A road construction contractor reported that a Kobelco SK330LC-8 excavator's hydraulic oil sight glass showed a "chocolate milk" color, and simultaneously, the engine coolant temperature was creeping into the red zone after only one hour of trenching work.
The technician's first hypothesis was condensation from short daily operating cycles, but the machine worked full shifts, so he ruled that out. Milky oil combined with overheating pointed strongly to a cross-contamination leak between the hydraulic circuit and the engine cooling circuit. On this model, the only location where the two systems meet is the "Oil-to-Water Plate Cooler," mounted on the radiator assembly, which uses engine coolant to stabilize hydraulic oil temperature.
He performed the definitive test: he disconnected the hydraulic inlet and outlet lines from the cooler, capped them, and then pressurized the cooling system with a "Radiator Pressure Tester" to 0.9 bar with the machine off. Within five minutes, the pressure bled down to 0.4 bar, and coolant was visibly weeping from the hydraulic side outlet of the cooler. The internal core had cracked. The cracks had most likely been initiated by an earlier event: the site foreman admitted the machine had once been refilled with straight water instead of coolant mixture during an emergency, and the previous winter's freeze had expanded the ice inside the thin cooler plates.
The failure mechanism explained both symptoms in sequence. Under pressure, the higher-pressure hydraulic oil (approximately 20-35 bar in the cooler circuit) forced oil into the lower-pressure coolant side whenever the machine worked, producing the oily film in the radiator and raising coolant temperature by reducing heat transfer efficiency. Then, at engine shutdown, when hydraulic pressure decayed faster than residual cooling system pressure, the flow reversed: coolant seeped into the cooler's oil passages and drained into the hydraulic tank, emulsifying the oil into the chocolate-milk appearance.
The repair was extensive because the contamination ran both ways. The oil-to-water cooler was replaced with a new unit. The entire hydraulic system was flushed twice: first, the tank was drained, cleaned with lint-free wipes, and refilled with flushing oil circulated for two hours through a separate external filtration cart fitted with 3-micron elements; then fresh ISO 46 oil was installed. All five main pump pistons and all swing/travel motor cases were inspected for bearing discoloration-fortunately caught early enough. The radiator was removed, professionally boiled out at a radiator shop to remove the oil film, and refilled with proper 50/50 long-life coolant. After three weeks of monitoring, oil samples came back clean, and coolant temperature stabilized at 88°C even in heavy digging.
