Root Cause Analysis Guide & Checklist

🧭

Method Selection Guide

Choose the right RCA method for your situation

IEC / ISO / CCPS Process Safety

When to use which method

📊

Pareto Chart

Visualise which few causes account for most failures. Quantitative, data-driven. Use when you have frequency or cost data across categories.

Best for: prioritising where to focus resources

🔎

5 Whys

Drill through successive layers of cause by repeatedly asking "why". Simple, fast, no statistics required. Suited to single-path, straightforward problems.

Best for: incidents, deviations, equipment failures

🐟

Fishbone (Ishikawa)

Brainstorm and categorise potential causes across People, Machine, Method, Measurement, Material and Environment. Good for complex problems with multiple possible contributors.

Best for: brainstorming sessions, quality investigations

📈

Scatter Plot

Quantitatively test whether two variables correlate. Use to validate or disprove suspected causes identified in a fishbone or 5 Whys. Requires paired measurement data.

Best for: confirming or rejecting a specific hypothesis

⚙️

FMEA / PFMEA

Proactively identify failure modes, assess severity / occurrence / detection, and calculate Risk Priority Numbers. Used in design or process reviews before failures occur.

Best for: new designs, MOC reviews, HAZOP follow-up

💡

Combine methods

For complex problems: start with Pareto to prioritise, use Fishbone to brainstorm, validate with Scatter, and document actions in FMEA.

Methods work together, not in isolation

Complexity legend

Low effort / quick Medium effort / data needed High effort / team required

📊

Pareto Chart

Identify the 20% of causes driving 80% of problems — Pareto's law

Quantitative Seven Basic Tools of Quality

Method type

Descriptive statistics — frequency / cost analysis

Input required

Tally data — occurrence count per failure category

Reference

Joseph Juran — Seven Basic Tools of Quality

Effort

Low to medium — depends on data collection period

Procedure checklist

1. Preparation

Define the problem or process to be analysed

Identify the categories of failure / defect / cause to be counted

Include an "Other" category to capture unanticipated causes

Define the observation period (duration and frequency)

2. Data collection

Collect occurrence data systematically over the defined period

Tally occurrences per category

Sort categories in descending order of frequency

3. Chart construction

Create bar chart with categories on x-axis, frequency on left y-axis

Calculate cumulative percentage column

Add cumulative % line on secondary y-axis (0–100%)

4. Analysis & follow-up

Identify which categories cumulatively account for ~80% of occurrences

Focus RCA effort on top contributing categories

Apply sub-Pareto to top category if needed to drill further

Document findings and trigger corrective action for top causes

Live Pareto visualiser

🔎

5 Whys Method

Iteratively drill down through layers of cause to reach the true root cause

Qualitative Toyota Production System / Lean

Method type

Qualitative — question-based drill-down

Best suited for

Straightforward single-cause problems; incidents; deviations

Reference

Taiichi Ohno — Toyota Production System; CCPS Guidelines

Limitation

Not suitable for complex multi-root-cause problems — combine with fishbone

Procedure checklist

1. Problem definition

Write a clear, specific problem statement (what, where, when)

Ensure the problem statement describes the symptom, not an assumed cause

Assemble the right people (operators, engineers, maintenance) for the session

2. Asking the whys

Ask "Why did this happen?" and document the first answer factually

Use the answer as input to the next "Why?" — do not assume

Continue until the root cause (systemic, not symptomatic) is reached

Verify each answer with physical evidence or data where possible

3. Validation

Check: would fixing the root cause prevent recurrence? (not just this incident)

Confirm root cause is within the team's control to address

4. Corrective action

Update standard work or procedure that caused the deviation

Add finding to FMEA / control plan if applicable

Create audit question or verification step to confirm fix is sustained

Record the complete 5 Whys chain in the incident / NCR reporting system

5-Why Analysis Form

Section / Unit / Area

Date

Reference / Report No.

Prepared by

Problem Statement

Describe clearly what happened, where, when and what the impact is. Do not state a cause — describe the symptom only.

Short-Term Action Taken / Correction

Immediate containment action taken to stabilise the situation. This is NOT the root cause solution — it only addresses the symptom.

Root Cause Analysis — Why Chain

Each answer becomes the subject of the next "Why?". Support each answer with evidence. You may need up to 7 levels to reach a systemic root cause.

WHY 1

State the direct reason the problem occurred.

WHY 2

More concise — may branch to multiple causes.

WHY 3

Critical transition — obvious → non-obvious.

WHY 4

Clear your mind — explore all avenues.

WHY 5

Systemic cause likely reached here.

WHY 6

Only if still process-related after Why 5.

WHY 7

Only if systemic cause not yet reached.

Root Cause Identified

Summarise the confirmed systemic root cause.

Solution (Permanent Fix)

Describe the structural change that prevents recurrence.

