Response to
OPTA Consultation Document on
Interconnecting Leased Lines
EXECUTIVE SUMMARY AND INTRODUCTION
COLT, Energis, Versatel and WorldCom welcome OPTA's Consultation Document on Interconnecting Leased Lines (reference: OPTA/IBT/2002/2000027) that outlines OPTA's provisional judgement of KPN's Reference Offer for Interconnecting Leased Lines (ILL-RO).
We are happy that OPTA on the basis of the outcome of the market consultation between July and September 2001 concerning possible approaches to the future regulation of KPN's tariffs for leased lines has decided to give priority to the implementation of ILL in The
Netherlands. As we have argued during the consultation a suitable ILL offer is required to tackle the problem that due to a lack of competition tariffs for leased lines in The
Netherlands are excessive, and - as far as 2 Mbit/s is concerned - are the highest in the European Union1. We fully believe that if KPN were to make ILLs available under
cost-oriented, non-discriminatory and transparent conditions, competition in leased lines and downstream markets, were new entrant use leased lines a basis, would substantially increase and, as a result, tariffs would automatically fall to a more competitive level. We want to emphasize here that there will only be a level playing field if all players including KPN retail are confronted with the same conditions in terms of costs for and quality of access lines.
The comments included in this paper focus primarily on the discussion in the Consultation Document. Generally we agree with the analysis and preliminary conclusions contained in the Consultation Document.
Our key messages are:
• In our view there are four key principles that OPTA should use in evaluating KPN's reference offer for ILL and to prevent KPN from introducing any so-called "deal killers" which would prevent OLO’s from using ILL:
1. The principle of duplicability: the ILL offer should allow OLO’s to duplicate all retail services provided by the incumbent (all distances, bandwidth, technical conditions). According to us this will only work when the incumbent’s retail business is also forced to “buy” access lines from the wholesale services portfolio and will have no ways to bypass this regime.
2. The principle of efficiency: the ILL conditions should always improve the best retail offer both in prices and in SLA due to efficiencies derived from the contracting.
3. The principle of no unduly burdensome restrictions: The incumbent should have the burden of proof in relation to the inclusion or modification in the ILL-RO of any
1.
The fact that leased line tariffs in The Netherlands are excessive is also recognized by the European Commission in its 7th Implementation Report on the Implementation of the Telecommunications Regulatory Package.
restriction, conditions or penalties not defined in the retail leased line offer and/or not applicable to the incumbent itself
4. The principle of continuity: the leading principle on migration should be that no payment should be required in order to migrate to the new regime if the service is already being provided and the migration requires no significant technical
modification
• We agree with OPTA that on the basis of the European Interconnection Directive and the Dutch Telecommunications Act KPN is legally obliged to provide ILL as an interconnection service. Therefore KPN must provide such ILL services on non-discriminatory terms and at unbundled, transparent and cost-oriented prices.
• We agree with OPTA's definition of an ILL, being a partial private circuit between a customer location and a point of interconnection. Versatel wants to point out that it should also be possible to use ILL’s to fill a missing link in OLO’s access network (from MDF to PoP).
• We agree with OPTA that it must be possible to use ILL for leased lines and every possible down-stream service and that the current user restrictions must be removed from KPN's ILL-RO.
• We agree with OPTA 's preliminary viewpoint on the ILL capacities and the network levels. Indeed, there is no need for analogue ILL and we agree that under the current circumstances it would be inefficient to buy structured nx64kbit/s circuits at LAP-level (but this issue needs to be addressed in the future). It is also essential that ILLs at all speeds above 2 Mbit/s at LAP and RAP level are included. We agree that ILL at NAP-level is not needed.
• We agree with OPTA's viewpoint that it is unreasonable that ILL can only be bought in combination with the collocation service. In many cases this requirement
unnecessarily raises the costs for OLO’s.
• To protect the investments that OLO’s have already made in interconnects for switched services with KPN it is required that the KPN Cluster Points are in the same locations as the locations that are currently being used for the other
interconnects (RAPs). If new locations for Cluster Points are chosen then this will lead to massive destruction of capital and hence potentially higher end-user tariffs.
• OLO’s should be allowed to share carrier systems. Without this ability to mix traffic, OLO’s would be forced to support a full STM-1/4/16 to support each type of traffic., which is not efficient and would raise OLO’s costs. If the equipment and
• We agree with OPTA that with regard to the capacity of POC interfaces the ILL offer must also include the 622 Mbit/s (STM-4) and 2.5 Gbit/s (STM-16) capacities as well as the 155 Mbit/s (STM-1) capacity. The tariffs for these services should be cost-oriented. We also agree on OPTA viewpoint that KPN's restrictions on interfaces result in raising the costs of rival OLO’s. The imposed restriction of unbundling into individual circuits by KPN and rebundling by OLO entails an unnecessary cost increase and hence a restriction on the use of ILLs. We believe that this restriction must be lifted.
• We fully agree with OPTA that service levels form an essential part of the ILL product, and that parties up-front need adequate service levels that are identifiable and guaranteed and in addition sanctioned by uncapped penalty payments that do not prejudice OLO’s rights to claim damages. Without such service levels, it is
possible for KPN to reduce the usability of the ILL service. We agree with OPTA that based on the principle of non-discrimination, purchasers of ILLs must be able to offer a comparable quality based on 2*ILL as KPN is offering in its retail services. In
addition to the above, it is especially important that OPTA obliges KPN to report on all relevant KPI's to OPTA, so OPTA can monitor and ensure that KPN's
performance is in line with the SLA and to ensure that KPN is not discriminating.
• We agree with OPTA that it would be unreasonable if buyers of ILL would have to pay migration costs if no new physical connection is involved at the access locations or POC locations and therefore only a logical (and not a physical) migration to a new service is involved. In that case the migration should be free of any additional
connection fees and/or migration fees. Also no migration costs should be allowed in cases where migrations take place when KPN was initially asked to offer an ILL service, to be terminated on OLO’s carrier system, which OLO had already in place and where KPN was only willing to offer a bundled service.
• We also agree with OPTA that it would not be reasonable if KPN would charge costs for early termination of current retail contracts if these were replaced by ILL
contracts. We believe OPTA should decide that KPN is obliged to retroactively refund OLO’s all the costs that OLO’s have paid in excess as a result of KPN not providing ILL as an interconnection service.
• We fully agree with OPTA that the forecasting obligations must be removed. The forecasting requirements as proposed by KPN are unreasonable, burdensome and discriminatory.
comments to OPTA as soon as possible.
• Energis, COLT and WorldCom are extremely worried by the fact that OPTA intends to use the EDC-model to determine ILL tariffs. We believe that a key requirement (also to avoid price squeeze) is that ILL tariffs are determined on the basis of BU-LRIC, instead of EDC. KPN's EDC-model results in tariffs far above cost, thereby placing competitors at a cost disadvantage, but also increasing risks of price
squeeze. the above mentioned OLO’s strongly recommend OPTA to reconsider the use of EDC. Versatel has a different opinion on this matter (see Versatel’s separate letter on this issue)
• At the least OPTA should ensure that ILL tariffs in The Netherlands fall below the (corrected) price ceilings as recommended by the European Commission. Therefore, the EU price ceiling date 1999 should be corrected with minus 30% to bring it up-to-date and reflect erosion.
