LOPA for Refinery Unit is where quantitative risk reasoning meets the complex reality of hydrocarbons processing, turning qualitative HAZOP findings into numerical risk decisions that stand up to internal scrutiny and regulator review. In refineries—from crude and vacuum distillation to FCC, hydrocracker, reformer, alkylation, delayed coker, SRU/TGTU, tank farms, and loading terminals—the stakes are high: loss of containment, fires, and vapor cloud explosions are major accident hazards, and decision-makers need a consistent way to judge whether safeguards are sufficient or more layers are required. This article explains how Layer of Protection Analysis fits into a refinery’s process safety lifecycle, how it connects to IEC 61511 and corporate risk criteria, and how practitioners can make the method robust, transparent, and useful for operations, maintenance, and compliance alike.
Who Does This Affect?

Because LOPA sits at the intersection of engineering, operations, and governance, it touches a wide range of roles and functions in a refinery. Understanding who is involved helps set expectations and defines the quality of data feeding the analysis.
- Process safety engineers who translate HAZOP scenarios into quantifiable event frequencies and risk targets, and who align outcomes with IEC 61511 functional safety requirements.
- Operations leaders and control room supervisors who provide realistic insights on alarm response times, staffing levels, shift patterns, and human reliability under upset conditions.
- E&I/instrumentation and control engineers who own the basic process control system (BPCS), safety instrumented systems (SIS), fire and gas (F&G) systems, proof test intervals, bypass management, and Cause & Effect logic.
- Mechanical and process engineers responsible for pressure relief systems (PSVs/rupture discs), flare design, utilities, and interlocks that are often credited as independent protection layers.
- Maintenance and reliability teams that manage CMMS histories, proof test records, bypass logs, and failure data that give LOPA its empirical grounding.
- HSE managers and compliance officers who must demonstrate ALARP/Tolerability of Risk to corporate boards and UAE regulators, including ADNOC group requirements, OSHAD SF (Abu Dhabi), Dubai Municipality, Civil Defence, and MOHRE where relevant.
- Project and MOC teams that introduce new scenarios and design safeguards during upgrades, tie-ins, debottlenecking, or unit restarts after turnarounds.
Refinery unit managers, integrity teams, and even commercial planners are indirectly affected because LOPA decisions can set operating limits, dictate shutdown logic, or influence turnaround scopes. As such, it’s not just a “process safety exercise”; it’s a cross-functional risk decision framework with implications for production and reliability.
Why This Topic Matters in Refining Operations

LOPA’s importance flows from the unacceptable consequences of rare but catastrophic refinery events, combined with the practical need to prioritize resources among many competing risks. A well-run LOPA program helps leadership answer questions such as: Does this overpressure scenario truly require a SIL-rated trip in addition to the PSV? Is operator intervention adequately reliable during a night shift with parallel upsets? Can we justify crediting firewater deluge as an IPL for pool fire escalation, given inspection and impairment practices?
From a compliance standpoint, LOPA strengthens the link between hazard identification (e.g., HAZOP/HAZID) and functional safety (IEC 61511), including Safety Requirements Specifications (SRS), SIL determination, and SIL verification. UAE regulators and industrial stakeholders—ADNOC HSE guidance, OSHAD, and Civil Defence in particular—expect risk reduction claims to be evidence-based, not aspirational. LOPA provides a transparent way to compare initiating event frequencies to tolerable event frequencies, and to demonstrate that credited layers are independent, effective, and auditable. This is equally relevant for greenfield projects and for brownfield units where modifications accumulate over years and documentation must be reconciled with operating reality.
Finally, beyond compliance and the prevention of major accidents, LOPA helps drive operational discipline. By clarifying which safeguards carry the risk reduction burden, the method encourages targeted maintenance, realistic bypass governance, smarter alarm design, and pragmatic proof test intervals—aligning safety with plant reliability and uptime.
Key Points to Understand

While LOPA is often presented as a “simple” semi-quantitative method, its credibility hinges on consistent definitions, realistic data, and disciplined treatment of independence and common cause. The following elements are foundational when applying LOPA in a refinery unit.
1) Scenario definition after HAZOP

LOPA begins with a clearly defined hazardous event identified in HAZOP, Bowtie, or HAZID. Each scenario must specify the process context, initiating cause (e.g., control valve failure, cooling loss, operator mis-set), the hazardous consequence (overpressure, thermal runaway, loss of containment), and the boundary conditions (inventory, phase, pressure, temperature) that influence severity and ignition potential. HAZOP deviations such as “high pressure in debutanizer” or “loss of reflux to reformer stabilizer” become concrete LOPA scenarios once the initiating event and consequence are pinned down.
2) Initiating event frequency

