National Broadband Plan: Discussion (Resumed)

I welcome Mr. Matt Yardley and Mr. Pat Kidney who are both partners in Analysys Mason. They are here to discuss the national broadband plan.

By virtue of section 17(2)(l) of the Defamation Act 2009, witnesses are protected by absolute privilege in respect of their evidence to the joint committee. However, if they are directed by the Chairman to cease giving evidence on a particular matter and continue to do so, they are entitled thereafter only to qualified privilege in respect of their evidence. They are directed that only evidence connected with the subject matter of these proceedings is to be given and asked to respect the parliamentary practice to the effect that, where possible, they should not criticise or make charges against any person or entity by name or in such a way as to make him, her or it identifiable.

Any submission or opening statement provided to the committee will be published on the committee's website after the meeting.

Members are reminded of the long-standing parliamentary practice to the effect that they should not comment on, criticise or make charges against a person outside the House or an official either by name or in such a way as to make him or her identifiable. I remind members and witnesses to switch off their mobile phones as they interfere with the sound system. I ask Dr. Yardley to make his opening statement.

Dr. Matt Yardley

I am a partner in Analysys Mason and head of its Manchester office. I am also Analysys Mason's lead partner on the national broadband plan in Ireland. I am joined today by my colleague, Mr. Pat Kidney, who is also a partner and head of Analysys Mason's Dublin office, our day-to-day partner on the national broadband plan. I will provide in the following statement some background on Analysys Mason and take the committee through the scope of our engagement in the national broadband plan, followed by commentary and some aspects of our advice.

Analysys Mason is a sector specialist in telecommunications and works across three main areas, namely, strategy, transaction support, and policy and regulation. The company was established over 30 years ago and now has 260 staff in 16 offices around the world. Our headquarters are in London. I have been with Analysys Mason for over 22 years and have worked extensively on broadband-related projects for a wide range of clients, including operators, investors, regulators and governments. My particular focus is on national broadband policy and implementation which is an area in which I have been heavily involved in the UK and other international markets. I have given evidence to competition tribunals and in commercial arbitration cases in the UK and Ireland on high-speed broadband networks. My role in the project for the Department is to provide overall project direction and quality assurance to the team of technical experts working on the project. That team has expertise across fibre networks, fixed wireless access networks, network deployment, state aid, IT systems and other technical areas.

Following a competitive tendering process, Analysys Mason was appointed in January 2015 as technical adviser to the Department, working closely with its chief technical officer, Mr. Patrick Neary. Our scope of advice was in two main phases. The planning phase included technical strategy, network cost modelling, developing the broadband specification and identifying the premises in the intervention area. At the procurement stage, our work included developing various technical specification documents and technical evaluation criteria, interacting with bidders during the technical dialogue meetings, undertaking the technical evaluation and providing various technical inputs to the procurement documentation along with other ad hoc matters. In addition, we have provided various inputs to PwC and Mason Hayes & Curran on, for example, state aid matters and some regulatory issues.

Analysys Mason's role in the pre-qualification of bidders centred on assessing the technical and professional capacity of the consortium with regard to the deployment, operation and maintenance of networks of a similar size and scale to that envisaged in the national broadband plan strategy. When a change of consortium member occurred, we were part of the team that confirmed the consortium continued to meet the technical and professional capability criteria in accordance with the published tender documents. We were the lead author of all technical related procurement documents covering the technical solution, future proofing, products, deployment, operations and maintenance. We took a lead role in the technical dialogue with the bidders, clarifying bidder responses and evaluating the responses against the procurement criteria. We also provided technical briefing notes and made presentations on technology to various stakeholders. This included, for example, covering technology developments in fibre and wireless networks.

As the committee is aware, Eir submitted plans to expand its fibre network build to cover approximately 300,000 premises in the intervention area. In accordance with the published criteria, we carried out an assessment of the credibility of these plans from a technical and deployment perspective. We concluded that they were credible. The intervention area was changed as a result. Another area of focus for Analysys Mason was on network costing. This took place early in our engagement and was updated at various points throughout the process as more information came to light. Our costing work included capital expenditure, or CapEx, and operating expenditure, or OpEx, elements covering both the initial build of the network and its long-term operation. These costs were then taken by KPMG as an input into its subsidy model. We were also involved in technical aspects of the contract development, helping the Department to manage the various technical risks that had been identified throughout the process.

The final tender was received in September 2018. We concluded that the bidder proposed a technical solution that was capable of delivering on the Government's requirements. As part of that analysis, we noted that the total cost of CapEx plus OpEx was broadly similar to our costing model when differences in take-up were accounted for. I am happy to take questions on the advice we have provided.

Before I open the questioning to members, Dr. Yardley referred to a matter which has been raised a number of times here and elsewhere, namely, the capacity of the consortium to deliver this project. Is Analysys Mason confident the consortium has the capability and capacity to roll out broadband, in particular in light of the changes to it? Is that view based on the previous experience of consortium members? Can Dr. Yardley elaborate on Analysys Mason's confidence in the consortium's ability?

Dr. Matt Yardley

We understand this is a critical area for the Department to understand. I break it down to two stages. During the early stages of the procurement, namely, the PQQ stage, it was really about the consortia demonstrating experience of other projects of a similar scope. They did that across a range of areas. We were interested in experience in the deployment or build of similar networks, the long-term live operation of similar networks, their experience of operating wholesale networks, which is particularly relevant to any state aid project, and some issues to do with quality, health and safety and so on. At that point, we were confident they met all of the criteria laid out at the PQQ stage. Moving forward through procurement to, in particular, the final bid stage, the bidder provided a great deal more detail on the proposed network. In summary, the bidder is building a substantial organisation. That includes 265 people in the core National Broadband Ireland, NBI, business. Many of the senior managers of that entity are now public. The NBI website lists 12 people. What is interesting there is that many of those people are industry veterans. They have been working in the telecommunications sector for a long time, many of them in Ireland, for a whole range of operators. At the senior management level, we were confident that this was shaped in a well structured manner. Beneath that, we were looking at how all of the functions within the business were structured. That is still within NBI. Part of the evaluation criteria looked at that in some detail from an ongoing operations perspective. We turned then to all of the partners they would rely on for delivery of the network. We have there a range of companies which are very well established in Ireland and which work with all of the operators in Ireland themselves. These are the main subcontractors for the build and operation of the network. In that sense, NBI is not doing anything different from any other operator in Ireland. It is subcontracting to a range of very well qualified organisations. That was demonstrated at the PQQ stage and throughout the process to final bid stage.

