Maintenance: Fuel boost pump leak.

The first flight following my second annual was a ~200 NM roundttrip to have a "$100 salad" with a friend.  The plane was returned to the hangar following an issue-free flight.  Upon going to the hangar four days later, when I opened my canopy, I was greeted with a strong odor of avgas.  Clearly, there would be no flying that day.

Knowing that the only part of the fuel system that was touched during the annual was around the filter (to remove, inspect and replace the latter), I immediately removed the F-01452 Aft Tunnel Cover to inspect the area.  Fuel stains were located around the airframe's lower F-14108A Fuel Pump Bracket yet not the upper F-14108B Fuel Pump Bracket (see below diagram from Figure 31-11, brackets in red). 

Here is an image following removal of the pump and filter.  You can see the fuel staining on the lower Fuel Pump Bracket.

There was no staining near or on the boost pump's output side (thus the output fuel line remained sealed to the pump).  Additionally, no fuel staining was noted at the joint between the filter and pump or anywhere on the filter body, confirming that the leak was not due to the filter as that seal remained intact too.  Fuel stains were concentrated at the joint between the pump's segments at the lower-most, fuel input side (the pump is segmented, composed of three parts).

So what happened?  It is my hypothesis that when the fuel filter was reinstalled and joined to the pump (using standard NPT torquing guidance of "thread with fingers until resistance then 1.5 turns more" - see Section 5.27), the seal failed at the segment described above (drawn as a blue arc in the top-most plan-view image).

It appears that this same leak has happened to at least two other people.  I neglected to take a picture of the staining on my pump, but the pictures below indicate the same area of leakage.  Both images are from the thread referred to previously.

Upon contacting Andair, they requested that I send the pump back to them in England for complementary repair.  When received, they reported to me that they pressurized the pump and could not find the leak and requested more information about the location of the leak.  The following day they reported that the leak was found (no information given on what was done differently the second time around).  They said that they "replaced the pump body with the latest iteration (now machined as one part instead of three) and also replaced the pumps (sic) inner rotors to the latest spec and a new ESC."  So this new-ish pump should not be subject to the same leakage problem as the original.  I don't know what advantages the new rotors or an "ESC" (electronic speed control?) give me.

Here are images of the new pump with the single machined body.  It appears to me to be composed of two pieces with the fittings attached via hex screws.  The joint where it leaked before remains on the new pump body.

The new pump makes the same noise as the old one however its wires are now Tefzel.