A Knar RAM Analysis combines plant knowledge, reliability evidence, operating logic and financial assumptions in one traceable decision model. That integration is what allows technical findings to arrive at the approval table in the language of production, risk and value.

Bring one important decision

Do not begin with “we need a RAM study”. Begin with the decision:

  • Should the standby unit remain?
  • Which design should be approved?
  • Does the supplier guarantee protect production?
  • Should the turnaround date, duration or scope change?
  • Which bottleneck deserves capital first?
  • Will the expansion deliver the promised output?

Book a RAM Decision Review

Bring one pending design, investment, procurement, contract or maintenance decision. Knar will define the value at risk, the minimum credible model boundary, the data required and the management outputs. If a direct calculation is enough, that will be said. If RAM is justified, the path to a decision-grade answer will be clear.

Bring your decision to Knar

Put production, risk and value in full view.

Then approve the decision with confidence.

Keep exploring

Related methods and reading

RAM Analysis sits inside a wider body of Knar methods, topic guides, articles and applied cases. Start with the one closest to the decision you are weighing.

Methods

  • RAMgenThe standardized framework Knar uses to build RAM models consistently across assets and options.
  • Probabilistic methodsHow uncertainty is carried through reliability, operations and economics instead of being averaged away.
  • Life cycle cost measurement and optimizationWhere RAM results become lifecycle cost and value, so options compare on more than capital cost.

Topic guides

  • RAM analysis and RAM modelingA plain-language guide to reliability, availability and maintainability modeling.
  • Availability analysisWhy an availability percentage alone is not an answer, and what to calculate instead.
  • Monte Carlo simulationHow repeated simulation produces the P50, P80 and P90 outcomes behind a RAM recommendation.
  • Maintainability engineeringDesigning and analyzing systems so they can be restored to service quickly, safely, and predictably.

Articles and applications

  • RAM models: the foundation of value resilienceWhy a RAM model is a decision asset and not a one-time report.
  • Expected value: beyond availability metricsMoving from availability percentages to expected production and financial consequence.
  • Case studiesDocumented Knar work across oil and gas, power, mining and complex capital systems.
Knar Global LLC - Knowledge and Integration Architects

Knowledge and Integration Architects for Mission-Critical Industrial Systems

Houston, TX

info@knarglobal.com
+1 (469) 473-1708

About

  • Our Philosophy
  • KIAME Framework
  • Jorge Granada, Founder

What We Do

  • Architect Optimal Systems
  • Solve Your Challenges
  • How We Work
  • Business Value We Deliver

Solutions

  • Digital Asset Value Realization
  • Stop Unplanned Downtime
  • Optimize Maintenance
  • Assure Lifecycle Investment Value
  • Unlock Hidden Business Value
  • Capital Decisions
  • Energy Transition

Resources

  • Case Studies
  • Blog
  • Engineering Topics
  • KnarLAB
  • Contact Us

Newsletter

Subscribe to receive insights on industrial optimization, reliability engineering, and decision support systems.

We respect your privacy. Unsubscribe at any time. No spam, only valuable insights.

© 2026 Knar Global LLC. All rights reserved.

Installing cognitive infrastructure, not delivering reports.

Contact Us