OFCOM Fibre NGA cost model (for access fibre charge control)

OFCOM have published a consultation on their proposed cost modelling approaches for Fibre access networks (NGA) ready for the next Wholesale local access market review.

This sounds very dry but could have a major effect on the future BB technologies used in the access networks.

They are basing the models on the costs of FTTC networks only and will use these if they decide BT has SMP in the fibre acess sector, ( Highly likely on their record), to set Charge controls on BTs Fibre access products.

There is a risk that if they do this that BT will have no way of recovering the higher costs of FTTP network rollouts and this may consequentially reduce the scale of Future FTTP rollouts.

This may also affect the competiveness of alternative Fibre access networks in areas where they are in competition with BT.

The document is available for anyone to comment on but the suggested modelling is complex. There are only two questions to answer

All the documents are available from the OFCOM site

I am suspicious of any model that does not separate FTTC from FTTP.

There may only be relatively few number of FTTP end users. But the costs are very different, The inital end user connection costs are higher but you don't need a telephone exchange. Hopefully early maintenance costs are lower

Also FTTP is where there is opportunity to do things pro-actively to encourage its roll-out...

Me too, as the cost models are wildly different. Quite apart from that, what about the treatment of g.fast? It's pretty clear that OR will have to deploy g.fast pretty quickly in order to compete with VM on technical capabilities, so there will be a considerable amount of extra investment required on NGA networks. Or, alternatively, the payback period on at least some of the existing FTTC installation may not be as long as expected.

There is another issue in that (by my estimation given the number of VM subscribers) and that is that in the areas where VM & BT both operate, the number of "fibre based" connections from each must be broadly similar. I suppose the BT part may be rising faster as there are more retail operators, but in any event, at the moment, a question arises whether OR really has SMP in the areas where it competes with VM. I don't really think OR do have the capability of pricing GEA-FTTC in those areas at whatever level provides the best return. (And, with national pricing, that has a knock-on effect of course).

In any event, I think that OR will have already factored in reductions in the wholesale price into their investment model.

As far as FTTP goes, we know it's been OR policy to price functionally comparable GEA-FTTC and GEA-FTTP products at the same rate, but is this actually set in stone? Is there room to have different pricing, or will Ofcom mandate these have to be treated the same? (Or the same as any forthcoming GEA-G.FAST product).

One company that will be awaiting the outcome with a lot of interest (and some trepidation) will be VM. A substantial reduction on the cost of OR VULA products could cause them a lot of pain and make them redraw their investment plans.

Then there's the financing of any USC. If there are no "excess profits" to be made from the mainstream FTTC deployments, then it will surely impact the whole investment model to hit that 99% 10mbps USC service (or whatever has been half-promised).

The real issue in my view is the monopoly BT has in the local loop outside of the cabled areas. Overlaying another local loop is not going to be financially viable so the local loop needs to be removed from direct BT control. Floating BT Openreach off as a separate BT company should give sufficient separation in addition most of the UK does not have access to VM. They do serve a lot of the population as they are focused on the areas of highest population but geographically most of the UK does not have access to VM services

It would be interesting to properly work out the cost of bringing Fibre to say 90% of the UK. Not actually bringing it to the door but passing the homes. It should also take into account the cost savings from discontinuing the legacy copper network and the money realised by selling off all the old copper cables that could be recovered

Those were my concerns also.
Section 3.8 states
"We therefore propose, to the extent that NGA cost estimates are required as part of the WLA Market Review, to base the modelled costs on those of a national efficient operator, building an NGA network using FTTC and VDSL2 technology."

There is no allowance for FTTP costs so FTTP cost recovery will be difficult.

There is no account of G-Fast in the costs.

The output appears to be regulated GEA prices ( Technology non-specific).

The model only deals with costs for Commercial areas, but this leaves 2 possible scenarios,
1. The regulated prices will only apply in commercial areas
2. The regulated prices will apply in both commercial and BDUK areas.

Both of which have some unpleasant knock-on effects either for BT or some consumers

I would assume 2. as this is to do with BT's SMP.

The SMP decision is outside this BUT OFCOM have always ignored VM when looking at access markets historically, so BT has always come out as having SMP outside the Hull area. Whether this will change if VM cover the 60% they have targetted will be interesting.

There also appears to be NO allowance for the costs of providing power to Cabs in the model.

Edited by kitcat (Tue 10-May-16 11:43:52)

If you are willing to fund me to the sum of £2.9m then I'll happily work on a costing for FTTP to pass 90% of UK homes.

Initial question, is this a total new build? Or extending from existing aggregation nodes? Can Openreach duct be used where available? Is overhead fibre allowed?

You do realise that if one person in a street is passed but does not take the service then the copper cannot be recovered? i.e. you need 100% take-up so that voice is on the fibre before you can recover copper, unless voice USO is to be rewritten. Suspect battery backup will eat into the revenue raised by copper recovery.

Bob

This is all to do with the price OR can charge for access to fibre services. It is irrelevent to who owns it. If the price is set by FTTC costs any owner will be unable to recover the costs of FTTP rollout be that BT or a seperate OR, therefore FTTP rollout will be constrained ( as will G-fast).

The equation for benefits due to copper recovery and resale for scrap is extemely dependant on the price you can get for the copper. Historically this has only been positive at the top of the economic cycle and usually only spans around one in eight years.
At present prices it would not cost in.

Other factors that affect this are; the costs of repairing the damage to other cables in the duct ( eg the fibres you ave provided), the cost of seperating the copper from the plactic sheath in an envoironmentally friendy way, the physical cost of recovering small cables ( 400 pair and smaller) .

There is also the effect that putting large quantities of copper scrap actually have on the price you can get, ( you tend to crash the market!).

My experience is that only in two years out of the past 30 ( two seperate occasions) has telephony scrap recovery had a positive value. (Both times when the price of gold was high enough to help balance the equation)

Whoever owns the OR network is entirely irrelevant to the cost model. The same economics apply in both cases. In some areas it will certainly be uneconomic to have a second fixed network. It is certainly relevant for Ofcom to regulate prices in monopoly areas, but it's less obvious there is a case in the 60% of the country where there is infrastructure competition. If Ofcom force a lower wholesale price, then it will cause a lot of pain to VM and alternative operators and be a distinct disincentive to invest (including by OR).

Incidentally, OR have no powers whatsoever to remove copper. They have to provide MPF as it's a mandatory product. The value of any copper which is recoverable is a very minor issue, especially as the costs of getting the stuff out will be considerable in many cases (and it couldn't be done until every single premise in the area had been converted).

Tim Worstall once did an estimate that the copper in BT's network was worth £30bn, but he made two ludicrous primary schoolboy errors which meant that he'd over-estimated by a factor of over 10. The real value (as pure copper) might be about £2.5bn. As scrap need purification much less. After costs of recovery (where that's even feasible), then much less again.

Edited by TheEulerID (Tue 10-May-16 18:49:22)