Phil Golin
"It is amazingly obvious that they did/do not want to mention the actual water content of the fuel on the flight. It SEEMS to be mention only three times and almost only in passing at "35 - 40" ppm, HOWEVER, all the tests were run with concentrations of 90 ppm (or more). There is no examination in detail of the actual measured water levels or what the actual state of the fuel/water mixture might have been like at the landing stage."
1) 35-40ppm is what was left in the fuel after landing - in other words that does not include water which may have settled and been recirculated by the water scavenging pumps.
2) The estimates for water concentration at loading are (admitedly up to) 40ppm dissolved and 30ppm entrained - or up to 70ppm. To get from 70-ppm at the start to 40ppm at the end, the concentration of water in the fuel leaving the tanks needs to be higher than 70ppm. If you read the bit on water scavenge system, you will see that the low lying fuel (more likely to have higher concentrations of water) is re-injected near to the pump inlets - why, because you are trying to get rid of the water or it will just settle back to the bottom - the way out is through the engine.
3) "A target water concentration of 90 ppm (as defined in ARP 1401) was selected for all the tests." SAE-ARP 1401 being the international standards publication for "Aircraft Fuel System and Component Icing Test".
Its not a glaring gap - its not precise, but then again as they have stated, very little about reproducing this is precise, their testing was to prove the viability of their hypothosis, because it was basically a new hypthosis.