ESD SIL Verification SIS

SIL Verification for Safety Instrumented Systems - IEC 61511

SIL Verification Cheat Sheet – Zohaib Jahan

SIL VERIFICATION CHEAT SHEET

IEC 61508 / IEC 61511 · Voting Architectures · PFDavg · Proof Test · Field-Ready Reference
IEC 61511 IEC 61508 SIS Design ESD / HIPPS / BMS FAT / SAT Reference
SIL Level Quick Reference
SIL Level PFDavg Range (Demand Mode) RRF Range Risk Reduction Typical O&G Application
SIL 1 ≥ 10⁻² to < 10⁻¹ 10 – 100 1 order of magnitude General process shutdown, low-risk utilities trips
SIL 2 ≥ 10⁻³ to < 10⁻² 100 – 1,000 2 orders of magnitude ESD, BMS, compressor protection, HIPPS (lower tier)
SIL 3 ≥ 10⁻⁴ to < 10⁻³ 1,000 – 10,000 3 orders of magnitude HIPPS, well head control, fire & gas critical SIFs
SIL 4 ≥ 10⁻⁵ to < 10⁻⁴ 10,000 – 100,000 4 orders of magnitude Nuclear / rarely justified in Oil & Gas
Core Reliability Parameters
PFDavg — Avg Prob. of Failure on Demand
PFDavg ≈ λD × TI / 2
λD = dangerous failure rate (per hr), TI = proof test interval (hr). Valid for 1oo1, low DC.
RRF — Risk Reduction Factor
RRF = 1 / PFDavg
Directly tells you how much safer the SIF makes the process. SIL 2 needs RRF ≥ 100.
Dangerous Failure Rate (λD)
λD = λ × (1 − SFF) or from SERH/PDS
Always use certified/validated failure rate data — exida SERH, SINTEF PDS, or vendor SIL cert.
Safe Failure Fraction (SFF)
SFF = (λS + λDD) / (λS + λD)
Governs hardware fault tolerance (HFT) requirements per IEC 61508 Table 2/3. SFF >90% is Type B target.
💡
PFDavg is an Average — Not a Point Value

PFDavg represents the average probability over the entire proof test interval. It is NOT the instantaneous probability at any single moment. Just before the proof test is due, actual PFD is at its highest.

⚠️
SIL is for the Entire SIF Loop — Not Just One Device

SIL verification must cover the full loop: sensor(s) + logic solver + final element(s). A SIL-certified transmitter alone does NOT give you a SIL-rated SIF. Every element contributes to the overall PFDavg.

COLOUR CODE:
Sensor / Initiator
Logic Solver (DCS/SIS)
Final Element (Valve/Actuator)
Safe State / Healthy
Degraded / Risk Elevated

VOTING ARCHITECTURES — BLOCK DIAGRAMS & FORMULAS

1oo1 · 1oo2 · 2oo2 · 2oo3 · 2oo4 | Sensor, Logic Solver, Final Element focus
1oo1 SIL 1–2
PT
S1
LS
XV
SAFE
PFDavg
≈ λD × TI / 2
Single channel. Simple, no redundancy. Any dangerous failure means the SIF fails. Low cost, lower integrity.
Low-risk utility shutdowns · SIL 1 SIFs
1oo2 SIL 2–3
PT S1
PT S2
LS
XV
SAFE
PFDavg (no CCF)
≈ (λD × TI)² / 3
Either sensor trips the SIF. High availability, lower spurious trips than 2oo2. Watch common cause!
ESD sensors · Fire detectors · Well head PT
2oo2 SIL 2
PT S1
PT S2
LS
&
XV
SAFE
PFDavg
≈ λD × TI (worse than 1oo1)
Both sensors must trip. Reduces nuisance trips. PFD is actually worse than 1oo1. Use only where spurious trips are expensive.
Compressor protection · Burner management · Critical valves
2oo3 SIL 2–3
PT S1
PT S2
PT S3
2/3
VOTE
LS
XV
SAFE
PFDavg (no CCF)
≈ 3 × (λD × TI / 2)² + β × λD × TI / 2
Industry gold standard. Best balance of safety integrity and availability. Tolerates one failure without spurious trip OR missed trip. β-factor (CCF) critical!
HIPPS · Main ESD · LNG / Refinery critical SIFs
2oo4 SIL 3
PT S1
PT S2
PT S3
PT S4
2/4
VOTE
LS
XV
SAFE
PFDavg (no CCF)
≈ 6 × (λD × TI / 2)³ (simplified)
Very high redundancy. Often used in logic solver subsystems (e.g. TMR controllers). Rarely seen at sensor level in O&G. High cost, complex proof testing.
TMR Logic Solvers · High-integrity HIPPS · Nuclear-adjacent
Architecture Comparison at a Glance
1oo1
Channels
1
Fails to danger
High
Spurious trip
Low
PFDavg
λD·TI/2
Max SIL
SIL 2
1oo2
Channels
2
Fails to danger
Low
Spurious trip
Medium
PFDavg
(λD·TI)²/3
Max SIL
SIL 3
2oo2
Channels
2
Fails to danger
High
Spurious trip
Low
PFDavg
λD·TI
Max SIL
SIL 2
2oo3
Channels
3
Fails to danger
Low
Spurious trip
Low
PFDavg
3(λD·TI/2)²
Max SIL
SIL 3
2oo4
Channels
4
Fails to danger
Very Low
Spurious trip
Low
PFDavg
Very Low
Max SIL
SIL 3+

