Future of Gas Entry Tariff Regime

DOCUMENT TYPE: REFERENCE:

Information Note CER/14/127

DATE PUBLISHED:

6th June 2014

CLOSING DATE:

NA The Commission for Energy Regulation, The Exchange, Belgard Square North, Tallaght, Dublin 24. www.cer.ie

1

Table of Contents

1.0 Introduction 2.0 Background 2.1 Decision Paper Cer/12/087 “The Regulatory Treatment Of The BGÉ Interconnectors And Future Gas Transmission Tariff Regime” 2.2 European Developments 3.0 ACER Framework Guidelines On Gas Transmission Tariffs 3.1 Option 1 - Postage Stamp Approach 3.2 Option 2 - Capacity Weighted Distance Approach 3.3 Option 3 - Distance To Virtual Point 3.4 Option 4 - Matrix Approach 3.5 Secondary Adjustments To Chosen Options 3.5.1 Rescaling 3.5.2 Equalisation 3.5.3 Benchmarking 3.5.4 Storage 4.0 ACER Guidance In Selection Of Cost Allocation Methodology 4.1 When To Choose A Methodology Appropriate To Each System 4.2 Methodology Counterfactual 4.3 Entry-Exit Split 4.4 Capacity/Commodity Split 4.5 Historical Costs V. Forward Costs 5.0 Assessment Of Cost Allocation Methodologies Applied To Ireand 5.1 Circumstance Criteria 5.2 Assessment Of Postage Stamp 5.3 Assessment Of Capacity Weighted Distance Approach 5.4 Assessment Of Distance To Virtual Point 5.4.1 Assessment Of Virtual Point Variant A 5.4.2 Assessment Of Virtual Point Variant B 5.5 Assessment Of Matrix Approach 5.6 Conclusion And Initial Assessment 6.0 Modelling The Approaches Suitable To Ireland & Next Steps

3 4 4 5 6 6 7 9 11 13 13 13 14 14 15 15 17 17 18 18 22 24 25 25 26 26 27 28 29

2

1.0 Introduction In June 2012 the CER published a Decision Paper (CER/12/087) on the future gas entry tariffing regime that would apply in Ireland. The reform is required in the coming years due to changes in how gas is sourced in Ireland. The decision was subject to Judicial Review in the High Court. Judgement in the Judicial Review was given in December 2013. Since the publication of the CER decision, changes at an EU level have continued. As part of the Third Package of energy legislation, the Agency for Cooperation of Energy Regulators (ACER) has been tasked with the development of a number of Framework Guidelines covering various aspects of the Community wide gas market. On 29th November 2013, ACER published its finalised Framework Guidelines on Tariffs. The Framework Guidelines (FG) details the alternative tariffing methodologies that must be applied at all transmission entry and exit points in the EU. This information note considers these alternative tariffing methodologies in the context of Ireland and in the context of the June 2012 decision. In summary this information note is the first step in implementing the necessary reforms to the gas entry tariffing regime in the light of European developments.

3

2.0 Background

2.1 Decision Paper CER/12/087 “The Regulatory Treatment of the BGÉ Interconnectors and Future Gas Transmission Tariff Regime”

In June 2012 the CER published a Decision Paper on the future gas entry regime that would apply in Ireland. This reform is required in the coming years due to changes in how gas is sourced in Ireland. At present the majority of gas in Ireland is sourced from Great Britain via the subsea Interconnectors (ICs) to Moffat in Scotland. The remaining gas supply in Ireland is sourced from gas storage and production facilities at Kinsale. From 2015 it is expected that indigenous production from the Corrib gas fields will displace gas from the IC system. In addition a potential Liquefied Natural Gas (LNG) project at Ballylongford, Co.Kerry may contribute additional supplies to the Irish market. The reasons for gas entry reform are set out in CER/12/087 where it is stated that; The current transmission entry tariffing regime needs reform. Without reform, and assuming the investments in the ICs are not to be stranded, the reduced IC throughput (due to new sources of gas coming on stream) will increase the unit IC entry tariff, potentially significantly so. This higher IC entry tariff would, in turn, push up the wholesale price for gas in Ireland. This would be inefficient and damaging to both consumer interests and Ireland’s energy Competitiveness.” The CER is committed to a tariff policy which recognises and rewards new entry which can be shown to be efficient by reference to other entry, in particular to entry from GB across the ICs which, it is assumed, will continue to be the marginal source of gas supply for the foreseeable future. The CER will base this tariff policy on forward looking long run marginal cost (LRMC) considerations rather than historic cost considerations – e.g. a prescribed portion of the historic costs of the ICs which would be deemed to constitute an efficient price signal to new investors.” This preference for a forward looking LRMC approach is grounded on (i) sound economic principles (ii) the specific circumstances of the Irish gas system (iii) avoiding arbitrary regulatory judgment calls which would arise with the historic based approach and (iv) compatibility with the applicable EU rules.