The whole track of evaluation was very carefully done. NBI met all of the criteria. As such, it passed that test. Looking more broadly at how the organisation is gearing up, this looks and feels like a credible wholesale entity. It is at the right scale, which is a scale similar to what we see in other markets around Europe. It is partnering with well established and recognised organisations. As such, we are confident that a very credible bid has been put forward.

Is Analysys Mason confident that the cost of the proposed roll-out is satisfactory?

Dr. Matt Yardley

In summary, we are.

As the Chair can imagine, there is a significant amount beneath that. As I set out in my opening statement, understanding the nature of the costs in this kind of operation is complex. Probably the first piece of work we did, back in 2015-16, was to use an existing and incredibly detailed cost model which was developed to understand the possible costs of delivering this network. I can talk the committee through more of that detail if necessary. As we went through the process, we learned a lot from interacting with the three bidders that were in the running at that time. When the number of bidders went down to two, we started to improve and refine the assumptions in our cost model. The initial cost of deploying the network turned out to be very similar to our original estimates, that is, the cost to build the network. We were confident that the cost model in terms of homes passed and to cover all of the intervention area was very well behaved and came in very close to the detailed costed solutions put forward by the two bidders. That gave us some confidence that the model was sound and robust. Those are the key opening comments on the cost. I am happy to go through more of the detail behind that if the committee would find that useful.

I will roll the discussion out to members. We can return to the issue of costs should members wish to so do.

I welcome the representatives of Analysys Mason and thank them for attending. Obviously, cost played a crucial role in the procurement process. Analysys Mason is the technical adviser to the Department. What are the key factors in terms of the cost that must be borne by the taxpayer escalating from €800 million to €2.95 billion? That is one of the major questions around this project. In 2015, the figure of €800 million was pencilled in and circulated by those working on the project. It is significant that there has been a 350% increase in a little more than three years. The equivalent in terms of building a house would be for a house currently costing €200,000 to build costing €650,000 to build in three years' time.

Will I go through each of my questions?

Would Mr. Yardley like to answer the Deputy's first question?

Dr. Matt Yardley

I would like to deal with it now because there are a few aspects to it which I would like to go through in detail. We should be clear that the figures of €800 million and €2.789 billion quoted by the Deputy refer to the subsidy. Our role on this project is on the technical side and understanding the costs. As I stated in my opening submission, we are very much focused on capital expenditure and operating expenditure. The subsidy is separate. It takes those costs as an input, but it is also calculated taking into account future relevance-----

We have been told in the past week or so that the overall cost will be in the region of €5 billion.

Dr. Matt Yardley

That is the total cost, yes.

The total cost will be the €2.97 billion and another €2 billion approximately.

Dr. Matt Yardley

Those numbers are-----

I ask Dr. Yardley to explain the costs, including the subsidy and the overall cost and the difference between the two from his perspective. That might clear up the matter.

Dr. Matt Yardley

The €800 million and the €2.7 billion refer to the subsidy. There were changes to the project which impacted on our estimate of the cost. That was then taken into account in the subsidy model which influenced the numbers cited by the Deputy. From our perspective, there are three main reasons the subsidy figure increased. Two are cost-related and within our scope of work and, as such, I will address them. The first relates to Eir's commitment to roll out in the 300,000 homes area. When we constructed our model, knowing that Eir would build in that area, we made an assumption that we would buy a product from Eir to traverse that 300,000 area. That was a reasonable assumption as it would be the most cost-effective way for a bidder to get through to serve the remaining 540,000 households. Through the process, particularly at final bid stage, it became clear that that product was not fully consistent with what the bidder needed to meet all of its obligations in a 25-year contract. The bidder had some nervousness about relying on that product to traverse the 300,000 area because it was signing up to a range of obligations for 25 years and everything was dependent on the connection through that area. We took the decision on the cost modelling side to inform the Department that it would be more prudent not to assume that product was available but, rather, to build through that area with new fibre using duct and poles from Eir. That is more expensive because it involves deploying new infrastructure across that area. That is stage 1, and it was the most important cost change through the process.

My understanding from answers provided to the committee over the past couple of years - I am open to correction - by Ministers and senior departmental officials is that Eir, SIRO and Enet or, as it is now, Granahan McCourt, were in the game and then the first two bidders withdrew. Obviously, had Eir stayed in the race and won the bid, it would have rolled it out side by side with or in place of its copper network. We were told there would be fibre to every home. That claim was repeatedly made by the Minister. The only change that has taken place of which I have heard is that the promise to bring fibre to every home has been dropped in recent months. My understanding for the past two years has been that if Eir was unsuccessful or did not stay in the race, whatever entity won the contract to roll out the scheme would rent the poles and ducting from Eir at a cost of approximately €900 million. Dr. Yardley seems to be indicating that Analysys Mason costed for a new network being built alongside the existing infrastructure. He referred to new poles and cables.

Dr. Matt Yardley

I apologise if I said that. I meant new fibre using existing poles and ducts. The Deputy is correct that there is a time aspect to this. Originally-----

That indicates that the original cost of €800 million was predicated on Eir being the successful bidder and, as such, the contractor for the work. If another bidder - let us call it Matt Yardley limited - won the contract, it would have to lease the poles and ducting from Eir which owns them and run the cable along that network.

Mr. Pat Kidney

Deputy Stanley referred to an original cost of €800 million, but that was the original subsidy, not the original cost.

Sorry, it was the original subsidy.

Mr. Pat Kidney

There is quite a distinction between the two, but the terms seem to be used interchangeably at the committee. The subsidy does not just depend on the network capital cost and the operating cost; it will also depend on revenue, take-up and many other factors of which we did not have sight. In fact, we were not aware of the €800 million cost until-----

Dr. Matt Yardley

It is a subsidy of €800 million.

Mr. Pat Kidney

Sorry. We did not know about the €800 million subsidy until September, quite late in the process. We can get the exact date in that regard for the committee. The two terms seem to be used interchangeably.

We received detailed network costs and operating costs from two bidders - Eir and the now preferred bidder, Granahan McCourt. The level of costs of the bidders were quite similar. I need to dispel any indications to the contrary.

Dr. Matt Yardley

On the Deputy's question on timing, the €800 million was based on a model whereby we were renting a service to the 300,000 area where Eir was rolling out.

We were building thereafter. For the reasons I have given, that was felt later on not to be suitable. The cost went up because we were then building-----

Dr. Matt Yardley

I am just talking about cost at this point. The cost increased because we were deploying new fibre in the area of the 300,000 premises. That is element one of explaining the cost chain. Since Eir's deployment in the area of the 300,000 premises, we went through a process of asking whether it comes off the map for state aid reasons. We concluded it did. It was then a matter of having to build out fibre through an area where the bidder cannot generate any revenue. There is a double whammy at that point. We have increased our costs, which are all upfront, so the hit is very hard, and we cannot get any revenue from the area of the 300,000 premises. That is a very material change for the subsidy. That is element one of the cost change. I am explaining how it also relates to subsidy.

