Hazard identification: HAZOP and HAZAN

HAZOP for identifying deviations with guide words on the P&ID, and HAZAN for quantifying how often and how big, including fault trees and trip fractional dead time.

Drafted with Aria, reviewed by the AiCanCode.org team. Spotted an error? Use Give Feedback at the bottom of the page.

Why it matters

Major accidents in chemical plants almost always come from a deviation nobody thought through: a valve left closed, a reverse flow, a trip that had failed unnoticed. HAZOP is the structured way to find such deviations before the plant is built; HAZAN puts numbers on the ones that matter so that money is spent where it reduces risk most. Both are standard practice and frequent interview topics.

Key ideas

Hazard identification versus hazard analysis. Kletz's distinction: identification asks "what can go wrong?" (HAZOP, what-if, checklists, FMEA); analysis asks "how often, how big, and so what?" (HAZAN, quantitative risk assessment). Identification comes first; there is no point quantifying a hazard that has not been found.

HAZOP (hazard and operability study).

  • A multidisciplinary team (process, instrumentation, operations, maintenance, safety, led by an independent chairman with a scribe) examines the P&ID line by line.
  • The plant is divided into nodes (a line or vessel). For each node the design intention is stated (e.g. "transfer 10 m³/h of toluene at 3 bar to the reactor").
  • Guide words are combined with parameters (flow, temperature, pressure, level, composition, reaction, phase) to generate deviations:
    • No / None — complete negation (no flow);
    • More / Less — quantitative increase or decrease (more pressure, less flow);
    • As well as — qualitative increase, something extra present (water in the feed, an extra phase);
    • Part of — qualitative decrease, something missing (one component absent);
    • Reverse — the logical opposite (reverse flow);
    • Other than — complete substitution (wrong material, a different operating mode such as start-up or maintenance);
    • time-based words such as Early, Late, Before and After are used for batch and sequential operations.
  • For each meaningful deviation the team records causes, consequences, existing safeguards and actions (design changes, alarms, trips, procedures). Operability problems (off-spec product, difficult start-up) are recorded too.
  • HAZOP is qualitative, thorough and time-consuming; its quality depends on the team and on an up-to-date P&ID.

Related identification methods. Checklists (quick, but only find known problems); what-if analysis (less structured brainstorming); FMEA (failure modes and effects analysis, equipment-by-equipment, often ranked with a risk priority number = severity × occurrence × detection).

HAZAN (hazard analysis).

  • How often? Event frequencies are built up with fault trees (top event traced back through AND/OR gates to basic failures) and outcomes traced forward with event trees (ignition or not, escalation or not).
  • How big? Consequence models for fire, explosion and toxic release estimate the effect on people and plant.
  • So what? The calculated risk is compared with criteria, e.g. a fatal accident rate (FAR) target or an individual-risk limit, and additional protection is added until the risk is acceptable or as low as reasonably practicable (ALARP).
  • Protective systems. A trip that fails unrevealed is "dead" between tests. Its average unavailability (fractional dead time) depends on its failure rate and test interval; the hazard rate is the demand rate times the fractional dead time. Testing more often, or adding a second independent trip, reduces it.

Layers of protection. Inherent safety (less inventory, milder conditions), basic process control, alarms and operator response, safety instrumented trips, relief devices, bunds and emergency response — each independent layer multiplies down the frequency of the final event. LOPA (layer of protection analysis) is a simplified semi-quantitative form of HAZAN.

Formulas

Risk (expected loss rate): R = F·C OR gate (independent rare events, frequencies): F_top ≈ F₁ + F₂ + … AND gate (a demand at rate D meets a protective layer failed with probability p): F_top = D·p AND gate of independent probabilities: P = P₁·P₂·… Fractional dead time of a single tested trip: fdt ≈ f·T / 2 (valid when f·T ≪ 1) Two independent identical trips (either can act): fdt ≈ (f·T)² / 3 Hazard rate: H = D·fdt Fatal accident rate: FAR = fatalities per 10⁸ hours of exposure

Symbols: F frequency (events/yr); C consequence per event (e.g. ₹ or fatalities); R risk (consequence per yr); D demand rate on a trip (/yr); f failure rate of the trip (unrevealed, /yr); T proof-test interval (yr); fdt fractional dead time (–); H hazard rate (/yr); p probability of failure on demand (–).

Worked examples

Example 1 (standard). A storage tank overflows if filling is not stopped. Filling fails to stop if the level controller fails (0.2 /yr) or the operator mis-sets the batch (0.1 /yr). An independent high-level trip has a probability of failure on demand of 0.01. Find the overflow frequency.

  1. Demand on the trip (OR gate): D = 0.2 + 0.1 = 0.3 /yr.
  2. Overflow (AND gate): F = D·p = 0.3 × 0.01 = 0.003 /yr.
  3. Overflow frequency = 3 × 10⁻³ per year, about once in 333 years.

Example 2 (GATE level). A high-pressure trip on a reactor has an unrevealed failure rate f = 0.5 /yr and is proof-tested monthly. The demand rate (pressure excursions) is 1 /yr. (a) Find the hazard rate. (b) Find the hazard rate if a second, identical, independent trip is added and both are tested monthly.

  1. T = 1/12 yr; f·T = 0.5/12 = 0.0417 (≪ 1, so the approximations hold).
  2. (a) fdt = f·T/2 = 0.0417/2 = 0.0208; H = D·fdt = 1 × 0.0208 = 0.0208 /yr — once in 48 years.
  3. To reach H = 0.001 /yr with one trip would need T = 2 × 0.001/0.5 = 0.004 yr ≈ 1.5 days — impractical.
  4. (b) fdt ≈ (f·T)²/3 = 0.0417²/3 = 5.79 × 10⁻⁴; H = 5.79 × 10⁻⁴ /yr — once in about 1700 years.
  5. (a) 0.0208 per year; (b) 5.8 × 10⁻⁴ per year. Redundancy reduced the hazard rate about 36-fold.