4

2.2 European Developments Since the publication of the CER’s Decision Paper on Entry tariffs, changes at an EU level have continued. As part of the Third Package of energy legislation, the Agency for Cooperation of Energy Regulators (ACER) has been tasked with the development of a number of Framework Guidelines covering various aspects of the Community wide gas market. The Framework Guidelines on Tariffs is one amongst a number of Guidelines finalised by ACER which are then to be developed into Network Codes by ENTSOG (as the European Network of Transmission System Operators for Gas). The function of the Framework Guidelines is to; “…contribute to non-discrimination. Effective competition and the efficient functioning of the market.”1 ENTSOG is responsible for “a network code which is in line with the relevant framework guideline” which must be delivered “to the Agency within a reasonable period of time not exceeding 12 months”. The Network Code on Gas Transmission Tariffs is expected to be binding in all Member States in 2017. A finalised Framework Guidelines on Tariffs was published on 29th November 20132. This includes a definitive list of Cost Allocation Methodologies that may be applied in Member States. CER considers that as the current regime requires reform before changes to flows in Ireland (expected in Summer 2015) and that the options available in the ACER Framework Guidelines will be binding in 2017 ( via the Network Code on Tariffs) it is desirable to implement tariff changes now that are in line with EU guidance. This will ensure that the tariff regime is as future proofed as possible and allows regulatory certainty for stakeholders. The purpose of this information note is to outline the options available in the Framework Guidelines on Tariffs and to assess their suitability to the Irish transmission system. In addition, CER has examined the Framework Guidelines in light of the principles outlined in CER/12/0873.

1

Regulation (EC) No. 715/2009 Article 6 (2) ACER Framework Guidelines on Gas Transmission Tariffs 3 th Decision Paper CER/12/087 was the subject of a Judicial Review with judgment delivered 11 December 2013. 2

5

3.0 ACER Framework Guidelines on Gas Transmission Tariffs

The ACER Framework Guidelines establish that the Network Code on Transmission Tariffs will allow EU Member States to apply one of four cost allocation methodologies to their system4 with four secondary adjustments available as required. These options are laid out in Section 3 of the Framework Guidelines. The allowed cost allocation methodology options include;    

Postage Stamp Capacity-Weighted Distance approach Distance to Virtual Point Matrix Approach

The Framework Guidelines explicitly state that the Framework Guidelines apply at all Entry and Exit points on the gas transmission system5 operated by gas Transmission System Operators6 (TSOs)7. We will first detail the cost allocation methodologies before turning to an explanation of the four secondary adjustments in the Guidelines.

3.1 Option 1 - Postage Stamp Approach The Postage Stamp approach allows for a single reference price to be determined for all Entry points and a single reference price is determined for all Exit points. The reference price for Entry points can be different to Exit points. To determine the relevant tariffs the allowed revenues, split between Entry/Exit and assumptions on capacity bookings are required. The ACER Framework Guidelines anticipate that the Postage Stamp approach will only apply in specific circumstances. Specifically, this methodology may be used where a “significant majority” (at least 2/3) of capacity is dedicated to either the domestic market or to cross-border gas flows8.

4

It should be noted that the European Commission specifically requested that a defined list of cost allocation options be made available in the Framework Guidelines. See EC letter 5 ACER Framework Guidelines on Tariffs Section 1.2. Page 5/34. It is expected that one of the four methodologies will be applied to set tariffs at both a national and Interconnection Point (IP) level. 6

It should be noted that the Guidelines, and thus the Network Code will require additional tariff issues to be addressed including Bundled Capacity and Virtual Interconnection Points. For the purposes of this information note and indeed the current reform of the gas entry regime only the sections relating to cost allocation methodologies will be examined. 7

Transmission System Operator “means a natural or legal person who carries out the function of transmission. And is responsible for operating, ensuring the maintenance of, and, if necessary, developing the transmission system in a given area…” 8 An alternative based on average distance flowed is to be elaborated in the Network Code on Tariffs.

6

It should be noted that the Entry/Exit split has not been pre-determined by ACER and thus the discretion on the appropriate split is left to NRA’s (see section 4.3 below). The application of the Postage Stamp approach would be considered to lack locational signals for new entrants to a market, as all entry point tariffs and all exit point tariffs remain the same regardless of location.

Postage Stamp Approach Cost Input

Historical

Entry/Exit Split

Input. 50:50 unless a separate/different split between entry/exit is chosen

Restriction

2/3 of flows to be dedicated to either domestic or cross-border

Entry Tariff

One uniform tariff at Entry

Exit Tariff

One uniform tariff at Exit

Secondary Adjustments

Not required

Table 1.1 Postage Stamp Approach

3.2 Option 2 - Capacity Weighted Distance Approach The second option under the Cost Allocation Methodologies allows for a Capacity Weighted Distance approach which aims to ensure that the tariffs for each entry and exit point are proportionate to the capacity they provide for the transmission system. The allowed revenues of the TSO (historic costs) are allocated according to a weighting factor based on the capacity9 and the distance between the entry/exit points. The resulting tariffs from this cost allocation methodology are based on a uniform unit price per capacity per distance. In addition there are two variants to the methodology.

9

Both technical and booked

7

Variant A This variant allows for all entry and exit combinations to be used.

Variant A Cost Input

Historical

Entry/Exit Split

Input

Restriction

Uses all entries/exits

Entry Tariff

Single uniform price per capacity/distance

Exit Tariff

Single uniform price per capacity/distance

Secondary Adjustments

Equalisation

Table 1.2 Variant A

Variant B Variant B restricts the usage to dominant flow patterns. applicable where pipelines are no longer fully utilised.

