🎓 Lesson 8
D5
Bi-Fuel Generator Efficiency Curve Modeling
A bi-fuel generator efficiency curve shows how well a generator converts fuel into electricity when running on two fuels—like diesel and natural gas—at different mix ratios.
🎯 Learning Objectives
- ✓ Calculate electrical efficiency at any diesel–natural gas substitution ratio using empirical curve-fit equations
- ✓ Design a bi-fuel dispatch strategy that maintains ≥92% of peak diesel-only efficiency across 30–100% load range
- ✓ Analyze efficiency degradation trends due to methane slip and lean-burn instability above 70% gas substitution
- ✓ Explain the physical causes of the 'efficiency dip' near 40–60% gas substitution using combustion thermodynamics
📖 Why This Matters
In off-grid mines—especially in remote regions like Western Australia’s Pilbara or Chile’s Atacama—diesel supply chains are vulnerable to disruption, price volatility, and carbon regulation. Bi-fuel generators (diesel + pipeline/liquefied natural gas) offer fuel resilience *and* up to 25% CO₂ reduction—but only if operated within their efficiency sweet spot. Misreading the efficiency curve leads to hidden OPEX penalties: a 5%-point efficiency drop at 15 MW scale wastes ~$380,000/year in fuel. This lesson equips you to model, validate, and deploy bi-fuel systems without sacrificing reliability or ROI.
📘 Core Principles
Bi-fuel efficiency is governed by three interdependent layers: (1) Thermodynamic: Diesel pilot ignition enables stable combustion of gaseous fuel, but excessive gas dilution lowers flame temperature and volumetric efficiency; (2) Control-system: ECU maps adjust pilot diesel quantity, injection timing, and air-fuel ratio to maintain combustion stability—each setting alters brake thermal efficiency (BTE); (3) System-level: Efficiency curves shift with ambient temperature, altitude, and engine aging—requiring derating corrections. The curve is typically concave-down, peaking near 20–30% gas substitution, dipping near 50%, then recovering slightly up to 70% gas—beyond which misfire risk dominates. Real-world validation requires ISO 8528-1 transient testing, not steady-state nameplate data.
📐 Empirical Efficiency Curve Model
The most widely adopted field model is a second-order polynomial fit to manufacturer test data, normalized to diesel-only baseline efficiency. It enables rapid dispatch planning without full CFD simulation.
💡 Worked Example
Problem: A Caterpillar G3520C bi-fuel generator has diesel-only BTE = 42.1% at 100% load. Field test data yields coefficients: a = −0.0012, b = 0.048, c = 0.998. Calculate BTE at 45% natural gas substitution (G = 0.45).
1.
Step 1: Apply the normalized BTE formula: η_norm = a·G² + b·G + c
2.
Step 2: Substitute values: η_norm = (−0.0012)(0.45)² + (0.048)(0.45) + 0.998 = −0.000243 + 0.0216 + 0.998 = 1.019357
3.
Step 3: Multiply by diesel baseline: η_BTE = 1.019357 × 42.1% = 42.91% → Round to 42.9%
Answer:
The predicted BTE is 42.9%, which exceeds the diesel-only baseline—indicating measurement uncertainty or interpolation error; verified field data shows maximum gain is +0.4% near G = 0.25, so this value should be capped at 42.5% per Caterpillar Technical Bulletin TB-1172.
🏗️ Real-World Application
At Newmont’s Boddington Mine (WA), a 24-MW bi-fuel plant replaced 100% diesel backup during LNG pipeline commissioning. Engineers used efficiency curve modeling to set automated dispatch limits: gas substitution held ≤35% below 60% load (to avoid lean misfire), and ≥20% above 80% load (to maximize CO₂ savings). Real-time SCADA integration with curve-derived efficiency thresholds reduced annual fuel cost by AUD $2.1M and cut Scope 1 emissions by 11,400 tCO₂e—while maintaining <0.5% unplanned outage rate over 3 years (vs. 2.3% pre-conversion).