• We welcome that OPTA is paying special attention to the problem of price squeeze. We agree with OPTA's description of possible causes of price squeeze. In general we believe the most important cause of price squeeze to be that inputs (like ILL) are priced above the level of the underlying costs and that according to Energis, COLT and
WorldCom OPTA needs to use a BU-LRIC model, instead of EDC to determine ILL tariffs. It is important that all potential causes of price squeeze are addressed in OPTA's upcoming final decision on ILL, which in sum means that ILL tariffs must be unbundled and according to Energis, COLT and WorldCom based on BU-LRIC and that costs must be allocated properly. Furthermore KPN must be obliged to keep separate accounts and it must be ensured that tariffs are non-discriminatory.
However, we strongly believe that even if these issues are dealt with properly in the final OPTA decision, a separate price squeeze test is still needed as a safety-measure. We recommend OPTA to initiate a separate proceeding for the
development of such a price squeeze test. On the matter of methodology for cost price calculations Versatel has a different opinion (see Versatel’s separate letter on this issue)
PROBLEM DEFINITION
1. The Board is keen to ascertain whether interested parties agree with the definition of the problem outlined in this section.
We agree with OPTA’s problem definition:
• Relatively high retail tariffs for leased lines are the result of lack of competition
• Lack of competition is the result of inability of OLO’s to use KPN’s access network
• Inability of OLO’s to use KPN’s access network is due to lack of non-discriminatory, transparent, cost-oriented KPN ILL offer
• Due to lack of non-discriminatory, transparent, cost-oriented KPN ILL offer, OLO’s can not compete on other downstream markets
• Due to lack of non-discriminatory, transparent, cost-oriented KPN ILL offer, OLO’s are significantly price squeezed, having to buy retail circuits from KPN or being forced to buy two KPN ILL circuits
• KPN is effectively discriminating OLO’s in this respect
KPN is the sole owner of a very dense infrastructure (combining backbone and access facilities). Currently it is impossible to compete with KPN on the leased lines market and hence on other downstream markets. Other operators in effect still have to purchase retail services from KPN. This is due to the fact that the current KPN offer for
Interconnecting Leased Lines does not fit the need for access lines of its interconnect customers (i.e. operators with a registration to provide public telecommunication networks and services and leased lines). Any supplier other than KPN will, pursuant to the KPN ILL offer, have to buy two lines from the KPN access network to be able to deliver an end-to-end solution to the customer, where KPN only has to buy one line from itself. This means first that KPN buys from itself at better conditions than what is possible for KPN’s wholesale customers, which is discriminatory. Second, it means that a price squeeze can occur when KPN abuses its own lower cost base to determine its retail leased lines prices at less then double the price for its access lines plus a certain retail markup. We therefore fully agree with OPTA's observations in paragraphs 23 and 24 of the consultation document that the current pricing structure for leased lines effectively results in a price squeeze, which results in an inability of other OLO’s to effectively compete with KPN and prevents the development of competition on the market for leased lines and downstream services in The Netherlands.
We agree with OPTA that high retail tariffs for leased lines are the result of market failure, namely the absence of a cost-oriented, non-discriminatory and transparent offer for ILL, which is best attacked at the source. Such an ILL offer would allow for
competition to develop in the market for leased lines and other downstream markets for telephony and data communication services.
between retail and wholesale market segments, whereby the latter (wholesale) market should be further differentiated between access (or terminating) leased lines and trunk segment (or backbone). We believe that a market definition as proposed by OFTEL would be most appropriate (see DRAFT Direction pursuant to Regulations 6(6) of the Telecommunications (Interconnection) Regulations 1997 of 17 December 2001. Here OFTEL states that “At the wholesale level, Oftel still considers it appropriate to define separate markets for terminating and trunk segments. However, Oftel now believes that there are two separate terminating segment markets. The three wholesale markets as defined by Oftel are:
• a market for terminating segments up to and including 2Mbit/s (low bandwidth); • a market for terminating segments above 2Mbit/s (high bandwidth); and
• a market for trunk segments.”
We believe that this definition describes the market for wholesale leased lines most appropriately. The terminating segment is then similar to what OPTA describes in Figure 1 as the “access segment.” We urge OPTA to adapt this definition to ensure that ex-ante regulation has the right legal basis, where market definition and assessment of competition precede determinations. The market for terminating segments is not
competitive, whereas the market for trunk segments is competitive.
Such a market analysis is also important given OPTA's ambition to intervene only if the market fails to provide an adequate alternative for ILL. We believe the SMP
designations and the underlying market analysis as such are not sufficient, because the SMP analysis could potentially lead to the wrong conclusion that an ILL product for leased lines of >2Mit/s is not needed. In fact OPTA itself also recognizes this in paragraph 59 of the consultation document, where it further divides the market into a market for trunk segments (backbone) and a market for leased lines to access
THE SOLUTION: A SUITABLE ILL PRODUCT
2. The Board is keen to ascertain whether interested parties agree with the demand for leased half-lines outlined in this section.
We agree with OPTA’s analysis that the problem, or at least part of it, is caused by the structural discrepancy between what providers require and what KPN offers them and that the solution is to oblige KPN to provide a service that meets the market's demand for access services. Indeed, what the market needs is in fact a leased half-line or Partial Private Circuit (PPC) from the core of KPN's transmission network to the edge of this network, where the customer is located. Since the identified market failure relates both to leased lines and to the associated downstream services, it must be possible to use these leased half-lines for every possible service. Any attempts of KPN to restrict the use of PPCs are not acceptable (see in more detail below under question 5,7 and 8). We fully agree with OPTA that the implementation of such ILL is most urgent tot stimulate effective competition.
We confirm our large need for (PoP-to-end) access lines for all possible services to prevent the problems as described in the response on question 1. To be able to
compete with KPN in leased lines and other downstream markets, a cost-oriented, non-discriminatory and unbundled ILL offer that includes access lines is much needed. We strongly support OPTA’s analysis here.
3. The Board is keen to ascertain whether interested parties agree that ILLs can also be used to connect base stations.
The ILL offer has the sole purpose of increasing competition in the leased lines and other downstream markets. Connection of base stations does not fit this definition, and therefore we feel the ILL offer should not be used for this. Base station connections are needed by mobile operators that are in principle in a different business than fixed operators. Only fixed operators i.e. those operators with a registration to provide public telecommunication networks and services and leased lines should be eligible to receive an ILL offer from KPN. Versatel wants to point out that ILL services should be used as part of a fixed OLO’s access network which includes that ILLs can be used as a
LEGAL BASIS
4. The Board is keen to ascertain whether interested parties agree with the assessment of the legal framework outlined in this section and if not, why not.
We fully agree with OPTA's analysis of the legal framework and that ILL is an interconnection service that KPN is obliged to provide and that KPN's ILL service is governed by the general conditions that are applicable to interconnection services, i.e. cost-orientation, transparency, non-discrimination, accounting separation, etc.
Indeed, for ILL the relevant European directive is the Interconnection Directive (ICD) that is implemented in particular in Section 6 of the Dutch Telecommunications Act. Article 2.1(a) of the ICD defines "interconnection" as:
"The physical and logical linking of telecommunications networks used by the same or a different organization in order to allow the users of one organization to communicate with users of the same or another organization, or to access services provided by another organization. Services may be provided by the parties involved or other parties who have access to the network."
The crux of this definition is that networks are linked in order to communicate or to gain access to services. Since ILL is a service consisting of the provisioning of access capacity from a customer end to an OLO point of interconnection, the OLO's network will be physically and logically linked to KPN's network at the Point of Interconnection. Or to put it in other words, when an ILL is supplied, both networks are connected to enable communication or access to services. We agree with OPTA that, consequently, at least one of the two ends of an individual ILL must always be located in a connection location, i.e. a location of KPN or of the provider requesting interconnection.
