Nick: You are correct - the ability of esters to effectively transfer heat is somewhat less than traditional mineral oils. (Mineral oil = 125.7, natural ester = 111.8, and synthetic ester 110.7 - all units are W/m2/K).
The net result is that for a given volume of liquid and a given thermal "starting point", the traditional mineral oil will keep the windings and their insulating materials about 1.1 to 1.2 degrees C cooler than either natural or synthetic esters. However, esters will, in general, withstand higher temperatures before their insulating properties are degraded.
On the ester "plus" side is improved biodegradability of the liquid, lower toxicity, high flash point, and low caloric value. On the mineral oil "plus" side is lower cost-per-unit-volume and higher regional availability.
It comes down to this - the transformer vendor is more interested in volume sales than in providing a durable product (planned obsolescence? limited life?). Since the ester coolant can be worked "harder" (e.g. it doesn't catch fire as easily and, due to biodegradation and reduced toxicity, is a favorite of "green" thinkers) and the train of thought that "a winding can always be replaced", transformer vendors - or at least their coolant suppliers - tend to argue for ester. Does it really reduce the transformer physical size? Sometimes - but then the question becomes whether it is truly the change to esters or a change in the transformer design (i.e. cooling circuit) that does the trick.
Converting energy to motion for more than half a century