Conclusion — Validation Sentence

Complete this sentence to verify the 5-Why chain is coherent: "[Problem] occurred due to [Why 5]. This was caused by [Why 4], mainly because [Why 3] was allowed by [Why 2], which led to [Why 1]."

Action Points

List all actions required to implement the solution and prevent recurrence. Add rows as needed.

#	Action Point	Responsibility	Target Date	Actual Completion
1
2
3

Improvements Achieved

Describe the measurable or observable improvements after the solution was implemented.

Sign-off

To be completed by the responsible supervisor or area manager after confirming the solution has been implemented and verified.

Supervisor / Manager name

Function / Title

Date of sign-off

Signature

Click to sign

🐟

Fishbone (Ishikawa) Diagram

Organise potential causes into categories — the 6M framework

Qualitative Seven Basic Tools of Quality

Also known as

Ishikawa diagram / cause-and-effect diagram

Framework

6M — Man, Machine, Method, Measurement, Material, Milieu (Environment)

Reference

Kaoru Ishikawa — University of Tokyo; Seven Basic Tools of Quality

Best used

Brainstorming sessions; feed results into scatter plot or FMEA for validation

Procedure checklist

1. Problem statement

Define and agree on the problem (effect) — write it at the "head" of the fish

Confirm the problem statement is specific and observable, not an assumption

2. Brainstorming — People

Were operators properly trained and certified for the task?

Could human factors (fatigue, ergonomics, visibility) contribute?

Were maintenance / inspection personnel using correct tools and technique?

3. Brainstorming — Machine / Equipment

Was the preventive maintenance schedule followed?

Are there broken, worn, or missing parts on the equipment?

Is equipment calibrated and fit for purpose?

4. Brainstorming — Method / Procedure

Is the procedure / work instruction present at the point of use?

Does the procedure accurately describe the correct method?

Does the procedure account for this specific failure scenario?

5. Brainstorming — Measurement

Was the measurement system (gauge, sensor, analyser) functioning correctly?

Has gauge R&R been performed and accepted for this measurement?

Were measurement results interpreted correctly and within specification?

6. Brainstorming — Material

Was the correct material / feedstock used as specified?

Was material quality (composition, grade, cleanliness) within specification?

Was material handled, stored, and transported correctly?

7. Brainstorming — Environment / Milieu

Could temperature, humidity, or pressure conditions have contributed?

Was the work environment (ventilation, lighting, housekeeping) adequate?

Could external factors (utility fluctuation, weather) have played a role?

8. Finalisation

Prioritise most likely causes from all categories for further investigation

Validate top candidates with data (scatter plot) or testing

Interactive Fishbone worksheet

👤 People

⚙️ Machine / Equipment

📋 Method / Procedure

📏 Measurement

🧪 Material

🌡️ Environment / Milieu

▶ Problem / Effect

📈

Scatter Plot

Quantitatively test whether a suspected cause correlates with the observed effect

Quantitative Seven Basic Tools of Quality

Purpose

Test correlation between independent variable (suspected cause) and dependent variable (effect)

Limitation

Correlation ≠ causation. A strong scatter does not prove the root cause; it supports a hypothesis

Next step

Use R² regression to quantify strength of correlation; combine with 5 Whys or FMEA

Reference

Seven Basic Tools of Quality; ISO 9001 quality analysis

Procedure checklist

1. Hypothesis formulation

State the specific hypothesis: "X causes / correlates with Y"

Identify independent variable (X) — the suspected cause

Identify dependent variable (Y) — the measured effect or outcome

2. Data collection

Choose variables that are measurable, objective, and simply defined

Collect a representative data set of paired (X, Y) measurements

Ensure data covers sufficient range of X to reveal a trend if one exists

3. Plot and evaluation

Plot X on horizontal axis, Y on vertical axis

Visually inspect for pattern: upward slope = positive correlation; downward = negative; random = no correlation

Calculate Pearson correlation coefficient (r) or R² if strong trend is visible

4. Interpretation

Do not conclude causation from correlation alone — validate with other RCA methods

If correlation is confirmed, proceed to corrective action and FMEA update

Test remaining suspected causes from the fishbone / 5 Whys in separate scatter plots

Data entry & correlation indicator

X-axis label (independent / cause)

Y-axis label (dependent / effect)

#	X value	Y value
1
2
3
4
5

⚙️

Failure Mode and Effects Analysis (FMEA)

Proactively assess failure modes, risk priority, and corrective controls

Quantitative / Semi-quantitative IEC 60812 / AIAG Process Safety

RPN formula

RPN = Severity (SEV) × Exposure (EXP) × Barrier (BARRIER) — each scored 1–5

Score guide

1 = best / lowest risk | 5 = worst / highest risk. Max RPN = 125. See Risk Matrix tab.

