Craig and Eric,
What I call the lip is spring side of the seal. In other words, in my experimental configuration, the orientation of inner (next to the C-Drive) and middle seals was with the spring facing the propeller side (preventing water in) and the outer seal facing the C-Drive (preventing the oil from leaving).
I think I can clarify a little more my rationale in the sentences below
Before proceeding it is important to note that standard seals work to prevent motion of oil from the spring side, lip side, but not from the opposite side. As pressure builds on the spring side it creates a force that tends to close the gap between seal and shaft. In contrast, when the pressure builds on the opposite side of the seal, it tends to expand the lip outwards opening a gap and allowing motion across the seal. In standard applications, the seal lip faces the high-pressure oil side.
As Danny pointed out, the density of oil is only 90-93 % that of water oil would tend to move up relative to water if both liquids present in the same container. However, what drives the direction of motion of fluid across the seals is the net pressure difference across them. That difference is proportional to the difference of the products of the height of the column times the density of the fluid. So, because the oil reservoir is above the water level by more than 10% of the distance between the water level and oil level in the tank, under normal conditions the net pressure across the seals would tend to move oil out of the C-Drive in the absence of any seals. Note, however, that if there were no seals preventing such motion, oil would flow out until its height is just about 10% higher than that of the water and, at that point, the flow would stop (pressure difference becomes zero!. So even if the seal (or seals) preventing the oil from leaving were to fail catastrophically, the oil in the C-Drive would not empty completely.
So why do we get water into the C-Drive when using the "Amel way" of orienting the seals? Well, the above calculation is valid only on static conditions. During dynamic conditions, when the outer seal fails, the interface between oil and water is put in motion driven by the shaft, which tends to mix the two fluids forming the cream chocolate color emulsion. If the central and inner seals (spring side) are facing the C-Drive as recommended by AMEL, this emulsion can move easily across them into the C-Drive and is mixed with the rest of the oil by the recirculation action of the C-Drive inner pump. Over time the water entering the C-Drive makes the oil level on the reservoir increase.
Then, why did the seals and bushing lasted longer and water did not get into the C-Drive in my experimental configuration? First, the friction between the bushing and the seal surface, over time tends to wear both surfaces. When newly installed proper greasing mitigates this friction and reduces wear but as the grease deteriorates over time, friction increases and the seal wears out. A harder bushing would last longer but the seal wear is the same. In either case, the presence of a thin layer of fluid (oil or water) between seal and bushing acts to lubricate the surface and reduce friction. In my experimental configuration, the seals next to the C-Drive Oil allow oil motion towards the outer seal when the latter begin to lose sealing. This makes them lubricated by the oil moving out, thus prolonging their life. If at this point, the C-Drive oil is made more viscous, the additional viscosity of the oil elevates the inner pressure which tends to close the outer seal reducing its flow but still able to lubricate it. In the "Amel way" the two inner seals lips are facing the oil side and, the outer seal facing the propeller thus none of the seals tend to be lubricated once the grease ceases to work.
A long explanation for a simple question but I hope worth to understand why in this case the "AMEL way" is not the ideal one.
Jose Venegas (Not Jose Luis)
Ipanema SM2k 278