Such an approach may be

Variant B Cost Input

Historical

Entry/Exit Split

Input

Restriction

Restricted to dominant flows

Entry Tariff

Restricted to dominant flows

Exit Tariff

Single uniform price per capacity/distance

Secondary Adjustments

Rescaling, equalisation

Table 1.3 Variant B

The Capacity Weighted Distance approach may be considered as being more cost-reflective than the Postage Stamp approach as it introduces both Capacity and Distance as the cost drivers to the cost allocation process and can be expected to result in differing tariffs for Entry points ( and Exit points).

8

3.3 Option 3 - Distance to Virtual Point The Virtual Point approach keeps the distance cost driver introduced in the CapacityWeighted Distance approach but instead determines the distance from each Entry and Exit point to a Virtual Point (VP). The distance from each Entry to the Virtual Point gives rises to entry tariffs and the distance from the Virtual Point to each Exit point gives rise to exit tariffs. The capacity at each Entry point and Exit point is also a factor in the final entry and exit tariffs. There are two ways of determining the Virtual Point, either mathematically (Variant A) or geographically (Variant B). Variant A To determine Variant A a balanced peak day flow model is developed which represents the gas flows and distances on the system between all Entry and Exit points as well as nodes (points where gas flows through) on the system. The VP is moved around to get the chosen Entry/Exit split. The flow distances (capacity/km) are then multiplied by an “expansion constant” and an annuitisation factor is applied. This gives an annual tariff. Variant B Variant B may be used where a single dominant node can be identified on the system. Under this approach historical costs are attributed to entry and exit points based on their capacity weighted distance to the VP. Where a unique physical dominant node cannot be identified then the Guidelines allow for a point on the map to be determined10 based on the capacity weighted average geographical locations of all entry and exit points. As a result the entry/exit split is an output. The application of Variant A or Variant B is dependent on the network and whether dominant flows can be identified on that system. In addition Variant A allows for incremental cost (forward looking approach) to be used as a cost concept. Overleaf outlines the key differences between Virtual Point Variant A and Variant B.

10

In this case there is a physical location for the virtual point whereas in Variant A in a complex network there may not be a readily identifiable location for the virtual point.

9

Variant

Cost Concept

Steps

Approach for determining the location of the VP

Tariff calculation

Secondary Adjustment?

Variant A

Forward looking

Combination of Entry/Exit distances via reference nodes

Mathematical

Marginal distance *Expansion Constant+ Annuitisation Factor

Rescaling

Variant B

Historical

No reference node. The capacity/distance of each entry and exit is measured to the Virtual Point.

Geographical

Minimise difference between revenue obtained from entry/exits and required revenues.

Not required if physical dominant node on the network is used

Table 1.4 Virtual Point

Below summarises each approach.

Variant A Cost Input

Forward looking

Entry/Exit Split

Input OR Output

Restriction

A reference node is required which may then be adjusted if the Entry/Exit split is an input.

Entry Tariff

Different at each entry point

Exit Tariff

Different at each exit point

Secondary Adjustments

Rescaling, equalisation

Table 1.5 Virtual Point Variant A

Variant B Cost Input

Historical

Entry/Exit Split

Output

Restriction

A physical point where flows converge is required.

Entry Tariff

Different at each entry point

Exit Tariff

Different at each exit point

Secondary Adjustments

Rescaling (potentially), equalisation

Table 1.6 Virtual Point Variant B

10

3.4 Option 4 - Matrix approach The final Cost Allocation methodology under the Guidelines is a Matrix approach. In comparison to other methodologies this approach allows the cost concept used to be either historical or forward looking (incremental). The use of the Matrix may be considered to give strong locational signals as it is based on determining a unit cost between all combinations of Entries and Exits. This approach segments the system into sections between Entries and Exits and then determines a unit cost on these segments based on the capacity and the costs associated with that segment. The level of granularity of the Matrix is dependent on the number of Entry and Exits used in the Matrix. The Matrix approach aims to have the entry tariff (or corresponding Exit tariff) close to the unit cost that the matrix delivers. The approach measures the capacity and distance from each Entry point to each Exit point. There is considerable flexibility in the number of Entry/Exit combinations, which are then adjusted to entry and exit tariffs by an optimisation method (least square error). The flexibility of the matrix can accommodate a constraint to ensure that no negative capacity charges occur. In addition, the flexibility of the Matrix approach allows other measures such as the fixing of an entry/exit split to be pre-determined as an input. This would allow a specific element of allowed revenues to be recovered from entry and the remaining from exit.

Matrix Approach Cost Input

Flexible ( can be historical or forward looking)

Entry/Exit Split

Flexible ( can be an input or output)

Restriction

None

Entry Tariff

Different at each Entry Point depending on cost concept

Exit Tariff

Different at each Exit Point depending on cost concept

Secondary Adjustments

Rescaling, equalisation.

Table 1.7 Matrix approach

11

Entry 1 Moffat UK

-

Entry 2 - Inch

Entry 3 Shannon

-

Entry 4 Corrib

-

Exit 1 - Dublin Exit 2 - Galway Exit 3 - Limerick Exit 4 - Cork Exit 5 - Waterford Exit 6 – Cork-Dublin Exit 7 - NEP Exit 8 - PTTW Table 1.8 Blank Matrix example

Above is a sample Matrix that could be used in the case of Ireland. Each cell of the Matrix is populated with an estimate of the “true cost” of moving gas from a given Entry Point to a given Exit Point. Thus, looking down a column there will be a different “true cost” of getting gas from (say Moffat) to each Exit Point. However, only a single entry tariff is to be charged regardless of which exit is used. Clearly there will be “errors” when the single entry tariff is compared with the “true cost”. The least squares error method is used to find the Entry tariff that results in the minimum errors.

