Knowledge / Ship systems

Scavenge-air cooling: density, condensation and water carry-over

Connect compressed-air temperature, pressure dew point and condensate mass, then distinguish normal moisture removal from leakage, blocked drains and liquid carry-over.

On this page

Cooling the air after a turbocharger improves the charge supplied to an engine, but it can also condense water already present as vapour in the intake air. The cooler, separator and drain therefore form a linked system. A low outlet temperature is not sufficient evidence of correct performance if liquid is accumulating or being carried downstream. The useful question is how much heat is removed and how much moisture changes phase, and whether the installation handles both within its intended operating envelope.

Separate cooling, condensation and liquid removal

In a representative arrangement, the compressor supplies hot pressurized air, the cooler removes heat, and a mist catcher limits liquid carry-over. The Everllence G95ME-C10.7-GI project guide identifies these elements in its scavenge-air system. This is a specific published example, not a universal layout. Other engine arrangements may have different stages, coolant circuits or drain connections, so equipment identity remains part of the interpretation.

Three physical states should be kept separate. Water vapour is a gas constituent; cooling can turn part of it into liquid droplets; separating and draining those droplets is a further process. A mist catcher is not a complete vapour dryer. Air leaving it can remain close to saturation even when liquid separation is effective. If that air is cooled again farther downstream, more water may condense.

Use absolute pressure when reasoning about humidity

Relative humidity compares water-vapour partial pressure with the saturation pressure at the local temperature. Compression raises partial pressures if composition remains unchanged. Vaisala’s dew-point explanation emphasizes that a sample measured after expansion to atmospheric pressure has a different dew point from the pressurized process. A quoted dew-point temperature therefore needs its pressure basis; an atmospheric sample value cannot simply be applied inside a pressurized receiver.

For a preliminary ideal-mixture calculation, use humidity ratio w, in kilograms of water vapour per kilogram of dry air. The relation is w ≈ 0.622 pᵥ/(p − pᵥ), where both total absolute pressure p and vapour partial pressure pᵥ use the same pressure units. The constant is the approximate molecular-mass ratio. Gauge pressure in the denominator creates an error because the gas mixture responds to absolute pressure. Also distinguish dry-air mass from total moist-air mass when forming a water balance.

Set a transparent example with rounded property values

Assume intake air at 35°C, 70% relative humidity and 101.3 kPa absolute pressure. Use saturation pressure 5.63 kPa at 35°C and 7.38 kPa at 40°C, rounded from NIST’s water saturation table. The inlet vapour partial pressure is 0.70 × 5.63 = 3.941 kPa. Its humidity ratio is approximately 0.02518 kg/kg of dry air, or 25.18 g/kg. These are illustrative ambient conditions, not tropical design requirements for a named engine.

Now compress this air to 300 kPa absolute and suppose no moisture is removed before the cooler. The water-to-dry-air mass ratio stays the same, although temperature and partial pressure change. The calculated vapour partial pressure before condensation is 300 × 0.02518/(0.622 + 0.02518), approximately 11.67 kPa. Cooling to 40°C brings the gas below the corresponding pressure dew point, so some vapour must condense in this equilibrium model.

Calculate condensate without confusing the flow basis

At the assumed saturated cooler outlet of 40°C and 300 kPa absolute, the humidity ratio is 0.622 × 7.38/(300 − 7.38) = 0.01569 kg/kg of dry air. The Vaisala calculation reference gives the mixing-ratio relationship in grams per kilogram; this example uses the ideal-gas form in kilograms per kilogram. It neglects non-ideal enhancement and detailed transfer behaviour. The difference between inlet and outlet ratios is approximately 0.00949 kg of condensed water per kilogram of dry air.

For an invented dry-air flow of 10.0 kg/s, the equilibrium condensate rate is therefore about 0.0949 kg/s, or 342 kg/h. This is the rate of liquid formation predicted by the model, not a guaranteed measured drain rate or a drain-sizing recommendation. If 10.0 kg/s had instead meant total moist-air flow, the dry-air flow would be smaller and the calculation would need adjustment. A unit label alone is not enough; the mass basis must be named.

Know why the ideal result may differ from the drain reading

The example assumes a uniform saturated outlet, steady conditions, no upstream water removal, no external leakage and complete collection of the condensed liquid. A real cooler may have temperature maldistribution, finite residence time and droplets that do not follow the gas evenly. Some liquid can be temporarily stored on surfaces or in a drain chamber. Measurement delay and transient load changes can separate the timing of condensation from the timing of discharge.

A drain rate lower than the calculated value can have several explanations: actual humidity may be lower, collection may be incomplete, the drain may be obstructed or water may be carried onward. A higher rate can reflect changing ambient conditions, release of stored liquid or a cooling-water leak. Treat the mass balance as a way to ask better questions, not as a single pass/fail threshold. Compare an interval long enough to account for storage and timing.

Examine the drain as part of the pressure boundary

The published G95ME-C10.7 drain description includes a permanent mist-catcher drain, an orifice limiting loss of scavenge air and a level indication for excessive accumulation. This bounded manufacturer example illustrates that liquid removal must coexist with a pressurized air path. It does not authorize altering an orifice, bypassing a level device or copying another engine’s drain arrangement.

A drain can be open as intended yet ineffective because the downstream pressure or routing prevents the required liquid flow. Conversely, an unintended large opening can release air and disturb the system. Trace both the liquid destination and the gas-pressure connection. A useful diagnosis asks what pressure drives the drain, what can oppose it, where water can collect and which observation proves that liquid is actually leaving.

Interpret cooling performance and pressure loss together

Air temperature, coolant temperatures, air-side pressure difference and relevant flow or load must be read together. A hotter outlet can indicate reduced heat transfer, warmer coolant, changed airflow or a different compressor inlet state. A larger air-side pressure drop can arise from higher flow as well as increased restriction. Comparing only one temperature or one differential-pressure value at unmatched load can therefore misidentify fouling.

Cooling increases gas density at a given pressure, but humidity and condensed mass also affect a precise composition balance. Lower temperature is not an unlimited optimization target. The engine manufacturer’s allowable charge conditions, condensation management and corrosion considerations constrain the useful operating range. A cooler that achieves a low average temperature while allowing liquid carry-over has not demonstrated the required combined function.

Distinguish leakage from ordinary atmospheric moisture

Visible water downstream does not automatically establish a leaking tube. Intake humidity can generate substantial condensate, as the example shows. Conversely, calling every drain flow atmospheric condensate can hide a real coolant leak. Compare changes in water chemistry where appropriate, coolant inventory, pressure relationships, ambient humidity and operating history using the approved diagnostic method. No single colour, conductivity value or short observation should replace the complete evidence chain.

Consider an event that begins after maintenance. Plausible questions include whether drain paths were restored, whether separator elements are installed correctly and whether sensor connections represent the intended location. Those questions concern functions and evidence, not improvised work on a running pressurized system. Inspection, isolation and testing remain tied to the actual equipment procedures.

Make a useful operating comparison

A comparison record should state intake temperature and humidity, intake and receiver absolute pressures, the location of each temperature sensor, coolant state, engine load, dry- or moist-air flow basis and the drain measurement interval. Note whether conditions were steady or changing. If an atmospheric humidity sample is used, retain its sampling pressure and any correction to process conditions.

The resulting interpretation should distinguish heat removal, vapour condensation, droplet separation and drainage. Keeping those four functions visible prevents two common errors: using cold air as proof that water removal works, and using water at a drain as proof of a coolant leak. The worked balance explains scale and causality; an actual acceptance decision still requires the installation’s specified limits and verified measurements.

Sources