There was then an additional change to the cost model. It related to a more nuanced point, which was about the assumptions in the original cost model about the mix of underground and overhead infrastructure in the intervention area. This was very difficult to resolve because there is no obligation on Eir to publish the information. There is no public source at all to which one can go to ascertain how many poles and how many ducts there are in the area of the 540,000 premises in Ireland. At the time, we had a working assumption of a certain mix of underground and overhead infrastructure. We saw, when the two bids came in, that the bidders' assumptions were different from those in our cost model. That was quite important to us because it meant our cost was different from the one the bidders were putting forward. We spent a lot of time trying to understand why that was the case. Ultimately, it led to the Department issuing a request to Eir to provide its infrastructure database and infrastructure atlas detailing all the poles and ducts in the intervention area. We analysed these and then found out that the assumptions did need refining and that had a secondary effect on cost. That was not as big as the first element I described, which is the impact of the 300,000 premises, but it did have an effect that was quite material. We learned through the process. We got better information as we went through the process. Those are the two factors that drove the change in cost. With the revenue change also taken into account, it explains how we went from the figure of €800 million to what I believe is the right figure to compare with, €1.8 billion, on a like-with-like basis.

I have just two other questions because I do not want to delay the meeting. In the past couple of weeks, I have been told the cost is now in the region of €5 billion. Dr. Yardley confirmed what I have been told I was wrong about for the past two years and three months, namely, that the hiving off in respect of the 300,000 easiest-to-reach households has materially affected the process and has had a great economic impact on it. I thank Dr. Yardley for this information. He is aware that since the other two credible operators have pulled out — Eir and ESB-Vodafone, called SIRO – what I have described is what happened. It was obvious it was going to happen at the time. My view was certainly contradicted and I was told I was wrong and wide of the mark. I do not get any pleasure out of saying that it now seems I was correct, as clarified by Dr. Yardley today. I take on board all the points made in the past five or ten minutes in this regard.

I have just one question. Eir is being given €900 million or €1 billion - the amount sometimes changes - for the use of infrastructure, part of which is in very bad condition, as is very obvious when driving around the country. Since Eir was privatised, there has been little investment in parts of the infrastructure. Some of the poles have fallen over. Some of the poles, if they have been hit by cars or other vehicles, are broken and held up only by the cables. I do not know with the underground infrastructure is like. I am told by some individuals working in the industry that nothing has been done with it since the mid-1990s or late 1990s and that it is in bad condition. On the other hand, the ESB network is in better condition and has to be because it is carrying live electricity wires. Traditionally, it has been kept very well and has been well maintained. Electricity goes to every house except the houses of very few who, for whatever reason, do not have electricity. One could count them on one hand in any county. I am only aware of two or three people who do not have electricity in my county.

Why is Eir being handed the amount of money in question? It represents a huge slice of the subsidy. It is one third of the overall subsidy, at least. It is almost €1 billion of taxpayers' money for permission to use infrastructure that is in bad condition and that will incur a charge year on year throughout the term of the 25-year contract while the ESB network is in place and is ready to use. The ESB is already using that network to carry telecommunications fibre. In other countries, the same infrastructure is used for electricity and telecommunications cabling. Does Mr. Yardley believes this make sense here? Is it possible?

Dr. Matt Yardley

The fundamental principle on a project of this kind, which comes down from European Commission state aid rules, is based on reusing infrastructure, minimising duplication of assets to be cost-efficient, and minimising distortion in the market as far as possible. I would not characterise what is occurring as Eir being handed the money, in that it is perfectly possible for the bidder to use other infrastructure it thought could be used to meet its purpose and honour all its service level agreements under the contract. It is down to the bidder to choose. When the bidder put the final bid in, it had to have absolute confidence that the infrastructure could be used for the purpose of deploying fibre. At the time, I do not believe it was in a position to say that in respect of the ESB, but that could change. It is perfectly possible through this process for the bidder to engage with other providers of infrastructure if deployment is more cost-effective or if it has more confidence about delivery, timescales or whatever. There is nothing at this stage that prohibits it from happening.

On the comment about aerial versus underground, duct infrastructure is generally not maintained that much unless it needs to be. Typically, duct infrastructure tends to get maintained in more urban environments if it is dug up when there is some property development or such like. It generally does not get touched unless it needs to be but I accept the point that, in Ireland, particularly the west, where inclement weather could affect infrastructure, it will need to be upgraded over time by Eir, and it will recover the cost of doing that through the charges it imposes. On the policy objective, I have referred to what is going to Eir but NBI is delivering through its retail partners services to end users. Ultimately, therefore, the end user will benefit. If there are better ways of delivering in certain areas, the bidder is more than open to examining those options.

So the ESB network could be used.

Dr. Matt Yardley

There is a range of technical and health and safety-related issues that would need to be overcome but the Deputy is correct that there are examples in other countries-----

And in this country.

Dr. Matt Yardley

-----of the low-voltage distribution network being used. That is possible. Of course, it is being used here in Ireland in open areas. Therefore, it is possible.

Mr. Pat Kidney

May I expand on that? There were meetings with ESB Networks, particularly when SIRO was still in the process. They were with a view to making the ESB infrastructure available. I attended some of those meetings with the Department.

The use of the ESB infrastructure is not as straightforward as may have been set out in the media. When one starts to go into the matter, it is quite complex. There are significant health and safety issues involved. There are also issues with wayleaves because the infrastructure does not go along the roads, it goes direct. There are technical issues regarding how the fibre is put on the medium and low-voltage networks as well. There is an engineering solution but the infrastructure and products ESB Networks was looking at developing were not at a stage where bids could be put in and a bidder could commit to meeting deployment targets with certainty regarding costs, etc. The committee would have to ask the bidder why it chose Eircom. The latter, or Open Eir, was offering a service with a defined product and a regulated price. A bid needs to be put in - there is considerable uncertainty about how the ESB infrastructure would be used - and that bid has to be committed to. The bidder, left with that sort of choice, went with Eir. However, there is nothing to preclude the bidder from talking to ESB Networks with a view to using its infrastructure in all or part of their network. There is nothing in the contract which would prevent that. It would have to be done under change control. The change control mechanism allows for changes of this nature as long as the benefits to the project and the State are set out and agreed.

I want to clarify one small point. Had the SIRO bid been successful, am I correct in stating that the ESB would have used its network to roll this out? That was its intention.