Event frequency estimates should draw from site reliability data, CMMS history, relevant generic sources (e.g., OREDA, CCPS), and engineering judgment moderated by peer review. Nominal values (e.g., 0.1 per year for certain control loop failures) must be calibrated against local performance. Applying auditing discipline here prevents “wishful thinking.” Where data are sparse, conservative ranges are used with sensitivity discussion.
3) Enabling conditions and conditional modifiers

Enabling conditions (e.g., blocked-in operations, startup modes) and conditional modifiers (e.g., probability of ignition, probability of occupancy, probability that a release reaches a receptor) are applied only when justified and documented. Crediting a low occupancy factor in a constantly staffed control room, or a low ignition probability in an FCC unit with known hot surfaces, will rightly be challenged. These modifiers should reflect site specifics—layout, inventory, ventilation, and workforce patterns.
4) Independent Protection Layers (IPLs)

An IPL must be independent, effective, and auditable. Independence means the IPL’s success is not undermined by the same cause that created the initiating event (no common logic, power, or instrumentation dependencies). Effectiveness is quantified via probability of failure on demand (PFDavg) and must be credible given design, testing, and impairment practices. Auditability means there is a way to verify performance—proof testing records, impairment logs, bypass controls, and management of change. BPCS loops are rarely credited as full IPLs due to shared dependencies, while PSVs, dedicated interlocks, SIS trips, and fire/gas systems can be credited if they meet criteria.
5) Risk targets and ALARP

