Head gasket failure and oil consumption: telling them apart

An engine that is burning oil and an engine with a failed head gasket are the same kind of fault: something is crossing a boundary that is supposed to be sealed. They are told apart not by how bad it is but by which boundary, and by which direction the leak runs.

There are four boundaries and each one leaks somewhere you can observe from outside the engine:

• Oil into the combustion chamber — blue smoke
• Coolant into the combustion chamber — white smoke, and coolant disappearing
• Combustion gas into the coolant — overheating and a pressurised cooling system
• Combustion gas into the crankcase — pressure in the engine, and oil pushed out of it

Work out which one is happening and you have done most of the diagnosis. Everything below is how to tell.

A note on the original, because it matters to what follows. Its title promises the head gasket and the causes of oil consumption. Its text covers neither: it is a careful parts list of the cylinder head — valves, seats, guides, stem seals, camshaft — and it stops there. Both promised subjects were in the embedded video only. So the parts of this article that come from it are the valve-train sections, which are good, and the rest is what a reader who could not play the video never got.

Reading the smoke

Colour tells you which fluid. Timing tells you which part.

The first row is the original's best single observation and it is worth stating as clearly as it does: an engine that is slow to start after standing for a few hours and then blows blue smoke at the first press of the throttle has failed valve stem seals. Oil sitting in the head runs down the valve guides into the chamber while the car is parked, and the engine burns it all off in the first few seconds. It is a common fault, it is not the same as worn rings, and the repair is a fraction of the cost.

Head gasket failure is four different faults

This is the part the original never reaches, and lumping them together is why the phrase "head gasket" frightens people more than it should. The gasket separates several things at once, and which pair it stops separating decides the symptom entirely.

Combustion chamber to coolant. The commonest and the most misleading. Combustion pressure — many times higher than anything the cooling system sees — is forced into the water jacket. The engine overheats, the top hose goes hard almost immediately after a cold start, coolant is pushed out of the expansion tank, and the level falls with nothing on the driveway, because the loss is going out of the exhaust as steam. Gas does not carry heat and does not pump, so the overheating and the coolant loss are the same event.

Coolant to combustion chamber. The same breach, the other way round, when the engine is off and the system is pressurised while the cylinder is not. Coolant seeps into the cylinder overnight, and the engine is hard to start, blows white smoke for a minute, and may hydraulic-lock in the worst case. A spark plug that comes out steam-cleaned while its neighbours are sooty is pointing at its own cylinder.

Combustion to oil. Gas and coolant into the crankcase. Coolant and oil emulsify into the pale sludge everybody calls mayonnaise, on the filler cap and the dipstick, and the oil level rises. This one destroys bearings quickly, because emulsified oil does not carry a load.

Between two adjacent cylinders. No smoke, no coolant loss, no external sign at all — just a misfire on two cylinders that survives every ignition and fuel part anyone fits. This is the one a compression test finds and nothing else does, and it is exactly the two-adjacent-cylinders pattern described in compression testing.

And to the outside. A weep of oil or coolant down the side of the block at the head joint. The least serious and the easiest to see.

The tests, cheapest first

Look at the oil and the coolant. Pale emulsion under the filler cap means coolant in the oil — but note the false positive, because it is common: a car doing only short cold journeys condenses water in the rocker cover and produces a small amount of the same thing without any gasket fault. Judge it on the dipstick, not the cap. An oil film or brown scum in the expansion tank runs the other way and is more reliable.

Watch the expansion tank with the engine running from cold. Continuous bubbling that starts before the thermostat opens is combustion gas, not air. So is a top hose that becomes hard within a minute of a cold start.

Do the combustion leak test. A chemical block tester draws air from above the coolant through a reagent that changes colour in the presence of CO₂. It is cheap, it takes five minutes, and it is definitive. Why an engine overheats covers it, along with pressure-testing the system cold and leaving it, which finds the leaks that only exist when the engine is not running.

Then compression, and then leakdown. Compression finds the two-adjacent-cylinders signature and tells you a cylinder is not sealed. Leakdown tells you where the air goes — bubbles in the coolant for a gasket, hiss at the exhaust for a valve, hiss at the dipstick for rings. Both are in compression testing.

Oil consumption that is not a fault

Before diagnosing an engine, establish that it is actually using an abnormal amount, because the modern answer surprises people. Manufacturers commonly permit up to about a litre per 1,000 km, and several state that figure outright in the owner's handbook — turbocharged and performance engines especially, and some will not consider a warranty claim below roughly 1.5 litres per 1,000 km. Low-tension rings and long oil-change intervals both push consumption up by design.

So measure it before you chase it: note the level on a level surface, on the same dipstick, at the same temperature, and record the kilometres between top-ups. An engine using half a litre between services is almost certainly fine. An engine using half a litre a week is not.

The cheap cause that mimics the expensive one

The second piece folded into this article is about the crankcase breather, and it deserves to be the practical conclusion, because it is the one thing here that turns a rebuild into an afternoon.

The crankcase has to breathe. Some combustion gas always escapes past the rings, and if it cannot get out, pressure builds in the engine and pushes oil out of every seal it can reach — past the stem seals into the chamber, past the crank seals onto the driveway, and straight up the breather hose into the intake where it is burnt.

A blocked or stuck crankcase ventilation valve therefore produces exactly the symptoms of worn rings: oil consumption, blue smoke, oil misting the intake tract, and often a rough idle and pushed-out seals as well. The difference is that the PCV valve is a cheap part and the rings are an engine.

So the order is: check the breather system before condemning anything. Clear hoses, a valve that rattles and is not gummed shut, no oil standing in the intake pipework. If the crankcase still builds pressure after the system is confirmed clear, then the pressure is coming past the rings and the compression and leakdown tests above are the next step — and at that point the answer really is worn piston rings.

What the original gets right, and one thing to correct

Its valve-train material is sound and worth carrying:

• A burnt valve usually starts with clearance. The original names incorrect valve clearance and a wrong air-fuel ratio as the two causes, and it is right on both counts — valve clearance explains why the drift runs toward the dangerous side.
• A snapped cam belt bends valves on an interference engine, which is most of them.
• A seat can drop out of an overheated head, which is one more reason to treat an overheating event as a real event.

One claim needs correcting. It says a worn valve guide produces "no change in engine operation, just a drier, noisier head". The noise is real. But a worn guide is one of the main routes by which oil reaches the combustion chamber — the stem seal sits on the guide and cannot seal a stem that is no longer held straight — and a valve that is not held square to its seat does not seal properly and eventually burns. A worn guide is an oil-consumption fault and a future burnt valve, not just a noise, and the article's own list of stem-seal symptoms a few paragraphs earlier describes what it does.

The corpus records the underlying condition at low engine compression and burnt or leaking valve, and the misfire it produces at P0301 onward — which is where most people meet this problem before they ever see smoke.

Video guides

Video, for the parts of this that are easier watched than read. The English-language ones come first.