Networks Tariff Liaison Group Meeting€¦ · Bord Gáis Energy stated that they were supportive of the RPM as tariffs have been stable to date and worked as expected. SSE echoed

1

Networks Tariff Liaison Group Meeting

18th & 19th September 2018

Location: Ashling Hotel Dublin

1. Background & Context

The CRU (CEPA consultancy support) and GNI (Frontier consultancy support)

presented on the slides attached below.

Slides below discussed.

180913 TAR NC NTLG 1&2 - slides.pdf

List of attendees

180927 NTLG Attendee List 18Sept18.pdf

180927 NTLG Attendee List 19Sept18.pdf

2. Reference Price Methodology (RPM)

CEPA presented principles in TAR NC and those which underpinned previous tariff

reform.

Broad support for proposal of maintaining Matrix RPM.

From later discussions: Shannon LNG raised possibility of multiplicative rescaling,

with support from Pardus and counter argument from SSE on the basis that the

additive approach maintained the diversity premium between the entry points. CEPA

queried why this would be a better approach in the context of the principles.

CEPA asked for feedback on the material presented and the expectations of

participants. Manx Utilities stated that they were supportive of the approach taken and

noted that all aspects of RPM didn’t need to be reopened. Bord Gáis Energy stated

that they were supportive of the RPM as tariffs have been stable to date and worked

as expected. SSE echoed this view. Electroroute stated that market is changing and

that needs to be considered.

2

Action:

o Frontier to examine effect of the rescaling approach for next NTLG and provide

an optional LNG discount functionality so that the effects can be modelled.

3. Entry/Exit split

CEPA presented background and TAR NC position.

Broad support for proposal of maintaining split.

4. Multipliers & Seasonal Factors (M & SF)

CEPA presented TAR NC principles and current multipliers.

CEPA highlighted the effect of moving the multipliers into the bounds of 1-1.5.

Vermilion questioned whether each individual month had to be within the range. CEPA

to review.

CEPA indicated the effect of moving to the TAR NC calculation and the flattening of

the profile. Potential trade off in terms of the principles set out in Art. 28.

Potential for further reduction in multiplier bounds post-2023 was raised and if

transition has been considered. CEPA stated that if any change was significant a

transition would be considered, also there is a possibility to update annually through

the Art 28. consultation.

SSE highlighted that the effect on different customer types should be examined, e.g.

milk drying in summer and whether the rationale of when the M&SFs were set still

holds.

Electroroute questioned how the multipliers would transition to full compliance by 2023.

From later discussions: This topic was discussed further under the Demands

segment. CEPA raised the influence of M&SF on the demand calculations. Also

changing M&SF may lead to a change in shipper behaviour. The effect of M&SF on

LNG was discussed.

Actions:

o CEPA to review Vermilion’s query on whether each individual month has to be within the range.

o GNI to provide information on the level of short term bookings. Quarterly multipliers to be considered. Transition to full compliance by 2023 to be considered.

5. Initial Modelling results

Frontier presented initial modelling results and outlined that a detailed model

walkthrough will be provided at the next NTLG.

3

6. Demands

GNI presented the methodology for demand forecasts.

Vermilion queried the merit order. GNI stated that Moffat is assumed to be the marginal

source but open to input from participants.

Vermilion highlighted that multiple gas sources should increase overall bookings as

bookings will be less optimised. GNI stated that is possible but total bookings are

assumed stable over all scenarios.

Action:

o IOOA queried why 19/20 demand is much higher than 18/19. GNI to review.

7. Expansion constant and Annuitisation factor

GNI presented updates to both.

IOOA queried index and stated that full SWSOS costs should be used. GNI stated that

HICP is used and that PC4 allowance was included. GNI to update SWSOS costs to

reflect the gross estimated cost (i.e. including grant) as outlined in the PC4 decision.