Corporate tolerable event frequency (TEF) criteria—often set by corporate risk policy or aligned with API RP 754 consequence categories—anchor LOPA calculations. IEC 61511 ties LOPA outcomes to SIL determination: if the sum of credited IPLs cannot reduce risk to the TEF, additional risk reduction (possibly via a SIF with an assigned SIL) is needed. The ALARP principle guides decisions where risk is close to tolerability limits, requiring justification of further controls or documented reasoning for not adding more layers.
6) Verification and lifecycle integration
LOPA does not end with a number. If a SIF is derived, SIL verification checks that achievable PFDavg under realistic proof test intervals and diagnostic coverage actually meets the target. Cause & Effect matrices, the SRS, and bypass management translate LOPA decisions into operational logic. Functional Safety Assessments (FSA) validate the overall lifecycle consistency.
Illustrative scenario and LOPA arithmetic
Below is a simplified example that captures the basic arithmetic. Actual studies warrant deeper scrutiny of dependencies, test intervals, β-factor common cause assumptions, and human factors.
| Scenario | Initiating Event (IE) Frequency | Credited IPLs | PFDavg (per IPL) | Combined Risk Reduction | Residual Event Frequency | Target (TEF) |
|---|---|---|---|---|---|---|
| Debutanizer overpressure leads to column rupture | 0.1 per year | PSV to flare; High-High Pressure SIS trip | PSV: 0.01; SIS trip: 0.1 (SIL1) | 1/(0.01 x 0.1) = 1,000 | 0.1 / 1,000 = 1.0E-4 per year | ≤ 1.0E-4 per year |
In this simplified case, the combination of a well-maintained PSV and a SIL 1 SIS trip achieves the corporate target. In practice, ignition probability and occupancy may be considered if the hazardous consequence is “fatality” rather than “loss of containment,” but they must be justified and consistently applied.
LOPA for refinery unit contexts must also reconcile dependencies: if the SIS valve shares motive power with the BPCS final element, or if both rely on a common I/O card or power supply, independence erodes and PFDavg credit must be reduced accordingly. Documentation and site interviews are essential to flush out these details.
How This Works in Practice
Translating the method into day-to-day refinery decision-making involves data gathering, cross-functional discussions, and alignment with the plant’s management systems. The steps below illustrate typical practice without implying a fixed “project” format.
- Trace scenarios back to source nodes: Start with the HAZOP node and deviation, confirm the initiating cause, and define the hazardous event. For refineries, this often relates to utility failures (cooling water, steam), control loop failures, or operator errors during startup/shutdown.
- Pull local performance records: CMMS histories for valves, transmitters, relief devices, and interlocks—plus bypass logs, alarm floods, and proof test records—anchor the analysis. Refinery performance during summer heat loads and Ramadan staffing patterns can affect human response credit.
- Set risk targets: Confirm the corporate TEF and escalation thresholds (e.g., whether an event leading to offsite impact requires stricter targets). Align with ADNOC HSE expectations, OSHAD SF, and Civil Defence where applicable, particularly for events with fire/explosion potential.
- Test IPL independence: Validate that each credited layer has its own sensor, logic, and final element paths per IEC 61511 guidance. For PSVs, confirm relieving cases, mechanical integrity, and impairment management. For F&G, verify detector coverage and fault tolerance.
- Calculate and sense-check: Combine IE frequency with IPL PFDavg and any justified modifiers, then peer-review results for plausibility. If the numbers are “too perfect,” revisit assumptions.
- Integrate outcomes: Update Cause & Effect matrices, operating envelopes, SRS, proof test intervals, and alarm setpoints so operations can run the plant within the risk envelope LOPA assumes.
- Close the loop: Track actions via MOC and include LOPA results in pre-startup safety reviews (PSSR). Ensure proof testing, bypass management, and functional testing are scheduled in the CMMS.
This cycle helps keep the analysis a living part of refinery governance rather than a binder on a shelf. Importantly, it also feeds into broader risk tools—QRA for societal risk, Bowtie for barrier visualization, and KPI dashboards for functional safety and process safety metrics (e.g., API RP 754 Tiers).
Benefits and Practical Use Cases
When LOPA is performed with discipline and shared understanding across teams, the benefits show up in both safety and operational outcomes.
- Clarity on “what really keeps us safe”: Distinguishing between housekeeping safeguards and true IPLs focuses resources on the few layers that carry most of the risk reduction burden.
- Rationalized alarm management: Insights about human response latency during concurrent upsets can lead to fewer, more meaningful alarms and improved alarm shelving rules.
- Right-sized SIL and proof test intervals: Avoid over-specifying SIFs where PSVs and inherently safer design suffice, and avoid under-specifying where a SIF is the only credible last line of defense.
- Stronger regulator dialogue: Evidence-based discussion with ADNOC HSE, OSHAD, Dubai Municipality, and Civil Defence is more straightforward when assumptions and credits are transparent and traceable.
- Turnaround and maintenance prioritization: IPLs critical to risk reduction rise to the top of outage scopes and impairment planning.
- Integration with project decisions: Brownfield modifications and tie-ins can be evaluated for net risk impact, enabling smarter design choices and interlock strategies.
Concrete use cases in refineries include confirming whether hydrotreater reactor high-temperature SIFs require SIL 2 under hydrogen services, establishing realistic credit for operator response to fractionator tower flooding alarms during night shifts, validating PSV adequacy for coker drum overpressure during steam-out, and determining the role of F&G detectors for pump bay escalation scenarios adjacent to tank farms.
Best Practices and Points of Caution
The strongest LOPA programs are built on consistency, evidence, and humility about uncertainty. These practices help keep analyses credible and useful.
- Define and police IPL independence: Check for shared instruments, logic solvers, power, utilities, or final elements. Independence must be technical and organizational—an operator action cannot be credited as an IPL if alarm floods routinely overwhelm the panel.
- Use site data where it counts: Generic PFDavg values are a starting point, not a destination. Proof test intervals, diagnostic coverage, and bypass culture significantly influence real performance.
- Quantify human response conservatively: Consider workload, lighting, noise, alarm rate, and access routes. Human error rates improve with training, but not to perfection—document realistic ranges.
- Beware of double counting: Do not count the same safeguard twice (e.g., a BPCS loop and a SIS loop that share a sensor or a final element). When in doubt, discount or consolidate.
- Align LOPA with IEC 61511 lifecycle: If LOPA derives a SIF, trace that requirement into the SRS, proof testing regime, and bypass governance. Conduct FSAs to assure lifecycle integrity.
- Integrate with Bowtie and QRA: Ensure consistency in consequence definitions, ignition probabilities, and escalation pathways used across tools.