Pursuant to Articles 6.4 and 7.2 of the Telecoms Act, KPN has been designated by OPTA as SMP provider on the leased lines market. The Article 6.4 SMP designation means that KPN is obliged to provide ILL under the following conditions:
• non-discriminatory conditions (Art. 6.5 a and b)
• supply on request all necessary interconnection information (Art. 6.5 c) and use the supplied information for the purpose for which it was supplied only (Art. 6.5 d)
• transparent and cost-oriented interconnection tariffs (Art. 6.6, paragraph 1)
• sufficiently unbundled interconnection tariffs (Art. 6.6, paragraph 2)
• establish a cost accounting system approved by OPTA (Art. 6.6, paragraph 3)
• publish a reference interconnection offer (Art. 6.7)
• keep separate accounts for interconnection activities and other activities (Art. 6.8)
In addition to being regulated by Art. 6.4-6.9 of the Telecoms Act, OPTA also mentions that the SMP designation of KPN pursuant to Art. 7.2 of the Telecoms Act is governed by the obligations set out in BOHT in relation to leased lines. In our opinion leased lines supplied by KPN with a capacity of up to 2 Mbit/s and PVCs at least, are regulated by BOHT.
We also agree with OPTA that the wording of the Guidelines for interconnection that were published by the Minister in 1997 needs to be changed to avoid any
RESTRICTIONS ON USE
5. The Board is keen to ascertain the views of interested parties concerning restrictions on use, and the reasons for these views.
We agree with OPTA that it must be possible to use ILL for every possible service. In other words, any possible restriction whereby the ILL can only be used to supply leased lines to third parties should be rejected. The relevant restrictions referred to in the ILL-RO have to be removed. More specifically, this means, for example, that 'provision 1.2.2. v' in the Service Description section of the ILL-RO must be deleted.
The restrictions with respect to the use of ILL purely for end-to-end leased lines to end users at more than one location with administrative ties to KPN, prevents OLO’s from using the ILL offer to offer leased lines and other downstream services to customers with just one location.
CAPACITY AND NETWORK LEVELS
6. The Board is keen to ascertain the views of interested parties concerning the
capacities and network levels where ILL services must be available. Please make a distinction between your views on the presence or absence of alternatives, i.e. the level of competition, on the one hand, and your own requirements for specific
services on the other. In terms of the alternatives, please specify why you feel there are no viable alternatives. In terms of requirements, please specify exactly what your company's plans are for connection at specific network levels if an ILL service is offered. To what extent do you expect your company to purchase ILL services? Specify what tariff differences between NAP, RAP and LAP level would cause your company to connect at specific levels.
Below we discuss OPTA's preliminary judgement in more detail:
• Analogue ILLs at all network levels.
We agree with OPTA's arguments that the ILL offer does not need to include analogue leased lines because of a lack of market demand.
• ILLs at NAP level.
We agree with OPTA that there is no market need to provide ILL at the NAP-level. Indeed, several OLO's have already rolled out their networks between RAP
locations, which means that there is a viable alternative for ILL at NAP level available.
• 64 Kbit/s, nx64 Kbit/s and structured 2 Mbit/s ILL at LAP level.
Apparently this category of structured leased lines is always routed at RAP level, because KPN has installed its cross connects at the RAP level and not at the LAP level. Since this results in the strange situation where tariffs for ILL at the LAP level are higher than tariffs for ILL at the RAP level, we would agree with OPTA that interconnection at LAP level for ILL is currently fully inefficient. However, we believe OPTA needs to address this issue in the longer term, whereby the leading principle should be that ILL at the LAP-level must always be cheaper than ILL at the RAP-level. According to Energis, COLT and WorldCom one way to solve the problem would be to use a BU-LRIC cost-model, instead of EDC, to determine the ILL tariffs. Of course the situation needs to change as soon as KPN starts installing cross connects at the LAP level.
• 64 Kbit/s, nx64 kbit/s and structured and unstructured 2 Mbit/s ILL at RAP
level.
We fully agree with OPTA that there are no adequate alternatives for ILL at the RAP level. Indeed, with the exception of KPN, no provider has rolled out on a large scale to LAP level. Consequently, KPN is the only provider with national access
• 2 Mbit/s unstructured digital leased lines at LAP level
We agree with OPTA that indeed, there are no viable alternatives available. Because the cross connect issue that applies for structured ILLs at the local level does not apply in this case, 2 Mbit/s unstructured should also be provided at the LAP-level.
• ILLs >2Mbit/s at LAP and RAP level.
We agree with OPTA that there are no adequate alternatives at either LAP or RAP level for ILLs with a capacity above 2Mbit/s to access locations. We fully agree with OPTA that although KPN has a relatively small market share on the market for these lines (if trunk capacity and access lines are included), KPN is still the dominant provider for access lines with a capacity above 2 Mbit/s at LAP and RAP level and that KPN should be obliged to provide all types of ILL with a capacity above 2 Mbit/s (e.g. 34 Mbit/s, 45 Mbit/s, 140 Mbit/s, 155 Mbit/s, 622 Mbit/s) as well.
• PVCs at all network levels
We agree that in principle OLO’s requiring PVC access services can provide these services themselves in the same way as KPN, by using ILLs with fixed capacity. We also agree that this situation would change if KPN stopped using fixed-capacity leased lines when supplying PVCs. In this case, OPTA would need to reconsider the extent to which alternatives to such a service are available.
The Table below summarizes our viewpoint (the blue indicating what we believe should be included in the KPN ILL product), which coincides with OPTA's preliminary
judgement.
Capacity/ Network level
Analogue 64 Kbit/s nx64 Kbit/s 2 Mbit/s structured
2 Mbit/s unstructured
> 2Mbit/s PVC
See Table 1 < 2 Mbit/s 2 Mbit/s > 2Mbit/s PVC
NAP no no ILL-RO no no ILL-RO no ILL-RO no ILL-RO no RAP no yes ILL-RO yes yes ILL-RO yes ILL-RO yes ILL-RO no LAP no no ILL-RO no no ILL-RO yes ILL-RO yes ILL-RO no
LAP levels have less priority (especially if not coinciding with RAP locations), as (most of) the competing suppliers of leased lines are interconnected with KPN at the RAP level. In any case, we find it very important to ensure that all of KPN’s cost components are allocated to the right level of interconnection. For example, a supplier that is
KPN has to make for interconnection at the LAP level. This is also the case for voice interconnection, where different charges apply for, e.g., regional and local ports. It is sometimes argued that SDSL is an alternative for 2 Mbit/s leased lines. Although this could potentially be true from a pure technical perspective, we would like to underline that we believe that under the current circumstances SDSL will not a viable alternative in the foreseeable future. Firstly, currently there are no SDSL product offerings available yet. Secondly, if SDSL would become available, it would only be an alternative for lines that are shorter than 2,5 - 3,0 kilometers. Thirdly, the geographical coverage of SDSL, if it would become available, will only be very limited in scope. Fourthly, quality of service for SDSL will potentially be a problem, mainly due to the fact that KPN does not provide satisfactory quality of service guarantees for unbundled copper lines. And lastly, SDSL as a service of course largely depends on KPN's service conditions for the underlying ULL service, that - as OPTA is aware - are still far from what they should be, which also negatively affects the potential future availability and conditions for SDSL.