IOOA highlighted previous discussions with GNI and the difference between actual

flow on linkline and modelled flow. GNI stated that it is a theoretical flow and therefore

not suitable for comparison with individual pipes.

IOOA queried whether GNI could calculate the dry expansion constant to reflect the

actual flow on the network for comparison against the theoretical flow calculation. GNI

to consider.

IOOA requested that the 2015 expansion constant and annuitisation factor calculations

are circulated so that the updates can be reviewed.

IOOA raised the possibility of indexing annually. GNI stated that they are open to

annual indexation.

SSE highlighted that in five years it may no longer be appropriate to send an expansion

signal, in that the methodology itself may not be appropriate for Ireland in terms of

sending entry signals (if no new entry was required).

Vermilion questioned the methodology for how fuel opex costs are determined.

GNI/CRU to review the methodology and update calculation to take account of 2018

gas prices.

Pardus pointed out that LNG compression is carried out upstream and that this should

be considered when reviewing the fuel opex methodology, Ormonde Organics (OO)

stated that this is also the case for biogas. GNI noted that Corrib also compresses

upstream.

4

Action:

o GNI to update SWSOS costs to reflect the gross estimated cost (i.e. including

grant) as outlined in PC4 decision

o GNI to circulate 2015 expansion constant and annuitisation factor calculations.

o GNI/CRU to further consider as part of this NTLG process updating the

annuitisation factor methodology to take account of 2018 gas prices.

o GNI to provide information with respect to the compressibility calculation.

o GNI to consider annual indexation of expansion constants.

8. VRF

CEPA presented on VRF; its view on network codes, principles, and requested

feedback on the role of VRF in the Irish gas market.

SSE queried firm price to which discount is applied. GNI stated that this could either

be applied to Moffat entry or Moffat exit which has been built into the model. SSE of

view that it should be applied to Moffat entry, as VRF wouldn’t exist without Moffat

forward flows (FF). SSE also indicated that from the perspective of the EU target

model, the European Commission views forward and virtual reverse flow as being in

competition in terms of gas wholesale price discovery.

ElectroRoute’s interpretation of the TAR NC is that the probability of interruption should

be applied to the direction of the product it refers to. If talking about Moffat Entry, then

probability of interruption should be applied to the forward product, as opposed to a

virtual reverse product. Following same logic, probability of interruption for VRF

product in Moffat should be applied to Moffat Exit.

ESB stated that VRF gives domestic production a route to exit and access to a more

liquid NBP.

IOOA stated that if the price is too high no one will use it, resulting in swaps which may

then reduce capacity bookings.

GNI stated that everyone has different uses for it, e.g. balancing, domestic shippers

exporting.

GNI stated that seasonal influence shouldn’t reflect probability of interruption as its

availability is based on FF nominations increasing over the day.

Equinor requested that probability of interruption is analysed over 12 months. GNI to

examine this.

CEPA highlighted that economic “A” factor may need to be transitional. Nephin stated

that it has to be cost-reflective.

SSE stated that if other examples of VRF pricing is to be examined CRU should

consider if the interconnectors are merchant as that would not be a like for like

comparison. SSE noted that the intent of the TAR NC is to capture and remunerate

TSO regulated assets and not merchant interconnectors.

5

Action:

o GNI to examine probability of interruption over 12-month period.

o CEPA to reflect on discussion and present findings at NTLG 3.

9. Entry point discounts (LNG)

CEPA presented the principles and context of discounts within the TAR NC.

Discussion of impact of LNG on wholesale gas price vs Moffat setting price as marginal

source.

Shannon LNG and Pardus requested the inclusion of LNG discounts in the model.

Discussion on LNG Security of Supply (SoS) benefit. SSE indicated that a SoS issue

would have to be identified by CRU as the competent authority and GNI as the NGEM

before a discount for LNG should be considered.

