Coolant Loss With Negative Cold Pressure Test
2013 class 8 tractor, 14.9 L, 612,000 miles. Driver reports the coolant level light on every few days; the shop tops off roughly 1 gallon per 500 miles. There is no visible leak, no steam, and no white smoke at idle. The reservoir cap vents when pulling a long 6% grade, where coolant temperature climbs to 222°F before dropping on the descent. A cold cooling-system pressure test at 15 psi held for 30 minutes with no drop. A chemical block-test (combustion gas) at idle was negative. Oil analysis: sodium 14 ppm, potassium 8 ppm, iron 22 ppm - all within fleet limits. The engine has an EGR cooler, and the fleet wants the truck back in service today.
1. Both the cold pressure test and the idle block test came back negative. Explain why neither test rules out an internal coolant leak on this engine, and describe the next tests you would run with ordinary shop equipment.
Read the worked answer
A cold static test only proves the system is sealed at room temperature and 15 psi. This leak is load- and temperature-dependent: the head, liner seals, and EGR cooler expand and distort at 220°F, and cylinder pressure at full load is many times the 15 psi test pressure. Likewise, a block test at idle produces little combustion pressure and low coolant flow, so a small leak is diluted below the detector's threshold.
Next steps:
- Hot pressure test: bring the engine to operating temperature, pressurize to cap rating, and watch the reservoir for a steady bubble stream (not just expansion burping).
- Block test at high idle and, safely, under load on a chassis dyno or hard pull.
- Cold-start bubble check: pressurize, start cold, and watch for bubbles within the first 30 seconds before the thermostat opens - that points to a compression leak.
- Borescope a cylinder after an overnight sit to look for coolant pooling.
2. How would you separate an EGR cooler leak from a head gasket, liner seal, or injector-cup leak on this engine?
Read the worked answer
Isolate the EGR cooler's coolant circuit - loop or plug its coolant hoses per the OEM procedure - and repeat the hot pressure test. If the loss stops, the cooler is the leaker; confirm on the bench by capping the gas side, pressurizing the coolant side, and looking for moisture or pressure decay. Also inspect the EGR valve and intake for coolant residue or white deposits.
If the loss continues with the cooler isolated, work the combustion side:
- Per-cylinder leak-down with the cooling system pressurized and the reservoir watched for bubbles identifies the affected cylinder.
- A pressure spike in the reservoir within seconds of a cold start indicates a compression-to-coolant path (head gasket or liner).
- Coolant in a cylinder on borescope, or a clean steam-washed piston top, supports an injector cup or head gasket.
- Sodium at 14 ppm is a mild but real coolant marker; trend it rather than dismissing it.
3. The fleet dispatches the truck anyway unless you object in writing. What is your recommendation, and what would you log during a supervised road test?
Read the worked answer
Recommend against dispatch. An internal leak that only opens hot can progress to a hydraulic lock on shutdown, bending a connecting rod or cracking a piston - a coolant loss of 1 gallon per 500 miles is already a real leak, and the cause is unidentified. Document the recommendation and get the fleet's acknowledgment.
If a supervised road test is the compromise, log:
- Coolant temperature and reservoir level before and after, plus any cap venting.
- EGR cooler gas-outlet temperature compared with the inlet - an unexpected rise or drop can indicate cooler bypass/leak.
- Boost, derate status, and any white smoke at the stack under load.
- Stop the test at 210°F or on any level drop and re-test hot, not cold.