INTERCONNECTION LOCATIONS
7. The Board is keen to ascertain the views of parties regarding interconnection locations.
We agree with OPTA's statement in paragraph 64 of the consultation document that it is unreasonable that ILL services are only offered in combination with collocation services. Firstly, it may be that providers already purchase the Network Interconnection Service (NIS) for (telephony) interconnection and therefore already have transmission capacity between the KPN location an their own sites. Secondly, it is possible that providers already have equipment in the KPN building without purchasing collocation. This occurs, for example, in the case of reciprocal interconnection for telephony services between KPN and a provider. Thirdly, it must also be possible to interconnect via other providers that are already collocated at the KPN site. In all of these cases – and
perhaps even more – the ILL service may be delivered at the KPN site without requiring collocation. Collocation would in these cases unnecessarily increase the costs of using ILLs.
Unbundling KPN ILL service
KPN and customer site interconnect
We maintain that it should be the decision of the interconnect customer whether interconnection takes place on the KPN or the interconnect customer location. Picking up the traffic at the so-called Cluster Points is likely to be the cheapest solution. Having KPN deliver the lines at an interconnect customer location will be the most expensive, unless a requirement for cost-orientation is placed on this service. We agree with OPTA’s statement under point 64, that it is unreasonable that ILL services are only offered in combination with collocation services for the reasons as mentioned. ILL interconnect at EVKC, also over switched services interconnect infrastructure To protect the investments that the interconnect customers have already made in interconnects for switched services with KPN it is required that the Cluster Points are in the same locations as those that are currently being used for the other interconnects (RAPs). This use would be supported by OPTA’s verdict of July 28, 2000 concerning the Reference Interconnect Agreement where KPN is obliged to allow operators to share the same carrier system, independent of whether it would be used for switched services interconnect or ILL interconnect. This view is also supported by the OFTEL decision of December 17, 2001. Without this ability to mix traffic, OLO’s would be forced to support a full STM-1/4/16 to support each type of traffic. This would mean that
existing and new infrastructure would not be used efficiently.
If the equipment and transmission, already in use for the switched services interconnect with KPN indeed can be used and where OLO's have placed equipment at the KPN EVKC, it will also be necessary that KPN provide at cost-oriented tariffs the so-called patch-through service to its ILL equipment. This patch-through is absolutely crucial in order for the ILL offer to be viable.
Shared carrier systems
Following the OPTA verdict of July 28, 2000 concerning the Reference Interconnect Agreement, KPN is obliged to allow operators to share carrier systems. An OLO could therefore use its installed equipment and transmission at either the KPN site or the OLO site to offer capacity on this infrastructure to another OLO for interconnect with KPN for either switched or transmission ILL services.
INTERFACES
8. The Board is keen to ascertain the views of parties regarding the interface specifications.
The Interconnecting Leased Line will have to be connected to the network of the interconnect customer and the end-user information will have to be transported over that network. Within the interconnect customer network all transmission is based on SDH technology, in which all line speeds from 2 Mbps upwards are dealt with within an STM1, STM4 or STM16 transmission container. The 2 Mbps circuit only becomes physically visible at the outer borders of the network. An Interconnecting Leased Line should be considered as part of the total transmission network and therefore the 2Mbps circuit should only become visible at the end-user location. It is a waste of money and effort to create another physical 2 Mbps connection at the Cluster Point. This money and effort could better be spent on increasing the competition in other areas. The interface does not necessarily need to be a fiber optic interface but it should at least be an STM1 transmission container. This goes for both structured and unstructured 2 Mbps.
The imposed restriction of unbundling into individual circuits by KPN and rebundling by OLO entails an unnecessary cost increase and hence a restriction on the use of ILLs. We believe that this restriction must be lifted.
SERVICE LEVELS
9. The Board is keen to ascertain the views of parties regarding service levels. The Board would like to receive specific, detailed proposals relating to service levels.
We fully agree with OPTA that service levels form an essential part of the ILL product, and that parties up-front need adequate service levels that are identifiable and
guaranteed. Without firm service levels, it is possible for KPN to reduce the usability of the ILL service. We agree with OPTA that based on the principle of non-discrimination, purchasers of ILLs must be provided with at least the same quality KPN delivers to itself internally for retail leased lines.
In our opinion the leading principle should be that the service levels for ILL should always be better, and at least be as good as, the quality guarantees for KPN's retail LL's. We want to add here that the service level an OLO can offer must at least be comparable to the service level of KPN’s retail service. This means that 2* ILL + backbone must offer the same quality to the end-user as a retail line. Therefore to create a level playing field the quality of a single ILL must be higher than KPN’s retail service (with the possible exception of unstructured 2 Mbit/s for which by definition a slightly lower quality is unavoidable).
We fully agree with OPTA's suggestion in the consultation document that service levels in principle should be guaranteed via two different - but complementary - routes.
1. The first route that is identified by OPTA is the route of the SLA that should be part of the ILL-RO2. Indeed, the SLA should include fixed, pre-defined quality levels for ordering and delivery, maintenance and repair and availability. We fully agree with OPTA that penalty payments, in case these levels are exceeded, must also be included in the SLA. However, the penalty system should be self-enforcing and deliberately be basic in form, to keep administration to a minimum. Further, it is important that the penalty payments are uncapped and do not prejudice OLO’s rights to claim damages. Another precondition is that KPN should never be allowed to allocate these penalty payments into the ILL tariffs. Otherwise, ILL buyers would end up paying the penalties themselves, in the form of higher tariff, which would of
course be unreasonable. More detailed proposals for what we believe should be part of the SLA for ILL can be found in Annex II.
2. The second complementary route is to govern the performance levels by
administrative law. In that context OPTA itself has a crucial role to play, primarily by ensuring that KPN is not discriminating and that KPN meets performance levels that are reasonable. An important precondition is that OPTA in advance makes clear what the performance levels are that OPTA expects KPN to meet and that OPTA monitors KPN's performance on a continuous basis3. In addition to a requirement for KPN to report to OLO's, KPN must be obliged to provide more aggregated reporting to OPTA to ensure that KPN is not discriminating against OLO’s in the provisioning of ILLs and to ensure KPN meets the pre-defined targets. Such reporting to OPTA would also play a preventative role by increasing visibility through a consistent and straight forward approach. These reports should be audited by a third-party to ensure accuracy, consistency and compliance. A dual reporting obligation (to the operators individually and to OPTA) with third party auditing would take a lot of the burden from the shoulders of OPTA. In case of under-performance and/or
discrimination OPTA should impose punitive administrative sanctions upon KPN. In paragraph 73 of the consultation document OPTA requests parties to specify their ideas on more specific and detailed proposals for quality guarantees. To meet this OPTA request we have specified what we believe should be included in the SLA for ILL in some more detail in ANNEX II. We strongly recommend OPTA to include these suggestions in the upcoming decision.
2.
OPTA mentions in paragraph 74 of the consultation document that the SLA will be governed by Civil Law. This is true as far as parties agree on the SLA and sign such an agreement with KPN. However, we would believe that OPTA has an important role to play by setting the minimum standards that must be included in KPN's SLA as part of KPN's ILL-RO. In other words, the upcoming OPTA decision on KPN's ILL-RO should also define the terms of the SLA.
3.
Service levels can roughly be divided in three types. Each of these three types of performance should be identified in the SLA and in the OPTA monitoring, and be measured and reported to both OPTA and the market parties on a monthly basis.
• Availability performance.
• Repair performance.
• Delivery performance.
Each of these types will be discussed briefly below. For further details we refer to Annex II.
Availability
In the ILL-RO Technical Manual (Section 4, pages 3-14), the quality of ILLs is defined with reference to three parameters: availability, errored seconds and severely errored seconds. The Table below includes the quality of service for different ILL types, according to the ILL-RO.