When to use

New process / equipment design; MOC reviews; HAZOP action items; post-incident investigations

Reference

IEC 60812; AIAG FMEA manual; CCPS Guidelines; RA-FMEA Risk Assessment framework

1. Scope definition

Define the process, system, or product boundary for the FMEA

Break the process down into sub-processes / steps / tasks

Assemble cross-functional team (process, maintenance, QA, HSE)

2. Hazard / failure mode identification

Identify all potential hazards or failure modes for each process step

Define the potential effects / consequences of each failure mode

Identify the potential cause(s) / mechanism(s) of each failure mode

3. Current state scoring (use Rating Matrix tab)

Document existing barrier analysis / current risk controls for each failure mode

Assign SEV score 1–5 based on severity of consequences

Assign EXP score 1–5 based on frequency of exposure

Assign BARRIER score 1–5 based on adequacy of existing controls

Calculate current RPN = SEV × EXP × BARRIER for each row

4. Recommended actions

Prioritise failure modes with highest RPN for corrective action first

Define recommended actions to reduce SEV, EXP or improve BARRIER effectiveness

Assign responsible person and target completion date for each action

5. Action results (post-implementation)

Document actions taken against each recommended action

Re-score SEV, EXP and BARRIER after actions are implemented

Calculate revised RPN and verify reduction is adequate

Update process control plan, HAZOP action list and audit checklist

Obtain sign-off from responsible engineer / QA manager; record revision number

Use the tables below to assign SEV, EXP and BARRIER scores (1–5 each). RPN = SEV × EXP × BARRIER. Maximum possible RPN = 125.

⚠️ Severity (SEV) — How bad is the consequence?

Score	Level	Description
5	Extreme	Fatality. Permanent extensive environmental damage beyond site boundary. Regulatory / legal citation. Exceeds permit requirements. Does not meet EHS legal requirements. Out of business.
4	Major	Non-reversible permanent disability; loss of sight; limb amputation; respiratory damage. Permanent environmental damage on site. Significant potential to exceed emissions limits / non-compliance. Bankruptcy.
3	Serious	Reversible loss of consciousness, burns, fractures. Reversible environmental damage on site. Potentially exceeds permit requirements. Significant property damage (>€10,000). Business loss >€1M.
2	Low	Recordable injury — requires medical attention. Adverse environmental effect below permit levels. Does not exceed IH exposure limits or EHS thresholds. Business loss ~€100,000.
1	Very Low	Minor bruising, cuts, light abrasions. Incident — no environmental impact. Full compliance with EHS requirements. Property damage <€250.

⏱️ Exposure (EXP) — How often is the hazard encountered?

Score	Level	Description
5	Frequent	Many cycles per day — continuous or near-continuous exposure during normal operations.
4	Daily	Exposure occurs every working day.
3	Weekly	Exposure occurs on a weekly basis.
2	Monthly	Exposure occurs on a monthly basis.
1	Seldom	Once a year or less — rare / infrequent exposure.

🛡️ Barrier / Probability — How effective are the existing controls?

Score	Level	Description
5	Expected	No barriers or controls in place for this hazard. Injury or environmental impact is likely to occur.
4	Likely	Single layer of control or barrier that can be unintentionally bypassed.
3	Moderate	Multiple layers of controls that can be unintentionally bypassed, or a single layer requiring deliberate bypass.
2	Unlikely	Multiple layers of controls that must all be deliberately bypassed to result in a failure.
1	Rare	Highly robust multiple independent protection layers. Failure requires extraordinary simultaneous events.

🎯 RPN Interpretation — Priority for corrective action

RPN Range	Priority	Required action
50 – 125	Critical 🔴	Immediate action required. Consider stopping work. Escalate to management. Implement interim controls at once before resuming.
20 – 49	High 🟠	Priority corrective action — assign ownership and target date within 30 days. Do not defer without documented justification.
8 – 19	Medium 🟡	Action required within 90 days. Review barrier effectiveness; consider additional or stronger controls.
1 – 7	Low 🟢	Monitor and review at next scheduled FMEA update. Document and retain as evidence of risk assessment.

Document Header — RA FMEA

Department

FMEA Number

Job Title

Date Prepared

Job Task / System

Date Revised

Prepared by

Revision Number

Reviewed by

Page

Identification & Ranking of Current Job Title / Task

Score each failure mode using the Rating Matrix tab. RPN = SEV × EXP × BARRIER (scores 1–5 each). Max RPN = 125.

Process Function & Requirements	Potential Hazard or Risk	Potential Effects of Failure (Consequences)	Potential Cause(s) / Mechanism(s) of Failure	Barrier Analysis (Current Risk Controls)	SEV	EXP	BARRIER	RPN
								—
								—
								—

Future / Completed Actions & Action Results

For each failure mode above, document the recommended action and re-score after implementation to show the revised RPN.

#	Recommended Actions	Responsibility	Target Date	Actual Completion	SEV	EXP	BARRIER	RPN
1								—
2								—
3								—

Root Cause Analysis — Method Guide & Checklist

Method Selection Guide

Pareto Chart

5 Whys Method

Fishbone (Ishikawa) Diagram

Scatter Plot

Failure Mode and Effects Analysis (FMEA)