12

3.5 Secondary Adjustments to Chosen Options In addition to the Cost Allocation Methodologies above the Framework Guidelines recognises that the calculation of tariffs that apply will involve what ACER calls “trade-offs”11. The Framework Guidelines state that; “NRAs may decide to adjust methodologies and associated initial tariffs at national level, via secondary adjustments” In addition the Guidelines state that only the adjustments outlined in the Guidelines may apply, “and the adjustments made shall be fully transparent and shall not undermine the initial decision to use a given methodology”. There are 4 secondary adjustments allowed. These relate to;    

Rescaling Equalisation Benchmarking Storage

3.5.1 Rescaling Rescaling may be applied to tariffs in limited circumstances to ensure that the allowed revenues are recovered, or to avoid negative capacity charges. To rescale tariffs the Guidelines allow for a constant to be added to tariffs or for tariffs to be multiplied by a constant. In this way the adjustments increases all tariffs by the same level. The Guidelines state that if the absolute difference in tariffs between entry points is a key economic signal, then adding the same recalling to all entry point tariffs maintains this signal12.

3.5.2 Equalisation Equalising allows for tariffs at sets of points to be set to the same level. The tariffs that are to be equalised must be the same type of point and cannot be a mixture of domestic and crossborder points. ACER states that the exhaustive list includes “Entry Interconnection points, Domestic points, Domestic entries, Domestic Exits, Entries from Storage, Entries from LNG terminals, Entries from production points”. The application of equalisation is limited to circumstances where  

The set of points serve a security of supply purpose Equalisation leads to price stability Equalisation fosters competition in the retail and/or renewable sector

11

Page 25/34 ACER Framework Guidelines on Tariffs Section 3.3.2.1 “In addition where a recalling is used to recover costs the assessment shall cover the consistency of this rescaling with the economic signals, locational signals in particular expected from the chosen allocation methodology. 12

13

Where an NRA has decided to apply equalisation to a set of points then this must be justified and take account of the potential benefit of equalising versus the efficiency gains of providing locational signals at the points. The application of the Postage Stamp approach to tariffs may be considered an equalisation approach. In this case all Entry points are equalised to the same tariff level and all Exit points are equalised to the same tariff level.

3.5.3 Benchmarking The concept of benchmarking is provided for in Article 13 (1) of Regulation 715/200913. Under this adjustment a tariff on one pipeline may be benchmarked against other pipelines where “effective pipeline-to-pipeline competition” exists. The competition between pipelines must indicate that a real choice for system users exists between systems. In addition, the effect that benchmarking has on the Entry/Exit split from the application of the cost allocation must be demonstrated. Finally, if an NRA intends to benchmark a tariff it must publicly consult on the proposed benchmarking and include the proposed increases that would arise for other tariffs. Benchmarked points on the system must also be communicated to ACER. For rescaling and equalisation there is no requirement to notify to ACER, or to consult on the proposal. However, where rescaling is proposed, the NRA must justify this at a national level.

3.5.4 Storage The final Secondary Adjustment relates to gas Storage. In the specific case of Entry/Exit points to and from Storage facilities NRAs may consider the benefits that Storage brings to the system as well as the need to promote efficient investment in networks14.

13

Article 13 (1) ….”and, where appropriate, taking account of the benchmarking of tariffs” CEER Interim Report on Changing Storage usage and effects on security of supply references that the development of the Framework Guidelines is likely to contribute to the facilitation of Storage as a flexibility tool. 14

14

4.0 ACER guidance in selection of Cost Allocation Methodology This section sets out the circumstances that ACER envisages for the application of the various cost allocation methodologies within each Member State. In essence ACER gives guidance to the NRAs, where the NRA’s can assess if those particular circumstances apply to their systems. For example, in their introduction to this topic15 ACER refers to incremental costs being appropriate in expanding systems “either resulting from an increase in demand, or triggered by a change in the general system sourcing (including a change proportion of domestic/cross-border flows). Conversely, the use of observed costs may be favourable for networks presenting constant or decreasing or decreasing consumptions, facing limited to no change in the system sourcing16”

4.1 When to choose a methodology appropriate to each system To accommodate differences in the national characteristics of systems the Guidelines explicitly allow NRAs a choice of cost allocation methodologies. As such the four options summarised above may have differing cost concepts and inputs depending on the system complexity, level of cross-border flows, expected demand growth in the system etc. Overleaf summarises the key differences between the methodologies inputs and cost concepts.

15 16

Section 3.2.1.2 Inputs criteria ACER Guidelines Page 15/34

15

Methodology

Cost Concept

Entry/Exit Split

Key inputs in tariff derivation

Cost drivers

Can it be applied in Ireland?

What do the ACER Guidelines say?

Postage Stamp

Historical

Input

Allowed Revenue/Capacity booking

Capacity

Ireland meets the 2/3 criteria set out in the Guidelines.

2/3 of transmission capacity dedicated to domestic or cross-border. Difference between average distance of cross-border and domestic flows not to exceed a threshold.

CWDA

Virtual Point

Historical

A Incremental

B Historical

Matrix Approach

Historical or Incremental

17

Input

Input

Output

Input or Output

Proportion of 18 capacity at each Entry and Exit points relative to 19 total capacity

Capacity

Expansion Constant ( €/GWh/Km)

Capacity

Minimise difference between revenues obtained from entry/exit points and required revenues

Capacity

Tariff to be as close as possible to the unit cost from the matrix.