Capacity 64kbit/s, 2Mbit/s 2 Mbit/s (unstructured) VC-12 VC-3 VC-4 Availability >99.9% >99.2% >99.98% >99.98% >99.98% Errored seconds ≤0.8% ≤0.12% ≤0.74% ≤1.39% ≤2.96% Severely errored seconds ≤0.025% ≤0.055% ≤0.037% ≤0.037% ≤0.037%
We fully agree with OPTA that the low availability of 2 Mbit/s transparent leased lines as suggested by KPN (99.2%) should be rejected. We agree with OPTA's observation that for providers who use ILLs to provide downstream services to customers, the availability of ILLs likewise represents the upper limit of the availability of the downstream service that they themselves provide. Therefore the availability level of 99.2% as suggested by KPN is totally unacceptable. We strongly recommend OPTA to decide that the
guaranteed availability of 2 Mbit/s transparent ILLs must be increased to 99.98%, and not the 99.9% that is suggested by OPTA in the consultation document. As can be learned from the report of the European Commission4 to which OPTA also refers in the consultation KPN has realized an availability of 99.98% for 2 Mbit/s transparent leased lines. We therefore believe this level of 99.98% should also be included in the SLA. With regard to unstructured 64 kbit/s - 2 Mbit/s ILLs, KPN guarantees a performance level of >99.9%. Although this level is equal to the KPN performance guarantees for standard Digistream retail products, we believe the availability level should be >99,98%. This level coincides with the KPN performance level for its retail Digistream premium. If
4.
OPTA would decide in line with what it has suggested in the consultation document, and impose a level of >99.9%, we would recommend OPTA to oblige KPN to follow an identical approach for ILL as for Digistream. This means that KPN should make available a premium ILL product in addition to its standard ILL product, identical to KPN's differential approach for retail Digistream leased lines.
Repair
As we have already stated in the above KPN's performance levels for ILL should at least meet KPN's performance levels for retail leased lines. The same applies to penalty payments. We therefore broadly agree with OPTA's suggestion to apply KPN's current retail penalty clauses for fault duration to ILL (see the table below). However, if the fault duration takes more than 8 hours, in addition to the reimbursement the OLO should also have the right to terminate the contract, without paying any termination charges and without effectively being prevented to claim additional damages.
Availability Fault duration 99.9% 99.98% < 2 minutes 0 % 0 2 - 5 minutes 0 % 20 5 minutes - 2 hours 0 % 100% 2 - 4 hours 20 % 200% 4-8 hours 100% 200% > 8 hours 200 % 200%
Ordering and delivery
We are worried by the fact that OPTA is not paying special attention to the issue of ordering and delivery. This is especially worrying given KPN's under-performance in the near past. We strongly request OPTA to include special requirements for ordering and delivery in its final decision.
In the current ILL-RO (version 0.4) KPN proposes a contractual supply time of 25 working days in which 95% of the lines must be delivered. KPN's (uncapped) penalty scheme includes a payment of 50% of the monthly recurring fee if the line is delivered between 15 and 26 working days beyond the agreed Ready for Service Date, and a payment of 100% of the monthly recurring fee if the line is delivered 25 days or more beyond the agreed Ready for Service Date5. We believe this is not sufficient. The suggested SLA, in the form as it now stands, is wholly inadequate for the purpose of encouraging KPN to provide reliable and efficient services to its competitors. A number of fundamental structural flaws must be addressed.
We agree with OPTA that delivery times and related penalty clauses for delivery times at least must be equivalent to the delivery times and penalty clause for retail leased lines. However, as already argued in the above we also believe it to be reasonable that ILLs requires a more comprehensive SLA than KPNs retail LLs. The SLA should:
5.
• Define precisely the provisioning intervals to be measured;
• Specify reasonable time-scales for all types of order;
• Institute a precise reporting mechanism to carrier customers on delivery performance; and
• Institute a precise reporting mechanism to OPTA whereby performance with respect to KPN's retail arm versus its carrier customers can be monitored and checked for any discrimination.
Another of the shortcomings of KPN's SLA is the lack of any detailed description of processes and procedures. To ensure transparency of the order process, the contract must clearly articulate the exact procedures to be followed, with service levels
applicable to each step of procurement. Definitions should be provided as to how performance should be measured and reported by KPN on a timely basis and, in the case of default, how (uncapped) penalties should be applied. Currently, the document is not structured in this way. It is more focussed on exceptions and penalties with no clear description of the process to be followed. For example, KPN provides no insight into whether there is an order receipt acknowledgement, whether there is an OLO requested due date, whether there is a firm order confirmation date, or the availability of an
expedite procedure. As a result, the Service Commitment on KPN's part is essentially no commitment at all because it is fraught with so many loopholes.
Further details of how we believe the service levels should be defined can be found in Annex II. We are fully aware that determining on all the issues we raise in that
document will mean that KPN must amend its ILL-RO substantially. However, it is absolutely essential in the short-term that the service levels for ILLs are adequate and at the least no worse than those currently available for retail circuits, and we look to OPTA to include the requirements outlined in the above and in the Annex in the upcoming decision very shortly.
MIGRATION COSTS
10. The Board is keen to ascertain the views of parties regarding migration costs. The leading principle on migration costs should be that no (extra) payment should be required if the service is already provided by KPN (i.e. pre-existing circuits), and the migration requires no significant technical modification. Cost for migration from the current situation to the ILL situation can only be justified if these costs are the result of new requirements from the buyer of the service, such as a new end-user location, a new contract duration or a modification in line speed. If the costs are solely the result of the ILL offer becoming available these costs should not be charged.
connection fees and/or migration fees. Migration fees potentially can only be applicable in case a new physical connection is involved. However, since there could also be some physical handling required to migrate from one interconnection point (current MCTN-offer) to another, we believe that also in cases where migrations need to take place that are the result of a forced use of a network interconnection point and service (for
instance because other alternatives would cause higher costs for OLO) there should be no charge for migrations.
We also agree with OPTA that it would not be reasonable if KPN would charge costs for early termination of current retail contracts if these were replaced by ILL contracts. Indeed, costs of this type could be justified only if the contractual relationship with KPN is totally terminated. This does not apply in to the migration from a KPN retail service to ILL, where one KPN contract is merely replaced by another KPN contract.
Furthermore, KPN has hitherto refused and delayed its ILL product-offer, so that other providers have always been obliged to make use of (the extremely expensive) retail offer. This should in principle result in an OPTA decision in which KPN is obliged to retroactively refund OLO’s all the costs that OLO’s have paid in excess during the last couple of years, due to the fact that KPN, against OPTA’s formal decision, has
FORECASTING
11. The Board is keen to ascertain the views of parties regarding forecasting.
We fully agree with OPTA that the forecasting obligations that are set out Section 5 of the Operations and Maintenance Manual of the ILL-RO must be completely removed from the ILL-RO, because such a requirement would be a huge obstacle to the use of ILLs.
We believe that the introduction of ILL will not change the total market but will merely create a different level playing field. KPN has a good deal of experience in serving this total market. Shifts to other operators do not necessarily create a bigger requirement for forecasting information for KPN. We believe a forecasting requirement would be
unreasonable, burdensome and discriminatory. Firstly, we are currently not required to provide forecasting under the retail leased line regime. We can order retail leased lines without any forecasting. This illustrates that KPN is very well capable of making its predications and planning of market demand without any forecasting. We do not see any reasons why this would be different with ILL. Secondly, if forecasting would be required we would be forced to undertake a disproportional amount of additional work to ensure that forecasts can be made, something that will also be quite difficult because it is almost impossible to say in advance where, when and how many new ILLs are needed. Thirdly, providing KPN with forecasts will give KPN an unfair competitive advantage since it will have visibility of our network planning, and insight in our customers. Lastly, the forecasting requirements would be discriminatory since these same requirements would not be applicable on KPN's own LL business and would also be discriminatory vis-à-vis retail KPN LL products, that do not require forecasting.
