Drop the tank and remove the hardlines, replace with -8AN from pump to rail.

Knightrous
Drop the tank and remove the hardlines, replace with -8AN from pump to rail.

I think I will hey. 

The stock fuel supply hard line has now been removed from the fuel tank and replaced with a braided teflon dash 8 fuel line directly from the in-tank fuel pump up to the fuel rail, via a new dash 8 Speedflow fuel filter (in the stock location) with stainless steel mesh filter element.
If I later decide to add a second pump in parallel in the tank, then the new dash 8 line will be sufficient to handle the flow of both pumps.
 

 

 

 

 

 

 
Soon I'll take the car back to MRC for another dyno session and see what we can do with a bit more boost and keeping an eye on the fuel pressure which can now be monitored with the sensor I installed recently.

Nice work dude. I see some unlucky mechanic got the fun job of removing your fuel tank :)

I'm excited about the more boost part :)

*EDIT*
The concerns mentioned below turned out to be due to my fuel pressure regulator, not the fuel pump itself. See here for more info.
 
I did some more detailed data logging with my ECU while accelerating at full power, and upon closer inspection of the data logs I noticed that the differential fuel pressure is dropping at high boost. I don't know if there's another explanation, but it seems that my fuel pump could be running out of flow, even with the dash 8 fuel supply line (it did help though).
 
I'm a bit puzzled because I revised my calculations over and over again and I always came to the conclusion that I should only be using about 85-90% of the fuel pump's flow capability!
 
With a full tank of 98 premium fuel I confirmed that the differential fuel pressure is fairly steady at full power. However, with a full tank of E85 the increased power/boost/fuel usage causes the differential fuel pressure to drop a little bit more than I'm comfortable with. It's been totally OK so far, and without the fuel pressure sensor there was no reason for concern because the AFRs have always been OK, but in the future there could be issues due to pump wear, fuel filter clogging, or a number of other factors.
 
So what does this all mean? Well, I may decide to add another matching fuel pump in parallel and wire it up with a separate relay so the ECU can switch it on only when its required (ie. a typical staged setup) so the return line doesn't get overwhelmed.
The dash 8 fuel supply line would be sufficient for the flow of both pumps, so the cost of installing the line wouldn't be wasted.
One of the concerns that people have with staged parallel fuel pumps is that the engine can still run if one of the pumps dies, so they wouldn't know about it until they go to full power and the AFRs lean out. However, I thought of an idea which I can implement in my ECU to relieve that concern. I can configure the ECU to alternate between the two fuel pumps (say every 10 seconds or so) while only one is needed, so if one of them has a problem I'll know about it before any harm is done. I don't know if anyone else has had a similar idea before, but I was pretty happy when I thought of it. 

robk
 I can configure the ECU to alternate between the two fuel pumps (say every 10 seconds or so) while only one is needed, so if one of them has a problem I'll know about it before any harm is done. I don't know if anyone else has had a similar idea before, but I was pretty happy when I thought of it.

 
Rather than having them switch every 10 seconds, would it work to have them used alternately on each starting cycle? i.e. it would solely use one pump (unless load requires the second to start) but the primary pump would change each time you hit the ignition. (dunno if that makes sense, i'm just waking up...)
 
The 3 series bmw I have apparently may have three fuel pumps! This I know because one is suspected of beginning to fail. Its got at least two. 

It seems to me that switching fuel pumps periodically may create more drama than it's worth. It must certainly contribute to the wear of the fuel pumps themselves.

Dudeman: I totally understood what you meant. Yes that's another option.
Bill: Yeah 10 seconds is probably too often. Perhaps every 2-10 minutes or so would be more suitable. If I don't alternate them too often, then I was thinking it may help by dividing the normal pump wear between the 2 pumps.

Nice idea. I'd also suggest adding some overlap peroid or at least preventing them switching over while at WOT

Have you considered what might happen to the return line if you add another pump?