Flexible

Distance

Yes, suitable for meshed networks where the Virtual Point may be too complex.

“…allowed revenue to be collected from each point should be proportionate to its contribution to the cost of capacity 20 on the system”

Yes

“Difficult to specify an appropriate geographical point”. “Reference node can be selected arbitrarily from the physical points in the network.”

Yes

“Appropriate in systems with a single dominant node which can be clearly identified as 21 reference point”

Yes

No guidance

Distance

Distance

Table 4.1 Comparison of methodologies

17

Applied to Total Allowed Revenues Booked or technical 19 Booked or technical 20 ACER Framework Guidelines at Page 18/34 21 Ibid at Page 23/34 18

16

4.2 Methodology Counterfactual In addition to the proposed choice of methodology the Guidelines state that a counterfactual must be examined22. The purpose of the counterfactual is to assess the relative benefits that the proposed methodology provides in terms of;    

Cost reflectivity Locational signals Transparency Tariff stability

The inputs for both the primary methodology and the counterfactual must be applied consistently to both methodologies i.e. they must be the same inputs23. This ensures that the relative pros and cons of each approach are assessed in terms of the objectives above. In addition the ACER Guidelines state that; “the choice for or against the matrix methodology, or the virtual point methodology relative to the capacity weighted distance methodologies, shall consider both the drawback of necessary network representation simplifications and the benefit in cost reflectivity, as compared to the capacity-weighted distance approach”.24 Therefore, the CER concludes that in assessing a chosen methodology, consideration must be given as to why a Matrix or Virtual Point approach is better suited to the system than the CWDA. This would, prima facie indicate that where a forward looking approach is chosen then it must be compared against the CWDA approach. In addition, the chosen methodology must be compared against a counterfactual, which must be based on the same primary data inputs and assumptions as the chosen methodology. For the purposes of this information note the CER will propose an initial evaluation of the methodology that is suited to the characteristics of the Irish system as well as consistent with the already established principles in CER/12/087.

4.3 Entry-Exit Split The split between the revenues to be recovered from entry points and exit points may be determined or approved by the NRAs. In applying the cost allocation methodology the split may be an input or may be an output. In particular, this split can be an outcome of the Virtual Point (Variant B) and the Matrix approaches.

22

For the avoidance of doubt all Member States may apply the Postage Stamp approach as the counterfactual, regardless of whether the Postage Stamp circumstances apply or not i.e. 2/3 of capacity dedicated to either transit or domestic consumption. See section 3.2.3 ACER Guidelines. 23 Page 17/34 ACER Guidelines 24 ACER Framework Guidelines on Tariffs at Page 14

17

The Guidelines state that in certain circumstances to avoid cross-subsidies between types of network-users a split based on cost drivers such as capacity and distance may be used25. The CER is of the view that the general principle of a 50:50 split between entry and exit will be followed. This would require the required entry/exit split to be an input into the calculation of tariffs. Decision Paper CER/12/087 stated that a 50:50 split between entry and exit was preferred26.

4.4 Capacity/Commodity Split The ACER Framework Guidelines require that “the collection of revenues shall be based on capacity charges27” There are specific cases where other charges may apply such as compressor fuel costs or charges for specific infrastructure such as metering. At present differing Capacity/Commodity splits are in place throughout Europe. The Guidelines will apply a uniform Capacity only basis in all Member States. ACER is of the view that as capacity i.e. the pipelines themselves are the primary cost driver in gas transmission systems, it is appropriate that the allowed revenues are recovered from tariffs based on capacity.

4.5 Historical costs v. Forward costs It should be noted that the inputs to each methodology are dependent on whether a historical or forward looking cost concept is used. As outlined above ACER considers that systems that are expanding or facing changes to system sourcing are most suited to incremental costs, whereas the use of observed costs are appropriate in systems with constant or decreasing consumption that do not face changes in system sourcing. From a European Commission perspective one of the primary objectives of the Guidelines, and subsequent Network Code, is to ensure that the allocation of costs between domestic and cross-border flows (exit) does not negatively impact cross-border flows of gas. ACER also refers to the various cost concepts and refers to both historic and forward looking costs28. Observed costs are divided into 

Historical Costs-The costs registered at the time of building the existing system Replacement Costs-Cost of building the system in a given year

25

This will be further detailed in the Network Code See Section 4.3 of CER/12/087 27 Page 13/24 ACER Guidelines 28 Section 2.3. General publication requirements 26

18

Incremental Costs reflect the costs of expanding the system. This includes;   

Long Run Average incremental Costs-Cost for developing capacity triggered by demand Standardised cost of expansion-Unit price for building new assets Investment plan based costs-Costs estimated in a specific investment plan

As indicated in the Guidelines there is a level of discretion left to the NRAs with regards how to approach either observed (historic) or incremental (forward) cost concepts. These may be demand based or project based for example. As the table on the next page shows, a forward looking cost concept can incorporate different sized increments, at different points on the system. As such, the choice of forward looking increment can be refined according to the characteristics of the network. For example, a project may be identified as particular to a specific point on the system due to changes in local flows. This would be considered an application of the LRIC approach. Alternatively, an average demand increment applied across the network would be an example of the LRAC approach.