PROOF TESTING · DIAGNOSTICS · COMMON CAUSE FAILURE

What kills your SIL — and what protects it
Proof Test Interval (TI) Effect on PFDavg
✅ Short TI — Better PFD
Halving the test interval approximately halves the PFDavg. A 1oo1 loop with TI = 1 year has double the PFDavg of TI = 6 months. TI is one of the most powerful levers in SIL design.
⚠️ PTC < 100% — Residual Risk Remains
Proof Test Coverage (PTC) means the fraction of dangerous failures your test actually detects. A 90% PTC leaves 10% of dangerous failures undetected — they accumulate until device replacement or a real demand.
❌ Skipped Tests — SIL Degradation
Skipping or deferring proof tests invalidates your SIL claim. The verification assumes TI is met. Every overdue test is an unverified SIF. Operations must NOT skip proof tests without an MOPC process.
Proof Test Formulas
PFDavg with Proof Test Coverage
PFDavg ≈ (1−PTC) × λD × TI/2 + PTC × λD × TI/2
Residual = (1−PTC) × λD × TI term. Higher PTC → lower residual dangerous failures. Target PTC > 90%.
Diagnostic Coverage (DC) Effect
PFDavg_eff ≈ λDD/(λDD+λDU) × λD×MTTR/2 + (1−DC) × λD×TI/2
λDD = dangerous detected, λDU = dangerous undetected. Online diagnostics (e.g. HART, partial stroke) reduce TI contribution significantly.
β-Factor — Common Cause Failure
PFDavg_CCF = β × λD × TI / 2
β typically 1–10% for redundant architectures. CCF contribution dominates in high-redundancy systems. Physical separation, diversity, and different manufacturers reduce β.
2oo3 with CCF (Full Form)
PFDavg ≈ 3(λDU×TI/2)² + β×λDU×TI/2
Second term (β×λDU×TI/2) often dominates when redundancy is high. Reducing β via design diversity is critical to achieve SIL 3 in 2oo3.
Diagnostic Coverage — Field Devices
Device / Feature Diagnostic Type Typical DC DC Level PFDavg Impact
Simple 4-20mA transmitter (no HART) None (manual proof test only) < 60% Low Full TI contribution to PFDavg
HART transmitter with EDDL diagnostics Online: sensor health, loop check 60–90% Medium Reduces residual PFDavg by detected failures
Smart transmitter + online proof test Full end-to-end functional test > 90% High Major PFDavg reduction. Supports longer TI.
Partial Stroke Test (PST) — ESD Valve Automated valve movement test 60–80% Medium Significant reduction in final element PFDavg
Full Stroke Test (FST) — ESD Valve Full travel proof test ≈ 99% High DC Near-complete dangerous failure detection
Logic Solver (SIL-certified, TMR) Internal self-diagnostics, watchdog > 99% Very High Logic solver PFDavg typically very small vs. field devices
⚠️
CCF Kills Redundancy

Two identical transmitters from the same vendor on the same impulse line share ALL the same failure modes. A plugged impulse line takes both out simultaneously. That is 100% CCF. Apply IEC 61508 Table D.4 separation guidelines.

How to Reduce CCF (β Factor)

Physical separation · Different impulse tapping points · Diversity in technology (DP + radar for level) · Independent cable routing · Independent power supplies · Different vendor devices where justified by risk.

