Generator Not Starting? A Step-by-Step Diagnosis for DG Sets

In short. Start by noticing exactly what happens when you press START. If nothing turns, the problem is almost always electrical — battery, connections, emergency stop or the controller. If the engine cranks but never fires, it is almost always fuel — empty tank, air in the lines, a blocked filter or a fuel solenoid that is not opening. If it starts and then stops, a protection is shutting it down. Those three paths cover most no-start calls we attend.

Who this page is for: For anyone whose generator will not start: decide in a few minutes whether it is a battery, controller, fuel or protection problem, what is safe to check yourself, and when to call a service engineer.

How a diesel generator starts

When you press START (or the AMF panel calls for a start), the controller does four things in order: it energises the fuel solenoid so fuel can reach the pump, it engages the starter motor, it watches engine speed through a speed sensor or the charging alternator, and it releases the starter once the engine is running on its own. A diesel needs three things to fire: fuel delivered at the right time, enough compression heat, and enough cranking speed to build that heat. Lose any one and the engine turns but does not run.

Diagnostic decision table

Safe checks you can make before calling

1. Read the panel. Note any alarm text or code exactly — photograph it.
2. Emergency stop and mode. Make sure the E-stop is released and the controller is in AUTO or MANUAL as intended. Reset any latched alarm.
3. Fuel. Check the tank level with a dipstick or gauge and confirm the fuel valve is open. Gauges on old sets often read wrong.
4. Battery. Check terminals are tight and clean and the isolator is on. If you have a meter, read voltage at rest and while someone cranks.
5. Oil and coolant. Check levels with the engine stopped. Low oil can trigger a shutdown immediately after start.
6. One attempt, then stop. If it still fails, stop. Repeated cranking drains the battery and can flood the cylinders.

Do not spray starting fluid into the air intake of an engine with a grid heater or glow plugs, bypass protections, or open a hot radiator cap.

Worked example (illustrative)

An illustrative case, not a project record. A 62.5 kVA standby set at an office fails to start during a power cut. The panel lights up and the starter clicks once. The battery reads 12.4 V at rest — looks fine. While cranking, the reading drops to 7.8 V. That collapse means the battery cannot deliver cranking current: it is sulphated or has a failed cell, even though its resting voltage looks healthy. Replacing the battery restores starting. The lesson: a resting voltage alone does not prove a battery is good; the voltage under cranking does.

When the usual assumptions fail

• “New battery, so it’s not the battery.” A new battery on a faulty charger, or with a corroded earth strap, behaves like a flat one.
• “The tank is full.” Fuel can be present but contaminated with water or sludge, or the pick-up pipe can be blocked.
• “It ran last week.” A set that ran on test but failed on a real outage may have an AMF wiring, sensing-fuse or mode-switch fault rather than an engine fault.
• Electronic engines. On electronically controlled engines a sensor or ECU fault can prevent starting; these need diagnostic tools rather than mechanical checks.

Preventing no-starts

• Weekly test run, with the result logged
• Battery charger checked and battery tested under load at every service
• Fuel kept topped up; water drained from separators
• Controller left in AUTO after every test
• Fuel filters changed, and the system bled, by someone trained to do it

A service engineer’s view

Most no-starts are not engine failures. They are batteries, fuel and switches — the small things that are skipped because the set runs so rarely. A weekly run with a written log prevents more breakdowns than any repair.

Cranking speed, compression heat and the starting circuit

A diesel has no spark. It fires only because air squeezed to a fraction of its volume gets hot enough to ignite the injected fuel. That heat depends on cranking speed. At a healthy cranking speed the air is compressed quickly, little heat escapes into the cold metal and little air leaks past the rings, so the charge comfortably passes the temperature at which diesel ignites. Halve the cranking speed and each compression stroke takes twice as long: more heat soaks into the liner and head, more air leaks away, and the peak temperature may fall just short of ignition. That is why an engine that “turns over” slowly can crank for a long time without a single firing stroke, even with perfect fuel.

Cranking speed in turn depends on the voltage that actually reaches the starter motor, not the voltage at the battery. A starter can draw several hundred amperes for a few seconds. At that current, every joint in the circuit — battery post, clamp, cable lug, isolator, starter solenoid, earth strap to the engine block — behaves like a small resistor. A few thousandths of an ohm that are invisible on a multimeter at rest become volts lost under cranking. The earth return is half of the circuit and is the half most often forgotten.