19

What Forward looking cost concepts are available? Below illustrates the different economic cost concepts that could apply to a future tariffing regime in Ireland given different concepts of long run costs.

29

Cost Concept

Description

Historic

Costs already observed on the network. Historic Costs can accommodate Long Run Average Cost below. May be described as the backward looking costs of building assets on the network that has already been built.

LRAC

Gives an average incremental cost of all points on the system based on historical costs.

(Long Run Average Cost) May be described as taking the average historical costs already seen on the system and applying them to give an average cost of maintaining the current capacity on the system. LRMC

LRMC approach considers a small increment of additional capacity on the system.

(Long Run Marginal Cost) This is the cost of delivering a small marginal or additional unit of capacity to the system, the cost of just one more “unit”. It would not be based on projects delivering realistically sized increments of capacity to the system. LRIC

The next appropriate investment at a point on the system.

(Long Run Incremental Cost)

This may be the next investment that would be required on the system such as a pipeline that would deliver a (realistic) increment of capacity rather than just a (theoretical) unit of capacity.

The key difference between Marginal and Incremental is that Incremental uses a “realistic” next increment of capacity whereas Marginal uses a “theoretical” additional unit of capacity LRAIC

Average cost of the next expected set of investments at a point on the system.

(Long Run Average Incremental Cost) May be described as what would likely need to be built on the system over time at each point. This could be one or more increments. These costs would be based on projects identified to deliver increment(s) of capacity. The average cost of all these projects leads to the LRAIC.

29

These descriptions are not to be taken as dictionary definitions; rather they aim to illustrate the various concepts. A more formal explanation is indicated in bold with an explanation in italics.

20

5.0 Assessment of Cost Allocation Methodologies applied to Ireland

As described in the previous section, in assessing the appropriateness of the chosen cost allocation methodology each NRA is expected to ensure that the chosen methodology complies with the objectives of the Guidelines. This includes





Circumstance criteria Is the chosen methodology suited to the characteristics of the Member State? This might consider such aspects as the number of entry points, the level of ullage30, or the requirements to consider whether flows are dedicated to cross-border or domestic use. Cost allocation test The cost allocation test requires the NRAs to compare the revenues recovered from cross-border and domestic points with an assessment of the costs that each impose on the network. 31 Counterfactual As detailed in Section 4.2 above the NRA must also provide a counterfactual methodology against which the chosen methodology must be examined.

30

Ullage refers to the level of unused capacity in a system or pipeline. It should be noted that the Network Code on Tariffs will contain a more detailed test. This will require NRAs to test the cost drivers based on a pure application of the chosen methodology as well as test the methodology after the application of secondary adjustments, if any are applied. See Section 2.1 above. For the purposes of the test an LNG terminal is considered a cross-border point ( Footnote 11 of the Guidelines at page 15/34) 31

21

5.1 Circumstance Criteria

Figure 5.1 Irish Gas transmission system

As can be seen from the map above Ireland is connected to the island of Great Britain via the Moffat Entry Point. This Entry point currently supplies over 90% of Irish gas requirements, and thus the price of gas at the virtual trading point in Great Britain (NBP) plus the cost of transporting that gas via the sub-sea interconnectors determines the wholesale price in Ireland. A second Entry point connects to the system at Inch, providing a small amount of production gas as well as Storage gas. As a result currently the majority of gas in Ireland converges at Loughshinny and Gormanston where IC1 and IC2 meet land off the sub-sea IC system. Therefore, the simple tariff system based on historical costs of the ICs has been suitable when flows have been primarily via that IC system. Over the coming years the number of Entry points and sources of supply is likely to increase. The table below32 indicates the changes to the entry points that are expected to occur in the coming years. These changes are expected to result in changes to flow patterns in Ireland, for example flows from Corrib towards the Southern section of the network33. Tariff reform must consider these changes to both the number of Entry points as well as consider the changes to flows within the island.

32

Source: Network Development Plan 2013 Similar flow changes can be seen in GB with flows from terminals at St. Fergus becoming less dominant as imports to England have increased e.g. LNG terminals at Teeside, Isle of Grain and Milford Haven. 33

22

Year

2014

2015

2018

Entry Point

Supply type

Supply 34 Volumes

Peak Day Demand

(GWh/d)

( GWh/d) 358

+19.4

365.7

+104.9

426.9

+191.2

Moffat

Interconnector

342.4

Inch

Storage/Production

35.0

Moffat

Interconnector

342.4

Inch

Storage/Production

35.5

Corrib

Production

92.7

Moffat

Interconnector

320.3

Inch

Storage/Production

33.8

Corrib

Production

72.9

LNG

191.1

37

Shannon

Supply Surplus

35

36

Table 5.2 Supply and Demand projections

CER explicitly recognised the impending changes to Entry in CER/12/087, where it recognised that a forward looking incremental cost such as LRMC would promote efficient entry onto the system. Considering the characteristics of the Irish system and the approach taken towards cost concepts in CER/12/087 the CER considers that in assessing the suitability of each Cost Allocation methodology the following criteria should be met; 1. A methodology that allows CER to apply a forward looking incremental cost concept. 2. Must take account of the specific characteristics of the Irish system, including the changing nature of flows and their impact on the system. 3. Should be transparent 4. Should be stable 5. Must be cost-reflective. The CER in 12/087 already identified that it wished to move from a historical cost allocation approach to a forward looking Long Run Marginal Cost (LRMC) approach. This was summarised as; “An LRMC approach to network tariff setting is well grounded in economic theory. The basic idea is to set tariffs to reflect the costs of providing an incremental unit of output at the different entry points to the system. Users will continue to purchase output at an entry point for as long as this is more valuable to them than the cost of