BYPASS / OVERRIDE · WORKED EXAMPLE · COMMON MISTAKES

What degrades SIL in real operations — and how to manage it
Effect of Bypass / Override on SIL Integrity
🔶 SIF in Bypass

    ⚠ The SIF is effectively disabled — PFDavg becomes 1.0 (100% failure probability)
    ⚠ Any process demand during bypass period is unprotected
    ⚠ If bypass duration is significant vs. TI, your SIL number is degraded
    ⚠ Multiple concurrent bypasses = compounding integrity loss
    ✗ Bypass without MOC/MOPC = non-compliance with IEC 61511 Cl. 11.9

✅ Best Practice for Bypass Management

    ✓ Require written MOPC (Management of Process Change / Override) for every bypass
    ✓ Define maximum allowed bypass time in the SRS (Safety Requirements Specification)
    ✓ Compensating measures required: increased operator rounds, independent monitoring
    ✓ DCS/SIS should log bypass start/end time with operator ID for audit trail
    ✓ IEC 61511 requires bypass management procedure in your SMS (Safety Management System)

📐
Corrected PFDavg with Bypass Time

PFDavg_total = PFDavg_normal × (TI − t_bypass)/TI + 1.0 × t_bypass/TI
Where t_bypass = total bypass hours per interval. Even 24 hrs bypass per year on a 1-year TI SIF adds ~2.7% to PFDavg on top of baseline. For SIL 2 border cases, this can push you out of SIL 2 band.

Worked Example — ESD Pressure Trip (1oo2 Architecture)
High Pressure ESD — Separator Overhead · 2× PT + SIS + XV
STEP-BY-STEP
Scenario: Separator overhead ESD. Two pressure transmitters (1oo2 voting at sensor level). SIL-certified SIS logic solver. One ESD valve (XV). Proof test interval = 1 year. Sensor λDU = 5×10⁻⁷/hr, PTC=90%. Logic solver PFDavg = 1×10⁻⁴. ESD valve λDU = 8×10⁻⁶/hr (with PST, DC=70%). β = 5% (sensors).
01
Sensor PFDavg (1oo2)
3.8×10⁻⁴
3×(λDU×TI/2)² + β×λDU×TI/2
= 3×(0.0044)² + 0.05×0.0044
= 5.8×10⁻⁵ + 2.2×10⁻⁴ = 2.8×10⁻⁴
02
Logic Solver PFDavg
1.0×10⁻⁴
From SIL certificate. TMR internal diagnostics. Typically very small contribution.
03
Final Element PFDavg
3.0×10⁻³
PST + FST annually. ESD valves are often the dominant PFDavg contributor in the loop!
04
SIF Total PFDavg
≈ 3.4×10⁻³
Sum = 2.8×10⁻⁴ + 1.0×10⁻⁴ + 3.0×10⁻³ = 3.38×10⁻³
→ SIL 2 ✓ (10⁻³ to 10⁻²)
RRF ≈ 295
💡
Key Lesson from This Example

The ESD valve (final element) contributes ~88% of the total PFDavg. This is common. Investing in PST + FST on the valve is far more effective than adding a third sensor. Do NOT over-engineer the sensor side and ignore the valve.

Top 6 SIL Verification Mistakes in O&G Projects
01
SIL-certified ≠ SIL-capable loop
A certified transmitter does not make your SIF SIL-compliant. The entire loop must be verified. Device SIL cert is just one input.
02
Ignoring Final Element in PFD calc
ESD valve PFDavg is typically the largest contributor. Many engineers focus only on sensors and skip the valve — this is a serious error.
03
Wrong β value for CCF
Using default β = 1% when no physical separation exists between redundant sensors is unconservative. Properly assess using IEC 61508 Table D.4.
04
Proof test coverage claimed but not demonstrated
Claiming PTC = 90% without a written proof test procedure that actually achieves it. PTC must be demonstrated in the proof test plan, not assumed.
05
Architecture mismatch to SIL target
Specifying 2oo2 for a SIL 2 SIF — this often fails verification because 2oo2 has PFDavg worse than 1oo1. The SIL target drives architecture, not the other way.
06
Outdated failure rate data
Using MIL-HDBK-217 or old OREDA data for SIL verification. Use current PDS 2013 or exida SERH. Failure rates can differ by orders of magnitude.

FIELD ENGINEERING NOTES — FAT · SAT · MAINTENANCE

Practical lessons for design reviews, FAT, SAT, and ongoing SIS integrity management
Field Notes by Phase
🔬
Factory Acceptance Test (FAT)
Design Review → FAT Stage

    ‣ Proof test simulation in FAT must replicate the actual field proof test procedure — NOT just inject a signal at the marshalling cabinet
    ‣ Test each voting logic permutation. For 2oo3: test 1-trip, 2-trip, 3-trip and verify correct response at each
    ‣ Verify bypass function: engage bypass, confirm SIS does NOT trip on live signal, confirm bypass alarm activates in DCS
    ‣ Record all logic solver diagnostic alarms and verify correct annunciation in HMI
    ‣ Document proof test coverage achieved per device — this becomes your field baseline
    ‣ Test SIL-required response time: confirm the SIF trips within the required process safety time (PST)