Mr. Pat Kidney

That is a fair assumption. We do not know what it would have done but that is a reasonable assumption.

I call Deputy Dooley.

I thank our guests for their presentation. It is helpful that they have set out their involvement in the engagement so far. Has Analysys Mason provided any advice or assistance to Granahan McCourt Capital at any stage?

Dr. Matt Yardley

No.

Is it fair to say that it has advised most telecommunications operators at some point?

Dr. Matt Yardley

Yes. We have advised many telecoms operators, as well as everybody else who is associated with our sector, including regulators, governments, vendors and investors.

It would not be unusual for Analsys Mason to have engagement with a company of standing in the telecommunications sector.

Dr. Matt Yardley

No.

Just last week, KPMG contended that its original costings were out of sync to the tune of €1 billion. It seemed to segue out of that analysis based on technical information that Analysys Mason had provided. Is it fair to assume that this related to the reappraisal of the job at hand as a result of the parallel overlay of fibre that was mentioned in response to Deputy Stanley?

Dr. Matt Yardley

The €1 billion is the change in the subsidy. Only part of that relates to the change in the technical cost.

Did Analysys Mason have a role in the soft costs, or the subsidy?

Dr. Matt Yardley

We had no role in that. As Mr. Kidney explained, we had absolutely no visibility of the subsidy until very late in the process.

Was Analysys Mason's advice all along specific to the technical solutions, the build cost and building maintenance?

Dr. Matt Yardley

It also related to the operating costs.

Mr. Pat Kidney

We advised on software, software support and telecommunications costs.

Were the rates of sign-up, what revenue might be generated, etc, all outside of that?

Dr. Matt Yardley

Anything on the demand side was outside our remit.

How much of the €1 billion could be attributed to the technical change?

Dr. Matt Yardley

We have numbers in one of the reports, but they are redacted. I do not want to give figures that are currently redacted in documentation. There is a material part of that that relates to cost, but the revenue impact was also-----

Were any of the redactions done as a result of our guests' desire or instruction or were they all done by the Department? I presume our guests were asked if it was okay for these reports to be published.

Dr. Matt Yardley

It was a combination. In terms of where we are at in the process, some of the technical costs could certainly impact on how the bidder deals with subcontractors, for example. We were genuinely being prudent on that. We do not have a contract at this point, so from a technical costing perspective the overarching opinion was that we should have some caution on that.

Mr. Pat Kidney

There is significant commercially sensitive information in our reports. We had sight of two bids and costs from competing operators. There is huge sensitivity regarding some of the figures. We have obviously used them, but-----

Can Mr. Kidney discuss the sensitivity of those costs and indicate how they might compromise the process if they were released now?

Mr. Pat Kidney

For example, one of the subcontractors that is being used by the preferred bidder is also being used by other operators in the market. They may have different costs.

Why would how they might display their costs elsewhere be of concern to either our guests or us? From our point of view, this is a State contract. We are charged with a responsibility to the taxpayer, so it is for us to try to get to the bottom of the matter. Is it not the case that there is no commercial sensitivity as it relates to this process?

Mr. Pat Kidney

It is commercially sensitive in the market, and the market is not just this process. It is much wider, and commercial companies have provided the Department, and us as its adviser, with commercially sensitive information. We have to respect the confidentiality of the process.

That was outlined last week as well, and with respect, we all understand the nature of something that is commercially sensitive.

Dr. Matt Yardley

I might add a point that may help address the question. The costs were looked at by us in incredible detail. Our work involves going into a lot of detail to understand the volumes and the unit prices involved in this bid, if the committee is concerned that these costs are not fair. In that report, which the Deputy can read into, there are certain things which give a sense of where we landed on this. We were happy with the costs at an aggregate level. We had question about certain elements beneath that which need working through with the bidder in the final stages. Due diligence has been done on that, but it is fair for us to be cautious on the commercial sensitivity point. That is separate to the understanding the State is exposed to.

The increase in the cost related to the decision to roll out a parallel fibre cable along that network of 300,000 premises. Could our guests give us some sense of that?

Dr. Matt Yardley

I am afraid we cannot.

Can they try to explain the cost differential between the purchase of the product that was mentioned, which was some capacity to splice into the existing network of 300,000 on the periphery of that outer wheel? What might the differential in cost have been had they proceeded with the purchase of that product, in other words, attaching it to an existing web?

Dr. Matt Yardley

I cannot do that either. It is the same question, in essence.

That is fine, but I still need to probe it. Can Mr. Yardley explain to us why the decision was taken to overlay a parallel network along that network of 300,000 homes? What was the guiding principle behind that?

Dr. Matt Yardley

We set off with what we thought the most cost-efficient way of doing this was, which was buying that product through the 300,000 home area. When the final bid is submitted, the bidder has to have absolute confidence that those inputs being taken from others are suitable for the duration of the contract.

We had some understanding of this. The bidder was concerned about taking this risk because the product is not being provided over a 25-year term. At that point, the bidder had no view on how it would be made available throughout the 300,000 premises and how it related to the intervention area. The product development was not specified, as I understand it. We have looked at future concerns about pricing, terms and conditions and range of other matters in some detail from a technical perspective.

Does Mr. Yardley not accept there is potential for that contract to breach state aid rules? On the basis it is rolling fibre past the homes of 300,000 people who have a service, what is to stop Granahan McCourt in the not too distant future from offering a connection service to those 300,000 in competition with the existing commercial provider? If that were to happen, it would mean using state aid to provide a benefit to Granahan McCourt to compete in an area not within the intervention area.

Dr. Matt Yardley

We are not a state aid adviser on the project but I have general experience of working on state aid projects in other places. That area is not in the intervention area so the bidder is not allowed to use public funds to serve the market in that area. It cannot use anything that is-----

How could it be prevented? If a fibre optic cable passes the door of somebody within the area of the 300,000 that is paid for by the taxpayer, what prevents a national broadband company-----

Dr. Matt Yardley

I can give an answer in principle on how I suspect it may work. I do not know for sure because I do not know this part of the contract. As I say, it is under the state aid provisions.

This is why it is useful to have the witnesses before the committee from a technical point of view. Last week, we had a discussion on what the net value of the company might be at the end. To me, the value of the company at the end has increased dramatically given the information the witnesses have provided us with, which is that this is not fibre optic connected to the outer extremities of an existing network but the building of an entirely new network servicing 840,000 premises and probably more because my understanding is-----

Dr. Matt Yardley

I do not think that is correct. That is not the characterisation. The project is serving homes in the 540,000 area-----

But the network has the-----

Dr. Matt Yardley

We just have to get there because of Eir's decision to invest there.