Common mistakes

  • Treating HAZOP as quantitative or HAZAN as a brainstorming method.
  • Applying guide words to a PFD instead of the P&ID, or without stating the design intention.
  • Mixing frequencies (/yr) and probabilities (–): an AND gate multiplies a frequency by a probability, never two frequencies.
  • Adding probabilities at an OR gate when they are not small and independent.
  • Forgetting the factor ½ in fractional dead time, or using it when f·T is not small.
  • Confusing "more" (quantitative) with "as well as" (qualitative addition).

For GATE CH

Expect conceptual questions on HAZOP guide words and their meanings, the difference between HAZOP and HAZAN, and simple numericals: fault-tree combination of frequencies and probabilities, fractional dead time and hazard rate of a trip, or FAR. Practise distinguishing frequencies from probabilities.

Quick check

  1. Which guide word describes "impurity present in the feed"?
  2. Demand rate 2 /yr; trip probability of failure on demand 0.005. Hazard rate?
  3. Trip failure rate 0.2 /yr, tested every 6 months. Fractional dead time?
  4. What document does a HAZOP team work from?

Answers: 1. As well as. 2. 0.01 /yr. 3. 0.2 × 0.5/2 = 0.05. 4. The P&ID.

Interview questions

All Plant Design and Economics interview questions

Try answering each one aloud before you open it.

  1. 1.What is HAZOP and why is it important in chemical plant design?Concept

    HAZOP, or Hazard and Operability Study, is a structured and systematic examination of a planned or existing process or operation. It is used to identify and evaluate problems that may represent risks to personnel or equipment, or prevent efficient operation. HAZOP is important because it helps in identifying potential hazards and operability issues early in the design phase, allowing for the implementation of corrective measures to mitigate risks.

  2. 2.Explain the concept of HAZAN and its role in risk management.Concept

    HAZAN, or Hazard Analysis, is a process used to assess the risk associated with identified hazards. It involves quantifying the likelihood and potential impact of hazardous events. HAZAN plays a crucial role in risk management by providing a basis for decision-making on risk reduction measures, prioritizing risks, and allocating resources effectively to manage those risks.

  3. 3.How does a HAZOP study differ from a HAZAN analysis?Concept

    A HAZOP study is primarily a qualitative technique focused on identifying potential hazards and operability problems in a process. It uses guide words to systematically examine deviations from design intent. In contrast, HAZAN is a quantitative analysis that assesses the risk associated with the hazards identified in a HAZOP study. HAZAN involves calculating the probability and consequences of hazardous events to prioritize risks and determine necessary control measures.

  4. 4.Why are guide words used in HAZOP studies, and can you give an example?Application

    Guide words are used in HAZOP studies to help systematically identify deviations from the design intent of a process. They prompt the team to consider how different parameters might deviate from normal operation. An example of a guide word is 'NO' or 'NOT', which might be used to consider what happens if a flow does not occur when it should.

  5. 5.What happens if a potential hazard is identified during a HAZOP study?Application

    If a potential hazard is identified during a HAZOP study, the team will assess the severity and likelihood of the hazard. They will then propose recommendations to mitigate the risk, which could include design changes, additional safety systems, or operational procedures. The goal is to reduce the risk to an acceptable level before the process is implemented.

  6. 6.In what scenarios would a HAZAN analysis be particularly useful?Application

    A HAZAN analysis is particularly useful in scenarios where there is a need to prioritize risks and allocate resources effectively. It is beneficial when dealing with complex processes where multiple hazards have been identified, and there is a need to quantify the risks to determine which hazards require immediate attention and which can be managed with existing controls.

  7. 7.What are the potential consequences of not conducting a HAZOP study in a chemical plant design?Application

    Not conducting a HAZOP study can lead to unidentified hazards and operability issues, which may result in accidents, equipment damage, or operational inefficiencies. This oversight can lead to increased costs due to downtime, repairs, and potential regulatory fines. Moreover, it poses significant safety risks to personnel and the environment.

  8. 8.In an FMEA, a failure mode is rated severity 7, occurrence 4 and detection 3. What is its risk priority number, and how is it used?Numerical

    RPN = severity × occurrence × detection = 7 × 4 × 3 = 84. FMEA (failure modes and effects analysis) is an equipment-by-equipment identification technique that complements HAZOP; RPN is used to rank failure modes so the highest are tackled first. It is a relative ranking on arbitrary scales, not a true risk value, so high-severity items should be reviewed even when their RPN is moderate.

  9. 9.If a HAZOP study identifies a deviation with a high likelihood but low severity, how should it be addressed?Application

    A deviation with a high likelihood but low severity should still be addressed, as frequent occurrences can lead to cumulative effects or operational inefficiencies. The team might implement measures to reduce the likelihood of occurrence, such as improved maintenance or monitoring systems, to ensure the process remains efficient and safe.

  10. 10.A hazardous event is estimated to occur 0.01 times per year and would cause a loss of ₹5 crore. What is the expected annual loss, and how is it used?Numerical

    Expected annual loss (risk) = frequency × consequence = 0.01 /yr × ₹5 crore = ₹5 lakh per year. It can be compared with the annualised cost of a protective measure: a trip costing ₹2 lakh a year that cuts the frequency tenfold reduces the risk by ₹4.5 lakh a year and is justified. Where people could be harmed, risk is judged against safety criteria (individual risk, FAR, ALARP), not cost alone.

Finished this topic? Mark it so your progress, study plan and readiness keep up.

Stuck on something here?

PreviousCost of heat exchangers, columns and reactorsNext Environmental and safety considerations in plant design