How Failed Fuel Injectors can lead to Diesel Engine Failures

 

Article by Bill Gillette (LogiLube) and Sanya Mathura (Strategic Reliability Solutions)

​

In heavy-duty diesel engines, failed fuel injectors are often the starting point for a much larger chain of destructive events. Across locomotive fleets, mining haul trucks, underground mining equipment, drilling machines, marine engines, and stationary power systems, injector degradation remains one of the leading contributors to fuel dilution, thermal stress, piston damage, and catastrophic engine failures. Understanding how we can prevent their failures are of utmost importance. 

​

Fuel injectors operate in very harsh and demanding environments with injection pressure as high as 35,000 psi. Often the operating conditions involve extremely high pressures, thermal cycling, microscopic tolerances, contaminated fuel exposure and continuous vibration. With these conditions, it is easy to understand how injector nozzles can erode, stick, crack, carbon-foul or lose atomization efficiency. A poor spray pattern can lead to incomplete combustion further causing excess diesel fuel to wash down the cylinder walls and migrate into the crankcase oil.  

​

The result is progressive viscosity collapse, increased soot generation, accelerated oxidation, and rapidly increasing wear rates.

​

Injector failures are commonly caused by:

• Fuel contamination

• Water in fuel

• Microbial growth in fuel

• Poor fuel filtration

• Low-quality fuel

• Fuel waxing / gelling

• Fuel lubricity deficiency

• Abrasive particulate contamination 

• Nozzle tip coking & deposits

• Hole erosion / cavitation erosion

• Thermal fatigue

• Poor combustion

• Extended idle conditions

By understanding these failure modes, we can better prepare maintenance plans and correlate the observed failures with oil analysis results. Utilizing LogiLube’s patented SmartOil® G3 Edge-AI Brain™ powering Autonomous Fluid Intelligence™, we can continuously monitor the engine oil condition and determine if intervention is required. Having a high fidelity, real-time, continuous data set correlated to actual operational activities can also help to determine the remaining-useful-life (RUL) of the oil, not just for the engine being sampled, but for the entire fleet of engines. 

​

Depending on the oil viscosity parameters initially programmed during SmartOil® commissioning (upper/lower control limits, rate of change), proprietary algorithms processed by the G3 Edge-AI Brain™ can detect a lubricant condition anomaly very early. The SmartOil® G3 Edge-AI Brain™ can autonomously trigger an Exception-Sample™ to capture a forensic-grade, ASTM-ready oil sample that can reveal abnormal changes in viscosity per ASTM D445. A rapid shift in the oil viscosity could be representative of an inflection point resulting from a compromised fuel injector. Left unattended, this condition could bring upon the onset of fuel dilution; the SAE40W engine oil can quickly be reduced to SAE5W engine oil through dilution from diesel fuel contamination washing down from the affected cylinder.

Real-time oil viscosity trend data correlated with forensic-grade used oil analysis lab data (ASTM D445 - viscosity, ASTM D7593 - fuel dilution) can give operators a high degree of confidence to take corrective action, thus preventing a potential engine failure. Additionally, this corroborated data ensures maintenance crews have a better understanding of overhaul/repair timing which can essentially save valuable planning resources and shop time while increasing machine availability.

​

By trending the volume percentage of diesel fuel ingression into the crankcase oil and the decrease in real-time viscosity values, operators can now schedule maintenance in advance to ensure there is minimal disruption to machine availability and uptime. Maintenance can also guarantee replacement fuel injectors will be available to avoid further downtime of the equipment. Through Autonomous Fluid Intelligence™ operators can now manage their diesel engine fleets more efficiently.

​

Autonomous, human-free oil sampling provides a new level sampling consistency and safety in the locomotive industry. This changes the game entirely as real-time sensor measurements correlated with forensic-grade sample analysis are used to “train” the local language model (LoLM) for the entire fleet. Essentially, the fleet becomes smarter simply through continuous monitoring, autonomous sampling and anomaly detection and threat assessment compared against a fleet-based LoLM.  

 

Copyright ©2026 LogiLube, LLC