🏭
Site Acceptance Test (SAT)
Commissioning → Pre-Startup

    ‣ SAT is NOT a repeat of FAT. SAT verifies the installed loop in real field conditions — actual process connections, real cables, real earthing
    ‣ Loop test must be end-to-end: apply stimulus at the sensor, confirm trip at the final element
    ‣ Verify impulse lines: no plugging, correct slope, no dead legs, correct manifold configuration (5-valve vs. 3-valve)
    ‣ Confirm ESD valve stroking: record open/close time, check actuator spring return, verify fail-safe position
    ‣ Check all SIS panel earthing, cable shields, and cable segregation from HV/power cables
    ‣ Pre-startup SIS punch list must be cleared before mechanical completion sign-off

🔧
Maintenance & Proof Testing
Operations Phase · Ongoing

    ‣ Never miss a proof test deadline. Overdue tests invalidate the SIL claim — escalate to FSM if test cannot be performed on time
    ‣ Use the approved Proof Test Procedure (PTP) every time. Do not improvise or shortcut — PTC claim depends on it
    ‣ Document every proof test: start time, end time, as-found condition, as-left condition, technician ID, any deviations
    ‣ As-found failures must be reported to the Functional Safety Manager (FSM) and trigger a revalidation review
    ‣ PST frequency on ESD valves must match what was assumed in the SIL calculation — verify during management review
    ‣ Trend your as-found failure rates. If actual λD exceeds assumed λD, your SIL claim is degraded — revalidate

Design Review — Common Gaps
FEED / Detailed Design · SIL Verification Stage

    ‣ SRS (Safety Requirements Specification) missing or incomplete — IEC 61511 Cl.10 mandates it before design starts
    ‣ SIL allocation from LOPA not reviewed against hardware architecture — mismatch causes late redesign
    ‣ No defined proof test procedure at design stage — PTC assumptions in SIL calc cannot be verified later
    ‣ ESD valve SIL certificate does not cover the actual trim, body, and actuator combination in the datasheet
    ‣ Logic solver SIL certificate scope does not match installed configuration (e.g. different I/O modules than certified)
    ‣ Functional Safety Assessment (FSA) not scheduled — IEC 61511 Cl.5.2 requires FSA before startup

Key Standards Reference
Standard Scope Key Clauses for SIL Verification Who Applies It
IEC 61511 (Ed.2) SIS for Process Industry Cl.10 SRS · Cl.11 SIS Design · Cl.12 Integration · Cl.16 Proof Test Process plant engineers, FSM, EPC contractor
IEC 61508 (Ed.2) Functional safety of E/E/PE systems Part 2: Hardware · Part 3: Software · Part 6: Annex B (PFD formulas) SIS vendors, system integrators, device manufacturers
IEC 61511-3 Guidance for LOPA & SIL determination LOPA methodology · Risk graph · SIL allocation Process safety engineers, HAZOP/LOPA team
ISA-84.00.01 US equivalent of IEC 61511 Aligns with IEC 61511 — additional US-specific guidance US-based projects (refining, chemicals)
Architecture Selection Quick Guide
🎯 SIL 1 Target
1oo1 is usually sufficient. Simple, cheap, maintainable. If spurious trips are a concern, consider 2oo2 with a redundant sensor — but verify PFDavg still meets SIL 1. Check final element condition carefully — it dominates.
🎯 SIL 2 Target
1oo2 at sensor side is the most common solution. Provides redundancy without the availability penalty of 2oo2. For final elements, a single SIL-certified ESD valve with PST often suffices. Verify the whole loop — don't just count channels.
🎯 SIL 3 Target
2oo3 at sensor side is standard. Requires careful CCF analysis — β factor management is critical. Final element: two independent valves or one high-integrity valve with PST + FST + full SIL cert. Logic solver must be TMR or equivalent. Independent SIL verification tool required.
📌
Remember — SIL is a Lifecycle Concept, Not a Design-Phase Checkbox

SIL integrity must be maintained from HAZOP through decommissioning. A properly designed SIF that is poorly maintained, bypassed without control, or never proof-tested to plan is NOT a SIL-rated SIF — regardless of what the design report says. The Functional Safety Manager (FSM) owns this lifecycle.

👷
Zohaib Jahan
TÜV Certified Functional Safety Engineer (FSE)
ICSS Engineer with 11+ years of Experience in Oil & Gas and Mining including Greenfield & Brownfield Projects. Let's discuss if I can be of any assistance!