Coolant Loss With No External Leak on a Tier 4F Loader
A 5,400-hour, six-cylinder turbocharged diesel in a wheel loader (cooled EGR, DOC) loses about 4 L of coolant per 10-hour shift but shows no external leak, no steam and no coolant on the dipstick. Coolant is a fresh-looking 50/50 mix. Under load at 34 °C ambient the top tank reaches 105 °C and the ECM has derated once. The radiator core is roughly 30% plugged with dust externally. The pressure cap releases at 88 kPa against a 100 kPa specification. Infrared readings at operating temperature show 96 °C at the top tank and 88 °C at the bottom tank. A cold pressure test pumped to 100 kPa falls to 55 kPa in 15 minutes with the system capped and dry. A block tester held blue at idle. Oil analysis returns potassium and sodium slightly above trend with silicon normal, and a cold start produces white vapour from the exhaust that clears within a minute.
1. What do the pressure decay and the negative block test actually prove, and what is your next test sequence?
Read the worked answer
The cold pressure decay (100 → 55 kPa in 15 min, dry and capped) proves an internal leak. A sound system holds pressure with only minor decay; a 45% loss with no external trace means coolant is going where it should not - combustion space, intake, oil or exhaust.
The block tester staying blue at idle proves very little. It is a screening test: at idle there is little cylinder pressure and little EGR flow, and a crack or leaking EGR cooler core often only opens when hot and loaded. A negative at idle does not rule out combustion gas. Repeat it at operating temperature, at high idle or under load, with the coolant level correct and the sample drawn from the top tank neck.
The 105 °C has simpler contributors that do not explain the loss. An 88 kPa cap against a 100 kPa spec lowers the boiling point of the 50/50 mix, and a core 30% plugged with dust loses heat rejection. Fix both - but neither removes 4 L per shift.
Next tests:
- Pressurize with UV dye at 100 kPa, run under load, inspect with a black light - radiator, water pump weep hole, EGR cooler, oil cooler, aftertreatment.
- Isolate circuits one at a time: plug the EGR cooler coolant ports and pressure-test the rest, then bench-test the cooler hot.
- Pressure-test the oil cooler separately and compare potassium/sodium against the last three samples, not one result.
- Look for glycol residue in the exhaust and at the DOC/DPF face.
- Finish with a hot pressure test at 100 kPa at operating temperature.
2. May the machine keep working while parts are on order? Justify your answer.
Read the worked answer
No - and the reasoning matters more than the rule.
- The leak path is unknown and one likely path is destructive. If coolant is passing through the EGR cooler into the intake, it enters the cylinder as liquid. On shutdown or a hydro-lock event, incompressible coolant can bend a connecting rod or lift a head - a cooler replacement becomes an engine overhaul.
- The aftertreatment is at risk. Glycol and coolant ash poison the DOC and plug the DPF. The derate may be the ECM protecting the engine, but the exhaust hardware is not protected from chemical contamination.
- Oil contamination compounds it. Potassium and sodium trending up means coolant is reaching the oil; coolant in oil attacks bearings and emulsifies into sludge. Every hour of running accelerates wear that fixing the leak will not repair.
- The overheat contributors are still present. 105 °C with a 30% plugged core and an 88 kPa cap means the machine is near the boil even when the leak is slow; a hard pull on a hot day pushes it into derate, and a derated loader in a production cycle creates pressure on the operator to keep pushing.
If the machine is essential: allow only a defined light-duty task with a daily coolant check, a hard shutdown temperature and written customer acceptance - and only if the leak is confirmed external. With an internal leak of unknown path, it parks.
3. How do you isolate the three likely internal leak paths, and what must be checked before and after the repair?
Read the worked answer
The candidates fail in different places, so the tests must distinguish them.
- EGR cooler: coolant lost to intake or exhaust, often only when hot; white vapour; coolant residue in the intake; the cooler bench-tests leaking when heated even if it passes cold. This fits a decay test with no external trace and only slight oil contamination.
- Oil cooler: coolant in the oil (rising potassium/sodium, milky oil, rising oil level) with no exhaust vapour. Pressure-test it isolated.
- Head gasket, liner or head crack: combustion gas in the coolant (block tester positive hot and loaded), coolant pushed out of the top tank on shutdown, bubbles at the radiator neck, a hard top hose. The cold block test does not exclude this.
Before repair: drain and capture the coolant as evidence, confirm the circuit by isolation, and inspect the intake and aftertreatment for coolant residue.
After repair - shut-down checks before restart:
- Hydro-lock check: if coolant may have entered a cylinder, bar the engine over by hand through two revolutions with the injectors out on the affected cylinders before touching the starter.
- Aftertreatment: inspect the DOC/DPF face for glycol residue or ash; a contaminated DOC can derate the machine after an otherwise perfect repair.
- System: new 100 kPa cap, clean the core from the fan side, vacuum-fill to avoid air locks, use the specified coolant chemistry (do not top up an OAT system with conventional green coolant), run to temperature with the heater open, recheck level after cool-down.
- Verification: hot and cold pressure test at 100 kPa with no decay; a loaded run at ambient with a stable top tank and no derate; oil sample at 50 h to confirm potassium and sodium returning to trend.