Size, noise and independence from ambient air. Water carries far more heat per unit volume than air, so the cooler is much smaller and quieter, and it holds oil temperature on a hot day when an air-blast unit would be struggling. The condition is having a water supply.

Water availability. Marine installations have effectively unlimited seawater and use these almost universally; industrial plant with a cooling water circuit likewise. Mobile equipment has no water supply and is therefore always air-blast, whatever the other advantages might be. It is the first question to settle, before any sizing is done.

Fouling on the water side - scale, silt and biological growth - which reduces heat transfer steadily and invisibly. The symptom is oil temperature creeping up over months with nothing obviously wrong, and the cure is cleaning the water side rather than investigating the hydraulics.

Tube and tube-sheet materials specified for seawater, since a cooler built for fresh water will corrode rapidly. This is a specification point rather than an upgrade, and getting it wrong produces a failure within months rather than years on an otherwise correct installation.

Water enters the oil, which is worse than the overheating the cooler was fitted to prevent - it degrades additives, corrodes components and destroys lubricating films. Rising water content in oil analysis on a water-cooled system should be treated as a cooler leak until proven otherwise.

By regulating the cooling water flow, usually with a thermostatic valve on the water side, so the oil is held in its working band rather than simply cooled as much as possible. Overcooling raises viscosity and is not harmless, particularly on start-up.

Periodic cleaning of the water side to remove fouling, inspection of tubes and tube sheets for corrosion, and monitoring of both oil temperature and oil water content. The temperature trend is the practical early warning, which is why a temperature sensor pays for itself on this type of installation.