[170212] in North American Network Operators' Group

home help back first fref pref prev next nref lref last post

RE: Level 3 blames Internet slowdowns on Technica

daemon@ATHENA.MIT.EDU (Naslund, Steve)
Mon Mar 24 23:57:05 2014

From: "Naslund, Steve" <SNaslund@medline.com>
To: Owen DeLong <owen@delong.com>
Date: Tue, 25 Mar 2014 03:56:14 +0000
In-Reply-To: <52CB56AB-6E46-4C1D-A2E0-3ED4154E2916@delong.com>
Cc: NANOG <nanog@nanog.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org

You are right but that is usually how it works with fiber because that last=
 drop to the home is a pretty expensive piece that you don't usually want i=
nstalled until it is needed.  The LECS usually don't even light a building =
unless there is a service that requires it.  I was trying to make the point=
 that $700 - 800 per premise as quoted seems extremely low to me.  The cost=
 of the cable, splices, cases, MPOEs, and especially labor make that number=
 unbelievable to me.  I am coming at this as someone who was in charge of a=
 similar project that connected every building on US Air Force bases to a f=
iber backbone.  An Air Force base is very similar to a suburb in a lot of r=
espects in terms of density and utilities structure.  I was responsible for=
 the design, pricing, procurement, and contractor management on that projec=
t.  We had 3,000 buildings in approximately a eight square mile area and th=
e total project cost was in excess of $12 million dollars which equates to =
something like $4000 per building.  Granted we were doing 12 strands per bu=
ilding but cable costs have fallen since this project so they should be pre=
tty close.

That project included the backbone and the drops into each building.  Betwe=
en those two, the drops into each building was the biggest challenge for un=
derground deployments since no underground conduits were usually available =
and there was a lot of existing infrastructure to be avoided.  I would imag=
ine that if it was a new subdivision it would be much easier but in a 50 pl=
us year old neighborhood there are tons of unknown obstacles and challenges=
.

The labor for splicing and cable pulling itself was provided by Air Force c=
able technicians so did not factor into the costs.  The costs were mostly c=
ivil construction under streets where duct were full and the addition of ma=
ny manholes and handholes because original manholes were not in the right p=
ositions to support the infrastructure or were decayed from being in ground=
 for 50 plus years.  I would say that about half of the money went for civi=
l construction of duct infrastructure and the remainder went to cable and v=
arious hardware items.  Yes, you could go all direct burial but under stree=
ts, that is a maintenance nightmare that you are going to pay for someday. =
 The Air Force required manholes and conduit under streets to allow for fut=
ure serviceability and it was probably a good move since we did use a lot o=
f pre-existing conduit going from copper to fiber.


Steven Naslund

>>This assumes installing a single home on demand.

>>In reality, if you're going to implement what Jay and I are suggesting, t=
hen you dig up a neighborhood at a time and drop a bunch of strands of fibe=
r (I'd guess 8 or 16 as likely numbers) per >>household.

>>Owen



home help back first fref pref prev next nref lref last post