INTERNATIONAL ILL
12. The Board is keen to ascertain the views of parties regarding the requirement in the market for this service and the content of the offer.
OTHER ISSUES
13. The Board is keen to ascertain the views of parties regarding any aspects not dealt with in the preceding paragraphs. The Board requests that you distinguish between content-related and textual points and, as far as possible, refer to the paragraph numbering of the KPN ILL-RO.
In the preceding paragraphs we have commented on the main issues as identified by OPTA. In addition we have included our further and more detailed comments on the integral ILL-RO (version 0.4) in ANNEX I.
PRICE SQUEEZE ISSUES
14. The Board is keen to ascertain the views of parties regarding the issues dealt with in this paragraph. If parties believe that a different squeeze test is necessary, they are
requested to indicate as specifically as possible how a test of this type would be carried out. We welcome that OPTA is paying special attention to the problem of price squeeze. We agree with OPTA's description of possible causes of price squeeze and the need to find adequate remedies for each of these causes in the context of OPTA's final assessment of KPN's ILL product offer.
The most important cause of price squeeze is that inputs are priced above the level of the underlying costs. This is why the key remedy against price squeeze is that OPTA must ensure that tariffs for ILL truly reflect underlying costs. We therefore fully agree with OPTA that it is essential that the ILL tariffs proposed by KPN are carefully assessed by OPTA in terms of cost orientation. We also agree with OPTA that if it would appear that ILL involves "wholesale specific costs", these costs will be treated in accordance with OPTA policy for voice interconnection services, meaning these costs should be proportionally allocated not only to OLO’s, but to KPN as well. However, we would like to underline that cost orientation not only involves cost allocation issues, but - even more important - involves making the right choice for the cost-model to be used. As we will argue in more detail below when we discuss the tariffs as proposed by KPN, Energis, COLT and WorldCom strongly
recommend OPTA to use a BU-LRIC model instead of EDC to determine the tariffs for ILL. In fact, we believe that it is OPTA's commitment to the use of the EDC cost-model that makes The Netherlands so sensitive for prices squeeze. Versatel has a different opinion on the method of cost price calculations (see Versatel’s separate letter on this issue)
sufficient. In addition OPTA should also introduce element based charging for the different ILL elements in line with the picture below
It is especially important the MUX at the customer-end and the POP-end is treated as a separate unbundled element. The pricing of the MUX should reflect the underlying cost structure, meaning that tariffs should reflect economies of scale.
Further, and in line with OPTA's recent consultation on integral tariff regulation for end-users and interconnection services6, we strongly recommend OPTA to oblige KPN to keep separate accounts for its retail and wholesale LL business and all of its
downstream retail services. For further details we refer to the ACT response to the OPTA consultation7.
We agree with OPTA that discrimination, in terms of tariffs, can also be a cause of price squeeze. If KPN internally provides leased lines under different conditions (i.e. lower tariffs or better quality) than the tariffs that apply for OLO’s, OLO’s are potentially
squeezed out of the downstream market by KPN. We agree with OPTA that it therefore must be ensured that KPN does not apply different conditions and tariffs to itself
(internal delivery) as to OLO’s. However, a non-discrimination requirement as such would not be sufficient to solve price squeeze issue. As said in the above the main remedy against price squeeze is a strict requirement for cost-orientation (by which we mean BU-LRIC). [Please note that Versatel has a different opinion on the need to use of BU-LRIC] Without a requirement for cost orientation the vertically integrated KPN would still be able to leverage it position by cross-subsidization, even if a non-discrimination requirement would be installed. When identifying the legal basis for the
non-discrimination requirement OPTA refers to Article 14 of BOHT. However, the obligation for non-discrimination is also unambiguously stated in Article 6.5 of the Telecoms Act. This Article should also be applicable.
Indeed, an additional potential cause of squeeze is that parties are incapable of making sufficient margin between the tariffs for ILL services and the downstream services that are based on these ILL building blocks. As we have argued in the above, all potential
6.
OPTA consultation document on integral tariff regulation for end-users and interconnection services, 26 November 2001.
7.
causes of price squeeze should be addressed in OPTA's final decision on ILL, which in sum means that tariffs must be unbundled and BU-LRIC-based and that costs must be allocated properly. Furthermore KPN must be obliged to keep separate accounts and it must be ensured that tariffs are non-discriminatory.
However, we strongly believe that even if these issues are dealt with properly in the final OPTA decision, a separate price squeeze test is still needed as a safety-measure, especially if OPTA sticks to the idea that ILL tariffs must be determined on the basis of EDC (which we - with the exception of Versatel - would strongly reject particularly because we believe the EDC model is one of the main causes for price squeeze in The Netherlands. Therefore we agree with OPTA, that in line with the existing price squeeze tests for voice services, a separate price squeeze test for LL needs to be developed directly after the final OPTA assessment of the ILL offer.
In the consultation document OPTA identifies the potential building blocks of a price squeeze test as:
Service Cost elements
ILL-LAP Embedded Direct Costs (TO+ LAP customer + LAP-POC) ILL-RAP Embedded Direct Costs (TO+ LAP customer + LAP +
RAP-POC)
ILL-NAP Embedded Direct Costs (TO + LAP customer + RAP-RAP + NAP-POC)
KPN retail leased line via LAP Fully Allocated Cost (TO + 2xLAP customer)
KPN retail leased line via RAP Fully Allocated Cost (TO + 2xLAP customer + 2xRAP-LAP) KPN retail leased line via NAP Fully Allocated Cost (TO + 2xLAP + 2xRAP-LAP + RAP-RAP) KPN retail downstream service ILL service + surcharge for specific costs for downstream
services
OPTA further states that two important aspects must be taken into consideration when defining a squeeze test. The first concerns the level at which the test is performed, in other words, on the basis of which (collection of) ILL tariff elements on the one hand, and which (collection of) retail leased line tariff elements on the other hand is the margin assessed. The second important aspect concerns the level of the margin, referred to as the retail surcharge. OPTA also states that any potential squeeze tests do not apply to retail services on markets on which KPN has no SMP, for example the current market for retail >2 Mbit/s leased lines.
We do not agree that a potential price squeeze test could not be applicable to retail (i.e. downstream) services where KPN does not have SMP and/or is not dominant. In other words, SMP and/or dominance on the downstream market are no preconditions for a price squeeze test. We believe that based on competition policy any price squeeze test can be developed for situations where a dominant position on an upstream market can potentially be leveraged at a downstream market. We therefore request OPTA to
prices by putting separate prices for access lines and services in their offers to the market.
GEOGRAPHICAL DIFFERENTIATION
15. The Board is keen to ascertain the views of parties regarding geographical tariff differentiation.
OPTA states that it is conceivable that the costs for ILL may be dependent not only on factors such as length and capacity, but also on the region where the connection points (or interconnection points) are located. This would then raise the question whether such differences in costs, if they can be proven by KPN, should result in geographical tariff differentiation.
Even in case there would be differential underlying costs for some regions (which indeed needs to be proven by KPN), we would strongly reject geographical
differentiation.
TARIFFS
KPN's tariff schedule is not complete
Table 11 of the Consultation document - projected below - sets out the ILL tariffs as proposed by KPN.