Correct but in doing so-----

Dr. Matt Yardley

That is very clearly demarcating-----

I accept that but in doing so a network is being built that will, perhaps, pass another 400,000 because Eir has indicated in its rolling out of the 300,000 that it has passed another 30,000 to 40,000. It is fair to assume that to get to the 540,000 premises another 300,000 or 400,000 premises must be passed, paid for by the taxpayer, which we accept is necessary, but it creates a long-term value in the company that would not have been characterised heretofore.

Dr. Matt Yardley

All we can say is that effectively the state aid provisions in the contract would need checking to see how they are dealt with. My expectation is it is very tightly managed within the contract.

We can follow that up.

Moving on to Granahan McCourt, the preferred bidder, Mr. Yardley replied to the Chairman's question on the technical capability and suitability of the company and whether it had met all the milestones set out. We are given those milestones as having been handed down on a tablet of stone from somewhere. When we compare Granahan McCourt with the other two main bidders, there certainly does not seem to be the same level of expertise, experience, project management and deployment of broadband services. Was a desktop study done? Did Analysys Mason examine networks that were built by the bidders? How did these companies pass the technical suitability test?

Dr. Matt Yardley

They had to put forward examples of their experience, as I said earlier, across a range of criteria, including building, operating, wholesale specific and other environmental and quality matters. They demonstrated that throughout the consortium. It is important to stress there is the National Broadband Ireland organisation and a lot of the build and operational expertise sits with its partners. There is information in the public domain on National Broadband Ireland's website and we can look at the senior management team and the expertise. I will read out the companies from where the expertise will come. They are Siemens, Enet, Airspeed, Level 3, Eircom, Magnet, Aurora, Rogers, which is a cable company in Canada, and Hibernia. The senior management team are all industry veterans, many of whom have worked in Ireland. That is a good starting point for building a new organisation.

We all accept it is important for that entity to engage with its partners and consortium members with very robust subcontracts that meet the requirements of the State. That is where we are at in the process.

To be clear, there is no consortium. There is a list of subcontractors identified as partners. It is language that has been used loosely throughout the process. There was a consortium at the beginning that included John Laing, SSE and Enet but over the course of the dialogue, the character and nature of that consortium changed to the extent that at the end the only party was a finance house out of Boston, namely, Granahan McCourt. There was recognition that there were subcontractors but not against which their reputation could be held to account as they were just subcontractors. They are not part of the consortium. They are not asserting anything other than-----

Dr. Matt Yardley

I will make an initial comment and perhaps Mr. Kidney will add a little more. That structure is the same as the way in which the other companies the Deputy has cited operate with these organisations. National Broadband Ireland is not doing anything differently in terms of its contracting arrangements compared with other large well-established operators in Ireland and this is a model we see everywhere. Telecoms operators generally outsource a lot of these functions to companies similar to this.

In those cases, the parent company usually is a telecommunications company.

Dr. Matt Yardley

We fully accept it is building a new telecommunications organisation.

Granahan McCourt is not a telecommunications organisation.

Dr. Matt Yardley

No, but National Broadband Ireland is a new organisation being staffed with people who have vast experience of the Irish telecommunications market. We accept this is not a large well-established company that has been operating in telecommunications for 20 years in Ireland but what it has put forward through the entire process at pre-qualification questionnaire stage and final tender stage against the criteria, which is the only reference point we have through public procurement, has met those requirements.

To move into an area that would be helpful to us in trying to understand costs, with regard to the technical solution, did Analysys Mason consider the balance between fibre and fixed wireless at all? The concerns for us are not only with regard to cost, although that is important. We are also looking at the rate of deployment and the speed at which people can access high-speed broadband. Was any consideration given to a mix of fibre and good quality fixed wireless?

Dr. Matt Yardley

This is a question we get asked a lot in other jurisdictions. It is a very important question in a project of this type. The Deputy termed it our solution but it is not our solution. We went to the market with a specification and the market told us the solution. At that point we had three bidders and they all came forward with a fibre solution.

They are the best entities to optimise the technical solution. We evaluate it and decide whether it meets the criteria but we are not buying a technology as we cannot to do that under state aid rules. We take an option that proposes the best solution for what we require. We looked at fixed wireless early in the process and we did costing work before we engaged any bidders but we had no idea what the solution would be at that point. We did a range of modelling exercises and looked at fibre-only options, mixes of fibre and fixed wireless and we also looked at the different qualities of service on fixed wireless. We concluded that, over a 25-year period, wireless would probably cost the same, or even more, than fibre would cost. It may seem an unusual finding and it might not be the case in many markets but it happens in Ireland because of the distribution of the rural population, which is sparse compared to, for example, England where clusters of villages or hamlets occur even in very rural parts of the country. The ribbon development in Ireland is a challenge for any technology, whether fixed or fixed wireless.

That is based on cost.

Dr. Matt Yardley

That is based on cost but we also looked at different levels of quality.

A third point of evidence relates to a report by Oxera Consulting with Real Wireless looking at high-speed mobile, which ComReg put out last week. ComReg explained that covering Ireland with high-speed mobile coverage, at a broadly similar specification as we have, would cost €1.9 billion, would require approximately 6,000 new sites and might take ten years to deploy. ComReg also looked at fixed wireless. It did not come up with costs but some work on that was done by Plum Consulting, which stated that it might require between 4,000 and 5,000 new sites.

There has been a lot of talk about the number of sites over the entirety of the intervention area. I am talking about the sweet spot that exists somewhere between 60% or 70% fibre and 30% fixed wireless, which might be used in an effort to get the necessary rate of deployment without it costing more. Dr. Yardley can talk about resilience over 25 years but who knows what fixed wireless technology will be like in ten, 20 or 25 years? We are trying to address the needs of consumers who are on the five-year, seven-year or ten-year roll-out. It is difficult because a life cycle has gone by for these people in that period. Is there a way to do this that will give quicker deployment?

Dr. Matt Yardley

We looked at the mixes. We were looking at the budget impact so, first and foremost, we were concerned with cost. It is an extended period and I can understand the frustration but there is no guarantee that moving to a mobile or fixed wireless infrastructure would occur any faster. In fact, the ComReg view was that building so many sites could take ten years. It seems a long period and is much longer than what we have from the bidder. The reason it could take so long is that we would need new towers and sites and a lot of these would be on private land. With fibre net, we are building infrastructure down the roads on poles and through ducts which exist. We do not have to build anything new but just deploy new fibre, which is very different from having to deploy a new mobile or fixed wireless site where we have to acquire the site itself and connect it to power and backhaul. We have to deal with private landowners, negotiate wayleaves and get planning permission. Those things tend to be complex and there is no assurance that moving to that mode of deployment would buy us any time. The bidder has put forward the best solution for the project. We did not dictate it.