34

Can be considered as equivalent to technical capacity. Including Peak Day demand for Isle of Man and Northern Ireland 36 Based on Year 3 of production as per the Network Development Plan 2013 37 No Final Investment Decision has yet been made for Shannon LNG. 35

23

investing in incremental output at that point, thereby maximising economic welfare.38”

Furthermore the Decision states that; “The Proposed Decision Paper drew attention to the advantages of an LRMC approach to transmission tariff setting. It is logical, it is based on estimated cost considerations at each physical entry point, it is forward looking and therefore more likely to encourage efficient behaviour than an historic based approach. It is also clearly compatible with – though not actually required by – the EU Third Package regime. Many respondents acknowledged these points in their submissions. Moreover, given the historic circumstances and shortcomings of the inherited current tariff regime as described above (single entry tariff covering both ICs, absence of any market test for the construction of IC2 etc.), the case for moving to an LRMC approach becomes stronger, in the CER’s view. However, it is acknowledged that LRMC based tariffs can be complex to calculate. There can be much debate on selecting an optimal methodology. It also has the potential to add to tariff unpredictability if the methodology chosen is not well bedded down at the outset or if it is not based on a reasonably lengthy time period. Such volatility is not good for improving the regulatory climate for actual or potential investors39.” The principles in CER/12/087 continue to stand and the basis for the CER moving to an LRMC approach have been the subject of a Judicial Review which considered the vires of the CER in moving towards an LRMC approach. As noted in CER/12/087 the calculation of tariffs can be difficult and complex. This complexity is in part due to the fact that there is a number of differing forward looking cost concepts that may be applied to Ireland. The approach considered in CER/12/087 stated that an LRMC approach would be applied. However, as already detailed forward looking cost concepts can include a number of differing concepts which were not explicitly explored in CER/12/087. These may provide CER with a forward looking cost that is reflective of the increments that may be expected on the network.

5.2 Assessment of Postage Stamp As over 2/3 of capacity is dedicated to domestic flows the Postage Stamp may, in theory, be applied to Ireland. However, as has been stated in CER/12/087 the CER is of the view that a Postage Stamp approach should not be applied to the Irish system. This is because the lack of locational signals would result in a lack of cost reflectivity, and could reward inefficient behaviour on the system.

38 39

CER/12/087 Page 14 Ibid

24

As the CER has ruled out the application of Postage Stamp in Ireland, the CER does not propose that this methodology be considered either as the chosen Cost Allocation Methodology or as the counterfactual comparator.

5.3 Assessment of Capacity Weighted Distance Approach The application of the CWDA approach may be considered as more cost-reflective than the Postage Stamp approach. This is because the inputs to the methodology consider the capacity and the distance between entry and exit points on the system. However, the application of the CWDA relies on the allocation of historical costs. As the CER stated in CER/12/087 the application of an LRMC approach versus historic was based on “sound economic principles to the specific circumstances of the Irish system”. Furthermore, that Decision Paper stated that; “For this very reason – the “price to beat” character of the Moffat entry point – it is essential that the methodology chosen for calculating the regulated entry tariff at Moffat is based on sound economic principles, even in the absence of congestion. An LRMC approach to calculating this tariff is more likely to produce an economically efficient and welfare maximising result than a more arbitrary approach based on some share of the historic costs of the ICs deemed to be a proxy for an efficient new entry signal”. In essence the particular characteristics of the Irish system (with the Moffat entry point continuing to be the price to beat) means that the application of a forward looking cost concept should produce more efficient tariffs. Therefore, the application of the CWDA to the Irish system would be less favoured by CER. However, as the choice of methodology must be examined against the merits of the CWDA approach40, the CER will model results for this methodology in the next stage of the entry tariff reform.

5.4 Assessment of Distance to Virtual Point As stated in Section 1 there are two variants of the Virtual Point approach. Variant A allows for a mathematical approach, whilst Variant B considers a geographically derived Virtual Point. In assessing the suitability of either approach to Ireland it is important to note that Variant A allows for a forward looking cost concept, whereas Variant B is based on historical costs.

40

As per Section 3.2.1.1 of the ACER Guidelines

25

5.4.1 Assessment of Virtual Point Variant A Virtual Point Variant A depends on expansion factors that are applied uniformly across the network. Whilst this input will be applied consistently across the network, it may lead to unstable tariffs at certain points where, for example flows (such as production) change over time. An example of this may be a production entry point where production flows diminish over time. This is of particular significance in Ireland where flows are likely to change in the coming years as new entry points begin to partially replace other entry points (Moffat). Variant A, due to its dependence on uniform expansion costs also fails to take into account projects on the system that have been identified as necessary in the medium term, such as those outlined in the Network Development Plan. Such projects may be necessary for certain sections of the network in the future as changes to flows occur on the system. As such, whilst expansion/reinforcement of the network may be required over time it is not necessarily the case that the entire network would require expansion at a given time. Rather, in light of the changes to flows and entry points (primarily from the east coast to the west coast) certain future points on the system may require reinforcement. To obtain a tariff the gas flow distances between entry and exit points are calculated and multiplied by the expansion constant41. However, as the forward looking cost concept identified in CER/12/087 is allowable under the VP Variant A approach it may be considered as consistent with CER/12/087.