LAP RAP NAP
64 kbit/s € 130 € 123 € 157
2 Mbit/s € 620 € 691 € 1,162
VC12 € 2,818 € 3,287 € 3,621
VC3 34/45Mb € 2,818 € 6,000 € 10,139 VC4155Mb € 2,818 € 8,910 € 17,414
First thing we would like to point out is that KPN's tariff schedule apparently is far from complete. We feel the following tariff components should also be part of KPN's proposal (and be judged by OPTA):
• the table includes the monthly recurring tariffs, but the non-recurring tariffs are missing. We would like to underline that non-recurring tariffs can be a "show-stopper" or "deal-killer" if these cost-components are not fully cost-oriented;
• a couple of speeds are not included in the table. More specifically, all possible speeds between 64 kbit/s and 2 Mbit/s are missing;
• also missing are the (non-recurring) tariffs for moving, dislocation and cancellation of ILL as well as upgrades or downgrades of ILL;
• in the consultation document OPTA mentions that the table tariff for 2 Mbit/s involves both structured and unstructured 2 Mbit/s and that the costs of structured are higher than unstructured. We believe both tariffs should be included on an individual
(separate) basis;
• and the last issue is that discount-schemes are missing. We believe the discount schemes should at least be equal to the discount schemes that apply to retail LLs. Itshould be prevented however that a discount scheme would be designed that solely benefits the incumbent provider or its subsidiaries as a large party taking up the ILL offer, as is for instance the case with the current offer of KPN Carrier Services for their Premium Transit services.
We believe OPTA should request KPN to ensure the tariff schedule is complete and to publish the whole tariff schedule, to allow market parties to send their comments to OPTA, so OPTA can take these comments into considerations in its judgment of the ILL tariffs.
KPN's tariffs do not reflect underlying costs
EDC-model and that the tariffs should not exceed the price ceilings that are recommended by the European Commission.
As we have already pointed out before, we strongly believe that ILL tariffs should be determined on the basis of a BU-BU-LRIC model. It is crucial to ensure that ILL tariffs reflect the forward-looking costs of an efficient operator in order to provide the right incentives to encourage productive, allocative and dynamic efficiency and to prevent price squeeze to take place. These costs may include an appropriate cost of capital employed and a "fair" contribution to common costs. OPTA's approach to determine costs for ILL (as well as other interconnection services, like originating services) on the basis of the EDC-model is not meeting these objectives and should be rejected8. Energis, COLT and Worldcom hereby again strongly request OPTA to reconsider the "Guidelines for Tariff regulation for interconnection and special access services" and to amend the Guidelines to be able to determine ILL tariffs on the basis of a BU-LRIC model.
Versatel is not agreeing with BU-LRIC. (See Versatel’s separate letter on this issue) However, would OPTA decide to use the EDC-model, it should use other instruments as well to ensure that the EDC-model does not result in unjustifiable high tariffs that have no relation to KPN's underlying costs (which seems to be the case if we look at the prices as they are proposed by KPN). These instruments include:
First of all it is important that OPTA re-introduces a bottom-up model that can be used for reconciliation purposes. The development of such a bottom-up model should be taken at hand together with relevant market parties. As is also prescribed by the OPTA Guidelines9, OPTA should install a working Group to involve other market parties in the assessment of the ILL tariff proposal of KPN.
Secondly, as is mentioned by OPTA, the OPTA Guidelines indicate that OPTA regards the price ceilings that are set by the European Commission as a significant testing framework for its final assessment, meaning that the ILL tariffs may in principle not exceed the price ceilings set in the Recommendation. However, as was already indicated by some market parties in an earlier consultation, since the recommended price ceilings of the European Commission are based on outdated tariffs (1999) the recommended tariffs must be corrected in line with price developments over the last couple of years. We believe that it would be reasonable to at least correct the price ceilings with minus 30%, which would result in price ceilings, that are included in the table below.
8.
For our main arguments we refer to the joint response of Energis, Telfort, Versatel and WorldCom to OPTA's consultation document on Tariff Regulation for Interconnection and Special Access, 26 January 2001, pages 8 -21
9.
Speed Length Price in Euro per month
64 kbit/s Up to 5 km 56
2 Mbit/s Up to 5 km 245
34 Mbit/s Up to 2 km 1260
34 Mbit/s Up to 5 km 1820
If one compares these (corrected) recommended price ceilings with the tariffs that are included in KPN's ILL tariff proposal, the conclusion should be that the monthly recurring prices are way above the EU price ceiling, and must be reduced by 47% for 64 kbit/s, by 60% for 2 Mbit/s10 and by 36% for 34 Mbit/s (see the table below).
Speed KPN proposal EURO/month Price ceiling EURO/month Reduction needed 64 kbit/s 130 56 47% 2 Mbit/s 620 245 60% 34 Mbit/s 2818 1820 36%
Alternatively, or in conjunction with benchmarking the proposed ILL tariffs to the EU price ceilings OPTA could also use a benchmarking with the ILL-tariffs in countries where ILL has been introduced already. However, a problem might arise from the fact that many of these ILL offers are new and that the tariffs might still be in a flux or not cost-oriented yet. In addition, and also recognized by OPTA in paragraph 25 of the consultation document, the typical Dutch circumstances would suggest that ILL tariffs in The Netherlands should be amongst the lowest in Europe. But this can potentially be tackled by the introduction of a correction factor. However, benchmarking the NL ILL tariffs with ILL tariffs in EU member states would be a useful tool, especially if OPTA sticks to EDC. The Graph below illustrates that rates in The Netherlands are way above EU best Practice, which is Sweden.
10.
Source: Worldcom, January 2002, retail offer and/or ILL offer (where available)
It is important for OLO’s to have a certain predictability on ILL tariffs, therefore it is recommended that for ILL services eventually a price cap would be put in place, as has been suggested for switched services as well by market parties in reponse to OPTA’s consultation on Tariff Regulation.
NRC and RC Tariff Analysis
5KM 2Mbps Leased Line Tail - One Year Contract
0% 20% 40% 60% 80% 100% UK PPC offer-new site UK PPC offer-existing site UK std-new site UK std-existing site Belgium zone 1 Denmark Average Europe France Austria Sweden Spain Ireland NL CTN Norway Germany Haupttrasse Switzerland NL Digistream Italy PPC US Verizon NRC% RC%
ANNEX I
INTEGRAL COMMENTS ON KPN REFERENCE OFFER FOR
(Concept version 0.4, 23 November 2001)
FIST Presentation June 20th 2001 by Paul van der Loo (KPN)
Slide 5
• MDF Access is not an alternative for ILL;
• Reciprocal offers are not valid as OLO’s are not parties with SMP (Aanmerkelijke Markt Macht);
Slide 7
• An ILL is projected as a leased line between two end-users not being a Telco on one end; This is not acceptable. An ILL is a half leased line, or partial private circuits between a customer and a POI.
• The Point of Interconnection is projected as being the Local Access point; This should include Regional Access points;
• Definition of an End-user is not relevant here; Slide
9
Please provide us with a list of LAP, RAP locations; Slide
10
Line speeds between 64 kb and 2048 kb and above 2 Mbit/s should also be offered; 2Mbit/s both structured and unstructured
Slide 11
• Tariffs should be cost-oriented
• Tariffs shown look extremely high, also if one compares these tariffs with the current commercial CTN proposal of KPN. Tariffs should at least be below the price ceilings set by the European Commission (and the EU price ceilings should be corrected by minus 30%)
• Add missing tariff components;
Raamovereenkomst Interconnecterende Vaste Verbinding Concept version 0.4
RO 1 1.1 Please provide version numbers and issue date with all Annexes; RO 2 1.3 How are addings and changes to the General Terms handled?