Mr. Pat Kidney

I need to emphasise the extent of the settlement pattern and how that influences these partial models. The figures for inhabitants per square kilometre in rural Ireland are pretty low and I believe they stand at 27. That does not tell the whole story, however. It is about road frontage and we were quite amazed when we did the analysis on this as part of a wider team. We have built along roads in a way that is amazingly consistent all over the country. Houses tend to be built over stretches of a couple of kilometres, with four or five in a row and then a gap until another four or five. There is a linear distribution of houses and premises in rural areas and this does not lend itself to a high site. We could end up building a high site that would only cover tens of homes and we may still miss some of them due to the topography and foliage etc., meaning we would have to build out to those houses. We did considerable work on this across the entire intervention area in every village and small cluster. We looked at going 1.5 km from each village and took four case studies. We looked at a detailed design of what would happen if we were to build out to homes and premises in this way. It left us with orphaned premises and 16% of them would fall out of coverage as a result. It is a considerable challenge to deploy wireless in an overlay sense. When people suggest 20% wireless, it would not be in an area contiguous to a village. It would be spread all over the place and would be difficult to get to. When the bidders came in they started looking at wireless and two of them had related wireless entities, but they decided on a 100% fibre solution. It comes down to road frontage and the settlement pattern in rural Ireland, consistently across the country.

I assume that while in the ESB network the poles go across fields, the network must nevertheless be configured around our roads-based development. There would not be the same problem with ESB poles as we would have with fixed wireless in those circumstances as the poles would generally be contiguous with the road. Is that a fair assessment?

Mr. Pat Kidney

It still has to reach the same homes.

We all know about our ribbon-based development and we have been giving out about it for 50 years but the ESB must be configured for a roads-based development.

Mr. Pat Kidney

Yes.

It would not be a disadvantage to use the ESB poles.

Mr. Pat Kidney

Yes, that is fair. The low-voltage ESB network delivers electricity to every home.

It is using to dealing with our pattern of development.

Mr. Pat Kidney

It is configured to deliver to every home.

Can the witnesses describe the network? In the case of the past few miles, where does the backhaul fibre start and where does it go to on this network?

Mr. Pat Kidney

Let us imagine a tree. This particular technology is Gigabit-capable passive optical networks, GPON, and it emanates from a point of presence or an exchange, goes out along various roads and then branches off-----

Does the backhaul start at the exchange then?

Mr. Pat Kidney

It does. It starts at the exchange or the point of presence. We are talking about the access network, just to clarify-----

This is all a wholesale access network.

Mr. Pat Kidney

Yes, it is. We generally divide networks, however. If we want to build a network that will cover homes around a particular exchange then we start at the exchange. We build out along a road, probably starting underground and moving on to a row of poles along that road at the edge of the town and then we continue out into the countryside. The network will be split as other roads are encountered, and that is why I suggest we visualise this network as a tree. The backhaul fibre is in the same cable as the cable that is used to drop connections to homes. For illustration purposes, let us say we end up with 36 fibres in a cable. Some 12 of those fibres might be going all the way back and be connected to the exchange while another 12 of those fibres would be continuously reused to bring to connections to splitters. Those splitters would then drop down onto the connections. It would be better if we had a diagram.

Some of those splitter connections are included as part of the backhaul.

Dr. Matt Yardley

I suppose this-----

Mr. Pat Kidney

It depends on the classification. Looking at the cable, it would not be possible to tell if it was backhaul or if it was access. In most cases, in reality, it is mixed. The cable that is going through the 300,000 would just be backhaul. In most cases, however, it would be mixed. The same fibre would be used but a fibre would be dropped into one part of the splitter and then perhaps four fibres from the same cable would be connected to the other side. It is a splitting of one to four. Then, moving on, there would be another splitter that drops onto the connection. This really does need a diagram to explain the concept properly.

Where is the exchange? Is the exchange a link point?

Mr. Pat Kidney

It can be. That cable will go back and into the exchange. It will be terminated on an optical distribution frame. Let us call that exchange an aggregate point. It collects a number of other cables. Let us suppose there are four roads going out of the town and four cables come in. In that case, that point could be a point of interconnection for retail providers or the aggregated traffic could be brought to a bigger exchange. It would depend on the geography.

Turning to the traffic from that exchange, where does the backhaul from that go to?

Mr. Pat Kidney

The products available on this network are wholesale products that run from a point of presence to a premises. The retail service providers have options as to which point of presence they go to but they provide the connection to the Internet. A retail service provider could have one national interconnection point with the entire network. Returning to the Deputy's question, in that case the traffic from the exchange would then all be brought back to that network.

Would the service provider have its own connection point from that exchange to-----

Mr. Pat Kidney

----its peering point at the Internet.

Dr. Matt Yardley

Some might do. The reason we are perhaps struggling with this is because different operators have different network configurations. Depending on what network assets operators have, one operator might be quite happy picking it up in exchange A, whereas another operator might not have anything there and may want to pick it up in exchange B. The network solution has to connect to those various points to serve its retail service providers. It is in its interest to do that because that is where its customers will pick up traffic.

I apologise to the Chair but I just want to tease this concept out some more because it is complex and it is good to try to understand it. I ask Mr. Kidney to think of a county. I do not know where he is from but he sounds like he is from Dublin.

Mr. Pat Kidney

I am from Cork but my wife is from Dublin.

That is fine. Mr. Kidney picked my favourite county, as my mother and father are from Cork. Where in the county is Mr. Kidney from?

Mr. Pat Kidney

I am from Cobh.

Cobh is not in the rural broadband area.

Mr. Pat Kidney

It is not.

Let us move out from Cobh to somewhere in east Cork, perhaps Youghal or beyond it.

Mr. Pat Kidney

Kilworth, perhaps.

Let select Glantane. Glantane would probably be in the rural broadband area.

Mr. Pat Kidney

I do not know offhand. I do not think that it is.

Where would the exchange be for that type of area? What is the typical type of town that Mr. Kidney talking about? Is it a village or is it bigger than that?

Mr. Pat Kidney

There is a village close to Cobh called Ballymore. The exchange for that village is in Cobh.

That is fine. Cobh is the exchange.

Mr. Pat Kidney

An aggregation point of some sort may be placed out in Ballymore and that might serve the surrounding area, recalling the tree structure we spoke of earlier. A single cable, or perhaps two if that aggregation point was serving many customers, would be brought in a ring into Cobh.