Electronic engines: the ECU must stay awake and in sync

On electronically controlled engines, including most newer CPCB IV+ sets, the mechanical picture is only part of the story. Before the engine control unit (ECU) will inject any fuel, three conditions must be met during cranking:

Each of these leaves a trace as a diagnostic trouble code. That is why the first step on an electronic set is to read the active and stored codes with the correct service tool, not to start loosening fuel unions. Some sets with emission after-treatment can also restrict power or operation after repeated emission-system warnings, depending on how they are configured; the codes show whether that is in play.

Where air gets into the fuel system

Most “cranks but won’t fire” calls on mechanical engines are air in the fuel. Air enters on the suction side, between the tank and the lift pump, where the line is under slight vacuum and a leak draws air in rather than letting fuel out — so there is often no visible drip. Typical entry points are a filter bowl or element seal not seated after a change, a hairline crack in a plastic pipe, a loose banjo union, a perished hose between tank and filter, and a tank level so low that the pick-up pipe draws air when the set tilts or vibrates. On sets that stand for weeks, fuel can drain back to the tank overnight through a leaking non-return arrangement, so the first start after a long gap fails while a second start “mysteriously” works.

In Indian conditions two seasonal effects make this worse. Through the monsoon, a part-filled tank breathes humid air in and out every day, and water condenses inside; it settles at the bottom, reaches the separator and, on a high-pressure system, can damage the injection parts. Through the hottest months, filters and hoses on sets in un-ventilated rooms age faster, and rubber seals harden.

No-start record sheet for site staff

Fill this in before the second attempt and before calling. None of it requires touching wiring or opening the engine.

After a failed start: try again, stop, or escalate

Where batteries, chargers or starting cables are the recurring issue, our battery and starting-system service page explains what a full starting-circuit check covers.

Second worked example: the battery is good, the cable is not

Illustrative case. An 82.5 kVA set with a 12 V starting system at a cold-storage unit cranks slowly and fails to start. A new battery was fitted two months earlier, and the panel shows 12.7 V at rest. The service engineer measures voltage in two places while the set is cranked:

• Across the battery posts: 10.6 V
• Between the starter’s main terminal and the engine block: 8.1 V

The battery is holding up well — 10.6 V is comfortably above the collapse seen in a failing battery. But 10.6 − 8.1 = 2.5 V is being lost in the cables and joints before it reaches the starter. A healthy circuit usually loses under 1 V.

Taking an illustrative cranking current of 500 A:

Circuit resistance = 2.5 V ÷ 500 A = 0.005 Ω (5 milliohms)
Heat in the joints = 2.5 V × 500 A = 1,250 W

Five milliohms is invisible to an ordinary meter, yet under cranking the joints dissipate heat like a room heater. Measuring the drop on each half of the circuit separately shows that most of it is on the earth side: a braided earth strap between frame and block has corroded under its lug in the damp of the previous monsoon. Remaking that joint brings starter voltage up to about 10 V, and the set fires on the first attempt. Replacing the battery again would have changed nothing.

Voltage-drop testing is carried out by a trained engineer, with the engine guarded against unexpected starting and hands clear of moving parts. Site staff should not probe starter or engine wiring.

Common mistakes

• Cranking again and again until the battery is flat, which turns a fuel or switch fault into a two-fault breakdown.
• Replacing the battery or starter first, without measuring battery voltage while cranking.
• Overlooking a pressed emergency stop, a closed fuel valve or a panel left in Off or Manual.
• Bleeding the fuel system without first finding how air got in, so the fault returns at the next outage.

What to send us

Send the set’s kVA and approximate age, the exact message on the controller, what happens when you press start (silence, a click, cranking without firing, firing then stopping), the battery voltage if you can read it, and when the set last ran and was last serviced. Use the service and AMC enquiry, or WhatsApp the details and photographs.

Related pages

• Generator Overheating and High Coolant Temperature Shutdowns
• Low Oil Pressure Shutdown on a Generator
• Generator Hunting
• Generator Troubleshooting
• Breakdown Diagnostics
• Generator Controller Alarms Explained for Site Staff

Technical content reviewed by D. Srinivas Reddy, Managing Partner and service engineer. Last updated 3 November 2026.