5.4.2 Assessment of Virtual Point Variant B Virtual Point Variant B on the other hand is based on the identification of a single dominant node on the system, based on physical flows or via a capacity weighted average geographical location. However, Variant B uses historical rather than forward looking cost inputs, and therefore would not be consistent with the forward looking approach taken in CER/12/087. The CER is of the view that the Virtual Point Variant B approach is not consistent with the principles set out in CER/12/087 as it is based on historic cost concepts.

41

€/GWh/km

26

5.5 Assessment of Matrix Approach The Matrix approach can be seen as the most flexible cost allocation methodology as it can incorporate either historical or forward looking cost concepts. The Matrix approach has been applied where locational signals are seen as an important objective for the tariff structure. As such, it may be considered to be consistent with the principles set out in CER/12/087. With changes to Entry points and flows arising in the coming years such locational signals will be important for the Irish system. In comparison to other approaches that use uniform expansion constants, the Matrix approach can facilitate assets that may be under-utilised and can accommodate different concepts of long run forward looking cost (see Section 4). To populate the unit costs in the Matrix inputs including the system capacity and costs associated with each segments are required. This may involve the application of either a historical cost concept or a forward looking cost concept. The Matrix is then created by populating each Entry/Exit combination with the costs (historical or forward looking projects) and the capacity of each combination of entry/exit. The complexity of the matrix is dependent on the number of Entries (columns) and Exits (rows) that the Matrix requires. As Ireland will have a limited number of entries, populating the Matrix Entry (columns) should be relatively straightforward. As Ireland has (to date) had a postalised Exit tariff the application of a methodology under the Guidelines would require choosing appropriate Exits (rows) for the matrix. Tariffs are calculated based on ensuring that the tariff at each entry and exit is as close to the corresponding value in the unit cost matrix. This is done by an optimisation algorithm that calculates the least squares error for the matrix. It should be noted that negative prices can be avoided by putting a constraint on the unit cost matrix. In addition, the matrix approach can take account of a prescribed 50:50 split between entry and exit. This would be consistent with the approach taken in CER/12/087. The flexibility of the matrix in allowing incremental costs and its consistency with the principles of CER/12/087 mean that the CER are of the view that it may be the most appropriate methodology for Ireland.

27

5.6 Conclusion and Initial Assessment The assessment below considers the different cost allocation methodologies that may be applied. The CER’s initial assessment of the available cost allocation methodologies indicates that;   

A forward looking cost concept such as the Matrix or Virtual Point approach would be considered appropriate for the Irish system. A historical based approach would not be appropriate to incentivise efficient entry onto the system. The Postage Stamp approach will not be considered as a counterfactual, although Ireland does, in fact meet the criteria for its application.

The use of Secondary Adjustments specifically Equalisation should apply to Exit points. The policy of equalising all Exit tariffs currently applies in Ireland. This was the subject of a policy direction given by the then Minister for State at the Department of Enterprise to Bord Gáis Éireann in 2001. This policy will continue to be the approach to Exit tariffs taken post this entry tariff review. Criterion

Postage Stamp

CWDA

Virtual Point

Matrix

Consistent CER/12/087?

with

No

No

Yes

Yes

Incremental concept?

cost

No

No

Limited

Yes

Yes

Yes

Yes

Yes

Low

Medium

High

High

50:50 Split?

Entry/Exit

Complexity

Table 5.3 Assessment of Cost Allocation Methodologies

28

6.0 Modelling the approach suitable to Ireland & Next Steps

As next steps the CER intends to publish modelling results (undertaken with our consultants and with information provided by BGN) on the Matrix approach. In addition, the Virtual Point (Variant A) will also be modelled. Lastly, as stated in the ACER Guidelines, the CER also proposes modelling tariffs for the CWDA as a counterfactual. This is to indicate the relative benefits of applying a Matrix or Virtual Point versus the CWDA approach as required by Section 3.2.1.1 of the Guidelines. As indicated in CER/12/087 the question of how to apply the LRMC principles and the so called common charge would be considered in some detail via a Networks Tariff Liaison Group (NTLG) before the CER takes a final decision. An initial meeting of the Group took place on 2nd April 2014 where these Cost Allocation Methodologies were outlined. It should be noted that since the publication of CER/12/087 the scope of the NTLG work has de facto been reduced, as the ACER Guidelines have provided a level of detail on methodologies and secondary adjustments that was not envisaged at the time of the CER Decision in June 2012. It should be noted that the primary purpose of this information note is to outline the cost allocation methodology principles allowed in the ACER Framework Guidelines and to outline the CER thinking on these. Following the publication of this information note more detailed examples of the modelled methodologies and the counterfactual will be published on the CER website.

In response to requests for indicative timelines for the NTLG’s work, overleaf is a proposed timeline.

29

Proposed Timeline Information Note

Published

NTLG meeting No. 2 (Modelling)

26th June

Consultation(incl. Models)

18th July – 29th August (6 weeks)

Public Workshop No. 1

8thor 15th August

NTLG meeting No. 3

11th September

NTLG meeting No. 4

30th October

NTLG meeting No. 5

4th December

Draft Decision Document

9th January 2015 – 20th February 2015 (6 weeks)

Public Workshop No. 2

30th January 2015

NTLG meeting No. 6

20th March 2015

Decision Document

21st May 2015

30