There is a reason I asking about this. Going back to a response to Deputy Stanley earlier, one of the increases in cost was explained as being because there was a need to jump over the 300,000 area. It surprises me that is such a difficulty. As I stated, all of those exchanges are already on the other side of that 300,000 boundary area anyway.

Mr. Pat Kidney

In general, they are not.

Dr. Matt Yardley

That is the issue.

That 300,000 area is very big then and covers a large area.

Dr. Matt Yardley

It tends to be clustered relatively closely around exchanges.

Was that not very clever?

Mr. Pat Kidney

I described the basics yesterday to-----

I have been talking to-----

We are nearly finished with this question and then I will let in the other members.

Mr. Pat Kidney

The 300,000 is basically the 300,000 premises near the exchanges.

They got us rightly.

I have been talking to people in the business who just signed 25-year backhaul contracts with Eir. It is not unusual to do a 25-year deal. If enough fibre optic cable is brought into the country, then the contract to serve a data centre has to extend over 25 years. It is standard business practice in this country to do a 25-year deal with Eir. That is not unusual.

Mr. Pat Kidney

That could be done if the product is available.

People are doing it.

Mr. Pat Kidney

It is very few people in very specific circumstances.

They have done it, however.

Mr. Pat Kidney

Data centres have done it.

Yes, but we have witnessed a major price increase. Why has that occurred? Did the bidder think that Eir would not have the capacity on its backhaul network?

Dr. Matt Yardley

As I said earlier, there were many reasons why the bidder was concerned about committing to that in the bid. I do not think, again, that there is any reason why that could not be reconsidered. All I was describing was the reason why the cost went up as shown from our internal costing work. If there is any potential to reduce costs, then of course everyone is incentivised to try to do that.

One part of the scheme that we hope takes off, please God, is a great increase in the Internet of things. We hope to see every dairy farmer sending in his milk figures in via Internet and every child using it to do his or her homework. There would, however, need to be a check that the network has a 25-year capability to cope with the increase in the volume of data that we want to encourage. Was there a concern in that regard or does the contract guarantee that sort of backhaul capability within the provisions?

Mr. Pat Kidney

I guess there is quite a bit of capacity on cable. The speed that one gets through a fibre optic pair depends on the equipment on the end of it. The network that is being designed here can provide 10 Gbit services.

That is more the case now, when there has been a necessity to create its own backhaul, effectively.

Mr. Pat Kidney

To explain this, it would probably be best for us to do a paper on the different components.

I do not know if that would be helpful to the Deputy but if Mr. Kidney wants to provide us with something technical-----

I apologise for taking so long but if Mr. Kidney could provide that, I would really appreciate it.

What is the use of the metropolitan area networks, MANs, in this tree?

Mr. Pat Kidney

To get from the edge of the town into the point of presence.

I think there are around 200 MANs. I cannot remember the final number.

There are 88.

Are they all being used?

Mr. Pat Kidney

I think 80 of them are being used. I can confirm the exact number later. I do not know it off the top of my head. The preference was to use the MANs and to use the point of presence or POP as part of that.

In terms of the cost increases, the witnesses said that when they looked at the atlas of ducting and overhead wires, it led to a cost increase. I presume that means less ducting and more overhead lines. Was that the reason?

Dr. Matt Yardley

The indication was that we had too much aerial so we changed it. We had to deploy more underground and that was more costly.

Was that because of the quality of the aerial connection?

Dr. Matt Yardley

No, just the fact of what was present. The assumption on the mix of whether it was duct or aerial was incorrect for the intervention area.

The amount of aerial had to be increased. Is that correct?

Dr. Matt Yardley

Yes. We changed the amount of underground to aerial and that led to an increase in costs because the underground option is more expensive.

The intervention areas did not have as much ducting as expected. Is that correct?

Dr. Matt Yardley

There was actually more duct being used than we thought and it is more expensive to deploy.

Is ducting more expensive?

Dr. Matt Yardley

Yes.

Dr. Yardley said that the third reason for the increase in the subsidy was not related to the physical construction costs of the network. What was it related to? He said there were three reasons. One was having to build through the 300,000 area and the second was the mix-----

Dr. Matt Yardley

The third reason is that with the 300,000 outside the intervention area, the project cannot generate revenue from that. We did not inspect the revenue side at all but we are trying to help the committee to understand why the subsidy can move around quite a lot. We are talking about a combination of a cost change, which happens upfront and hits quite hard but also the revenue building up over time is much less. When one looks at that from a cashflow perspective, it has a very strong effect on the subsidy. We never saw the effect in totality. All we saw was the change in cost as we changed our cost model.

I have talked to industry people with similar experience in other jurisdictions. They have told me that this situation is unusual in that we are subsidising the construction costs. In fact, we are paying for the construction. We are also subsidising the revenue risk. We are covering both in our subsidy. What I hear from others is that it would be standard or normal enough for the State to provide a backstop on the revenue side because we do not know what the take-up will be like. The State would cover that risk to get the project built. That gives any operator, developer or bidder certainty on the revenue side. Why is the construction side being subsidised? Ordinarily in this industry a developer would go to a financier with a backstop on the revenue side and seek to borrow for construction. Why are we subsidising the construction side, as well as guaranteeing the revenue side?

Dr. Matt Yardley

We are not guaranteeing the revenue. The Deputy is right that with all of these types of projects, the characterisation is that the market risk sits with the bidder.

Mr. Pat Kidney

That is the case here.

Dr. Matt Yardley

That is true in this case.

Why is the construction risk residing with the Government? Why are the costs being borne by the State? It would be more typical with a project of this type for the State to cover the revenue risk and let the developer cover the cost of construction and take some of the risk. The developer is taking no risk here. The developer is covered on the construction side and on the revenue side.

Mr. Pat Kidney

NBI has the revenue risk in its entirety. If the developer does not make its projected revenues-----

The witnesses have just said that the subsidy went up because the revenue fell.

Mr. Pat Kidney

A prospective 300,000 customers disappeared from the network but the developer still had to build through that area. There was a double whammy. The developer had to build at least part of the network, albeit not the actual drop connections, though the 300,000 area but with no revenue arising from the 300,000. To add to the vicious circle, the 300,000 are the customers who would be served first because they are the closest to the exchange. Once premises are passed, there is nothing to stop them being connected and starting to generate revenue, even before the deployment area is complete.

Dr. Matt Yardley

I would add that the reason I introduced the subsidy into the discussion is that I am concerned that people are mixing up costs and subsidies. We do not have any visibility of how the subsidy model works or indeed, anything on the revenue side of this equation. I was just bringing that to the table to explain that there is a third component which will affect-----

This is complex and we have to dig to get to the bottom of it, as best we can, with blacked out documents. The witnesses said that the cost increases were material to the subsidy increase but that they were only a small factor or part-----

Dr. Matt Yardley

I said they were material.