- Treat conditional modifiers with discipline: Occupancy, ignition probability, and weather-related dispersion should be evidence-based, not optimistic.
- Capture cybersecurity and impairment risks: Consider how cyber or common-cause impairments could degrade SIS or F&G effectiveness; coordinate with IEC 62443-aligned practices.
For practical governance examples, industry professionals often consult specialist libraries and curated technical notes. A useful starting point for structured materials and thought leadership is CABINET NPM, which many practitioners reference alongside standards like IEC 61511, ISO 45001, API recommended practices, and recognized CCPS texts to triangulate consistent assumptions and documentation approaches.
Frequently Asked Questions About LOPA in Refining
How is LOPA different from HAZOP?
HAZOP is a systematic, qualitative method for identifying deviations, causes, and safeguards within a process node; it excels at finding scenarios and generating actions. LOPA is a follow-on quantitative step that takes selected scenarios and estimates event frequencies, assesses safeguard credits, and compares residual risk to a tolerable threshold. HAZOP tells you “what could go wrong” and “what safeguards exist”; LOPA quantifies “how likely” and “how much risk reduction is enough.”
When should we apply LOPA during a project or modification?
LOPA is most effective after a HAZOP has identified scenarios and before detailed design locks in decisions on SIS, PSVs, and F&G coverage. For brownfield changes, apply LOPA during MOC and before PSSR to validate that new interlocks or safeguards deliver the required risk reduction. It is also common to re-open LOPA for units undergoing debottlenecking or for which incident learnings suggest that assumptions have drifted from reality.
Can operator response be credited as an IPL?
It can, but only with caution. The response must be specific, trained, and supported by clear alarms and procedures. Credit is typically modest (e.g., PFDavg values reflecting human error probabilities) and should consider staffing levels, alarm loads, and concurrent tasks. Night shifts, alarm floods, and complex manual actions warrant more conservative credit.
Is fire and gas detection a valid IPL for refinery scenarios?
F&G can be an IPL if it meets independence, effectiveness, and auditability criteria. For gas detection, you need credible coverage (detector density and placement), reliable voting logic, and demonstrable automatic actions (e.g., ESD, deluge) with proven availability. Credit should be scenario-specific; detection that only alarms without automatic mitigation may provide limited risk reduction.
How does LOPA relate to SIL determination under IEC 61511?
LOPA informs SIL determination by quantifying the shortfall between initiating event frequency and the tolerable event frequency, after accounting for credited IPLs. If more risk reduction is needed, a SIF may be assigned a SIL target. SIL verification then demonstrates that the designed SIF (considering proof test intervals, diagnostics, and common cause) achieves the required PFDavg over its lifecycle.
What data sources are acceptable for initiating event frequency and IPL performance?
Prefer site-specific data from CMMS, incident records, proof tests, and impairment logs. When local data are limited, use recognized generic sources (e.g., CCPS, OREDA) and align with corporate guidance. Always disclose sources and uncertainties, and use sensitivity checks to avoid overconfidence in single-point estimates.
How often should we update our LOPA for a refinery unit?
Review LOPA when there are significant changes: MOC affecting safeguards, alarm design, or operating envelopes; after notable incidents; during major turnarounds; or on a periodic cycle aligned with functional safety assessments. The goal is to keep the analysis synchronized with how the unit actually runs and how safeguards actually perform.
Does LOPA apply to utilities like boilers and flare systems?
Yes. LOPA is relevant wherever credible hazardous scenarios exist—boilers (water level, overpressure), flare systems (overload, flameout), power distribution (loss of critical supply), and cooling water systems (loss causing thermal stress). Many refinery upsets originate in utilities, so the same discipline applies.
Conclusion
In refining, the value of LOPA lies in how faithfully it mirrors the plant’s realities: actual initiating frequencies, genuinely independent layers, human factors under pressure, and alignment with IEC 61511 and corporate risk policy. By grounding assumptions in site data and integrating results with Cause & Effect logic, SRS documents, proof testing, and bypass governance, organizations avoid “paper safety” and achieve risk reduction that holds up in front of operations teams and regulators alike. When done well, LOPA becomes the connective tissue between HAZOP findings, SIL allocation, and day-to-day operating discipline—ensuring the refinery’s most critical barriers are the ones that receive attention, budget, and care.
To go further on this topic, NPM’s broader LOPA Study Services UAE page covers this in more depth.
Our References
NPM’s references illustrate how disciplined risk reasoning translates into safer, more reliable operations across a wide range of high-hazard industries. Our portfolio spans refineries, petrochemical complexes, gas processing plants, storage terminals, utilities, and power generation assets, as well as construction and logistics environments where process safety overlaps with occupational risk. In our reference projects, readers will find examples of how LOPA outcomes connect to IEC 61511 lifecycle deliverables, how HAZOP findings are elevated into actionable requirements, and how barrier management frameworks, including Bowtie and QRA, are reconciled with operating realities. The references also show how decisions about PSVs, SIFs, and fire and gas coverage are supported by transparent assumptions, realistic human factors considerations, and maintenance data that stand up to internal assurance and regulator review.
Beyond downstream installations, NPM’s references include complex brownfield modernizations where documentation and reality needed to be brought into alignment; turnarounds in which IPL availability and impairment risks had to be managed hour by hour; and multi-asset governance programs where consistent LOPA and SIL practices were rolled out at scale. Readers can observe how we synthesize initiating event data, CMMS histories, and proof test records into defensible risk models, and how management systems—MOC, PSSR, alarm management, and bypass control—are tuned so that credited layers remain effective through the full operating cycle. Whether the facility is a large integrated complex or a focused single-unit installation, the through-line in the references is the same: clarity of scenarios, integrity of assumptions, and a living connection between risk models and the way the plant actually runs.
For more information about our references, visit Our References.
Speak with NPM
If you would like to discuss how LOPA assumptions map to your refinery’s operating context, our team is available to share perspectives and help you think through practical considerations. Whether you are planning a HAZOP, reviewing SIL targets, or preparing for a regulatory dialogue, a brief conversation can clarify next steps. For any further questions or to speak with our team, Contact Us.