RO 3 2.4 This article is not acceptable. It is not of KPN’s interest whether OLO’s contracts with it’s end users cover the regulations set out in the
Raamovereenkomst. Two different issues which must not be connected; RO 4 2.6 We don't see any argument why KPN should not apply all regular articles in
the General Terms valid. It is our opinion that in this case the regular “Algemene Voorwaarden voor Vaste Verbindingen” apply; indeed KPN is always obliged to interconnect and deliver ILL
RO 5 2.7 See RO 4 RO 6 2.8 See RO 4 RO 7 2.9 See RO 4
RO 9 6.4 Individual Services are contracted for an initial term of 1 year followed by an indefinite period with a 1 month term of notice;
RO10 7 Add a term where the Raamovereenkomst supersedes any other contracts in case of disputes;
RO11 7 Add an article where the applicable law is the Dutch law;
Service Description Concept version 0.4
SD 1 1.1.1. ILL is a service that must be provided under the ICD and is subject to OPTA regulation
[Also see our answer to question 4 of the consultation document]
SD 2 1.2.1.i The KPN ILL should at least be provided in accordance with the General Terms & Conditions for Leased Lines instead of an alternative. KPN delivers leased lines for a long time under these GT&C.
SD 3 1.2.1.ii ILL should not be restricted to 64kb/ 2048 kb structured and 2048kb unstructured only. All in between speeds (n x 64kbit/s) and speeds above 2 Mbit/s should also be included.
[Also see our answer to question 6 of the consultation document]
SD 4 1.2.1.ii The type of connectors on both ends does not have to be the same. If for efficiency purposes delivering the ILL at Telco site on a fibre optic connector then this should be possible. Unnecessary use of TMX equipment should be avoided.
[Also see our answer to question 8 of the consultation document]
SD 5 1.2.1.iii • Collocation should not be a requirement for ILL [Also see our answer to question 7 of the consultation document]
• If there is already collocation either for MDF and/ or interconnect in the building installed this co-location should be usable.
• For areas where KPN cannot provide collocation KPN should provide backhaul into OLO facilities at no cost.
SD 6 1.2.1.iii Any KPN Collocation contract, where this kind of Collocation Service is provided, should be applicable;
SD 7 1.2.1.iii Patching between collocation cabinets should be done in accordance with OPTA guidelines for Special Access
SD 8 1.2.2 ii The ILLTU should be made optional if any alternative or groomed alternative is available;
SD 9 1.2.2 iii Forecasting was never an issue for retail leased lines and can not be an issue for ILL either. Any forecasting requirements must be rejected
[Also see our answer to question 11 of the consultation document] SD10 1.2.2.iv In case parts of the Access Capacity are not available due to KPN’s fault,
provisioning will not be held up;
SD11 1.2.2 v This criterion means that an ILL can only be provided between two end-users. One of the ends of the connection should be the Telco PoP with KPN;
SD12 1.2.3 ii The operational manual should at least include a service manual delivery and service timescales and a SLA including a penalty scheme that is applicable in case of non-performance. [see Annex II of pour response to OPTA
SD13 1.3.2 i See SD 9
SD14 1.3.2.ii Pre-ordering is the final phase of forecasting, see SD 9 SD15 1.3.2.iii See SD 6
SD16 1.3.2 v • The patching-issue within co-locations as raised for MDF/interconnection should be solved.
• And only if applicable. See SD 8
SD17 1.4 Quality of Service. KPN's QoS for ILL should at the least meet QoS of KPN retail LL's. [see Annex II of our response to OPTA consultation document for our detailed proposals for an SLA]
SD18 1.5.1 • The KPN access points are not yet included in the parameter schedule.
• These access points should be similar to the 20 RAP-interconnect points and should cover the full RAP-interconnect areas. There should be a tier2 offer where the LAP or NRC interconnect areas are covered.
SD19 1.5.5 Automatic interfaces to KPN management systems should allow for quick provisioning and service management resolution.
Tariff Schedule Concept version 0.4
TS 1 • Tariffs should be cost-oriented.
• Current tariff proposal (version 0.2) is not yet complete and should be assessed by OPTA first on cost-orientation;
TS 2 Tariffs at the moment are higher than the current commercial Wholesale Offer of KPN and the price ceiling recommended by the European Commission, and thus not cost-oriented
Parameter Schedule Concept version 0.4
PS 1 1.1.1.1 ILL is subject to regulation
PS 2 1.3.1. • A total initial delivery time of 20+10+25 = 55 working days is far too much. [see Annex II of our response to OPTA consultation document for our detailed proposals for an SLA]
PS 3 1.3.1. Supply Time Penalty shouldn’t be capped to 100 % MRC See our proposal in Annex II. Paragraph 3
PS 4 1.3.1 There should be a procedure to obtain an individual leased line as an urgency order within 48 hours at extra cost.
PS 5 1.4.1 • This is not acceptable. We refer to our proposals included in Annex II PS 6 1.4.1 Fault repair times should be focused on individual lines not on percentages of
lines;
PS 7 1.4.1 A yearly availability measurement should be included
PS 8 If contractual period are different we would like to see effect of longer periods on the actual tariffs as well as a volume discount;
PS 9 2.2.1 Cancellation of a non-placed order is not applicable. Forecasting was
discussed in SD 9. If forecasting will be an issue than is should be regarded as an intention and not binding (i.e. MDF Access). Delete this whole paragraph; PS10 2.2.2. This kind of terms is not acceptable and part of the risk being company with
Definitions Concept version 0.4
Def 1 1.1.1.1 ILL is subject to regulation (see OPTA's consultation document) Def 2 Interconnecti
ng Leased Line
All speeds between 64 Kb and 2048 kb and above 2Mbit/s should be included
Def 3 Cluster Point Should include Regional Access Points as well
Def 4 End User The phrase "end-to-end private circuit" should be deleted Def 5 Interconnecti
ng leased line
Definition should say LL from customer premises to Telco PoP Def 6 Point of
Connection
Point of Connection should be possible at Telco’s Technical Site as well
Technical Manual Concept version 0.4
TM 1 1.1.1.1 ILL is subject to regulation
TM 2 2.1.1.1 line speeds between 64kb and 2048kb and above 2 Mbit/s and Coaxial connections should also be offered;
TM 3 2.2.1.2 For the regional option we want to see an ODF (Optical distribution frame) as a line terminating unit for interconnection line speeds from STM1 and up (G707) (G957).
TM 4 2.2.1.4 For a regional option with an ODF we prefer to use fc fibre connectors. All equipment, interfaces and signals should comply with G703, G704, G707, G957, G958.
TM 5 2.2.1.5 This sounds like a bit trivial. This should be conform the normal standards for collocation as described in the regulated offer for collocation;
TM 6 2.3.1.2. i
The optional ITU-T G703 was deleted with reference to version 0.3, please put in in again;
TM 7 4 ILL Quality. See Annex II
TM 8 4.1.1.4 Planned Work should be only be distracted from the total time and thus not affect availability when Planned Work was well planned (minimum
notification of 10 working days), otherwise this planned work will be classified as service affecting and thus affect the availability
TM 9 4.1.1.5 ITU-T recommendation G821 doesn’t prescribe QRSS signals to be used. TM10 4.2.1.1/
3
Quality levels: there should be an option to get ILL with 99,985 % availability
Billing Manual
BM 1 1.1.1.1 ILL is subject to regulation
BM 2 2.1.1.1 If KPN can’t deliver the correct information, OLO states that this is a KPN problem from the start. Any claim towards OLO in order to come up with the correct info will be regarded as incorrect;
BM 3 No Billing Data Collection process was defined although presented in version 0.3;