The other issue was not just the loss of the 300,000. There were other subsidy factors on the revenue side, as well as the loss of the 300,000. We need to find out what those factors are.

Dr. Matt Yardley

That is correct.

Deputy Stanley wants to come back in briefly.

I have a quick question on the 300,000 households or premises area. On the day of the press conference, a map was handed out and it was very clear to me what would happen. I could identify areas in my constituency, which I know best, where there were clusters of houses and quite a high density of housing for what are rural areas. Other more-difficult-to-reach premises were out the road, up boreens and were more scattered. Mr. Kidney has referred to the difficulties in reaching them. We are talking now about having to roll out parallel fibre cables on existing poles or through ducting in the area that Eir now has, which is now covering approximately 340,000 households. It is my understanding from what the witnesses said that the original intention was that the developer would splice into that or would use the telecommunications network that is there already. To be clear, Eir, got that area and took it over. The developer is left with between 480,000 and 500,000 households. Dr. Yardley seems to be suggesting that the developer working on the area with 500,000 premises must overlay fibre cable because Eir has stated that it has the area with 300,000 premises and if the developer wants to pass that area, it must install its own system and fibre cables alongside those installed by Eir. In other words, the developer cannot use the Eir equipment. That is one point that I want the witnesses to clarify.

Dr. Matt Yardley

If I could correct the Deputy on that-----

My second point-----

I ask Deputy Stanley to let Dr. Yardley respond.

Dr. Matt Yardley

That is not correct. Eir did make an offer of a product to traverse the area with 300,000 premises but the bidder chose not to take that up in the end because of all of the issues to which I alluded earlier. We then adjusted the cost-----

Was one of the reasons that the price was too high?

Dr. Matt Yardley

No. The price in the cost model meant that the total cost would be lower, so from the State's perspective, that was better at that point. However, there was a whole range of other technical issues about which the bidder was concerned. We can provide further information on that.

I ask the witnesses to do that.

Dr. Matt Yardley

We had some sympathy with those concerns.

I ask the witnesses to send a list of those concerns to the committee. That would be very helpful.

Dr. Matt Yardley

We will do that.

I have one very small point of clarification, if I may. Obviously, in terms of using pre-existing poles and ducts, a subsidy would be paid for permission to pass the area. Eir owns the poles and the other consortium, if successful, will want to hang its cables on those poles and run its cables through the ducting underground. The consortium would pay Eir for the lease or rental. Is that correct?

Dr. Matt Yardley

That is correct.

This is difficult and we are only learning about it. There are 88 metropolitan area networks and a MAN is equivalent to an exchange. Is that correct?

Mr. Pat Kidney

Yes. There is a point of presence, POP, which would be an aggregate point.

How many other exchanges will we be using as POPs in the national broadband scheme?

Mr. Pat Kidney

I will come back to the Deputy with the number.

What is it roughly?

Mr. Pat Kidney

The total number is around 200. I do not have the exact figure.

MANs would account for about one third of those.

Mr. Pat Kidney

Yes.

Are all the other exchanges Eir exchanges?

Mr. Pat Kidney

Yes.

The ESB has a ring of eight in terms of fibre optic cable backhaul. I do not know if BT still has ducting along the rail lines. There is other State ducting with respect to Bord Gáis.

Mr. Pat Kidney

Most of the POPs and the MANs are connected to the network, as are any that are close to the ESB, for example, Cork city.

Typically, the Eir exchanges would exclusively use the Eir backhaul network. Is that correct?

Mr. Pat Kidney

I do not know the answer to that question.

I presume they would.

Mr. Pat Kidney

They certainly use Eir but whether they use another company or some other company is interconnected to them, I do not know. The bidder would build from the Eir exchanges into the POPs set out on the map.

Mr. Kidney said that it was decided that the bidder would provide its own network in the area of the 300,000 homes. To where exactly does that network go? Where is the point of presence for those additional fibres being put in?

Dr. Matt Yardley

There may be a slight misunderstanding. They are not passing all those 300,000 homes or building a duplicate network passing every home. They are traversing this area to get to the area beyond it. It is a much more discrete-----

Where are they trying to get to?

Dr. Matt Yardley

To the 540,000 homes that lie beyond that.

The 540,000 homes.

As they are jumping over the area with the 300,000 homes, where are they jumping from or what is the other side of that?

Mr. Pat Kidney

In the case of the POPs, or the MAN if it is a MAN town, they are jumping from the point of presence on the MAN, that is, the POP that was built when the MAN was built. In the case of a non-MAN town, they are going to the Eir exchange.

Dr. Matt Yardley

It is important to understand that the bidder is traversing rather than building past every single home in the area with the 300,000 homes.

They are traversing to an exchange on the urban side of Ireland, as it were.

Dr. Matt Yardley

There is a node, POP or some aggregation point into which many lines come.

They could be providing fibre, say, from Glenthorn to jump back to Dublin effectively.

Mr. Pat Kidney

No, they are not. It would not be that extensive. The Deputy should think of it as a central point, be it an exchange or a POP. Around that, there is normally a town or village and on the outside of that doughnut, as it were, there is the area with 300,000 homes, or part of that, which is served by Eir. It is not ubiquitous. Outside that, there is the area that is served by that exchange. The inner third is built.

Mr. Kidney said that most of the exchanges were part of that area with the 300,000 homes.

Mr. Pat Kidney

No, the area with the 300,000 homes would surround the exchanges. That area is outside the town that the exchange is in. That is obvious if one looks at the map with the light blue colouring on which it has been marked.

The bidder would be traversing the area of 300,000 homes served by Eir to get to the 540,000 homes. Dr. Yardley's point is that the bidder will not fully duplicate in the area of 300,000 homes.

Dr. Matt Yardley

Yes. I was conscious that one of the Deputies may have misunderstood that. We should be clear. On the value question, we are not passing all the homes. That is not what is happening.

Will the ducting and poles be used in that traverse?

Dr. Matt Yardley

Yes. That is correct.

Some of them will be used.

Dr. Matt Yardley

Yes.

Is it a large number?

Dr. Matt Yardley

We are not using every single pole past every single home.

Mr. Pat Kidney

If one thinks of the main road out of a particular town, the network would go out along that and when it got to the edge of the area with the 300,000 homes it would start to go left and right and then span out.

I thank Dr. Yardley and Mr. Kidney for their input and time this afternoon.

The joint committee adjourned at 6.35 p.m. until 2 p.m. on Thursday, 13 June 2019.