CEPA asked for feedback on importance of locational signals. OO stated it doesn’t

effect biogas production signal, may influence injection point. SSE highlighted the

importance of diversity premium last time and that it incentivises new entry being close

to demand centres and that there should be no preferential treatment for a certain type

of new entry. SSE also noted that where a discount on top of the diversity premium

was applied, this would have the effect of leading to a higher wholesale price in Ireland.

Action:

o GNI to include an optional discount for LNG entry point in the model with a view

to informing results for consultation.

10. Small scale entry

CEPA presented on possible treatment of small scale entry.

GNI note it’s important to take into account scale as there could be many small-scale

entry points.

Ceres note that in the UK most biogas production is Dx connected and locate where it

is technically viable with required demand. Tend to avoid trucking gas as it’s not

economic. Of view that locational signals wouldn’t be a factor in commercial decision

making relative to overall cost of project and that simplicity is important to producers.

Action:

o CRU to consider discussion and present at NTLG 3.

o CRU/GNI committed to analysing notional point for biogas entry.

6

11. Capacity/commodity split

CEPA presented on current approach and TAR NC with GNI presenting the impact

assessment. GNI will circulate their impact analysis to participants.

Vermilion believe optimisation of bookings would be required in analysis. GNI note

these are the extremes so the answer lies somewhere in between the two.

Tynagh believe 80:20 should be examined. CEPA highlighted that this would be

difficult to justify as cost-reflectivity is key component of TAR NC and capacity is the

network cost driver.

Vermilion raised issue of shrinkage being smeared across entry regardless of fact that

Corrib already compress upstream. Nephin agree those who cause costs should incur

them. Pardus of view this needs to be considered.

SSE highlighted importance of considering domestic customer. NDM must book 1in50.

SSE also noted that whilst there was guidance from Europe the purpose of the TAR

NC was to ensure a robust consultation process, which the NTLG process was so that

each Member State drove a decision that was right for its own particular

circumstances.

Tynagh note that balancing market price will be lower in ISEM, reducing ability of

powergen to recover short run marginal cost. The comment was in support of lower

capacity percentage.

GNI confirmed that the 90:10 capacity/commodity split is applied to the Allowable

Revenues which are mostly fixed and not variable costs.

Actions:

o GNI to circulate capacity/commodity impact analysis to participants.

o Shrinkage to be examined further.

12. Next Meeting

The next NTLG is scheduled to meet on 9th October 2018 in the Crowne Plaza

Dublin Airport, Northwood Park, Santry Demesne, Dublin 9.

www.cru.ie

Network Tariff Liason Group 1 & 2

September 18th and 19th 2018

TAR Network Code Workshop

www.cru.ie

Introduction

Update and goals for NTLG 1 & 2

1

Opening stakeholder forum on 8th of August.

CRU/GNI took stakeholder responses into account

GNI further developed model topology

Discussion informed agenda for NTLG 1 & 2

NTLG high-level goals

Over these two days the NTLG will examine the tariff structure and its parameters

NTLG participants feedback is essential to optimise for TAR-NC

This will inform preparation of impact analyses of potential changes, and a detailed examination of models at NTLG 3(Models to be scrutinised by CRU/CEPA and shared with participants at NTLG 3)

Page 2

OUTLINE OF THE WORKSHOP

Page 3

Outline of the workshop

Purpose

To ensure that the

interests of Irish gas

consumers are

protected…

…in the context of the

requirements set out

within the TAR Network

Code.

Page 4

Agenda

10:30 – 11:00 Introductions and outline of the day (CRU/CEPA)

11:00 – 12:00 High level model methodology (CRU/CEPA)

12:00 – 13:00 Presentation on tariff modelling (GNI/Frontier)

13:00 – 14:00 Lunch

14:00 – 14:30 Demands (GNI/Frontier)

14:30 – 15:15 Expansion constants and annuity factor (GNI/Frontier)

15:15 – 15:30 Break

15:30 – 16:45 Virtual reverse flow (CRU/CEPA)

16:45 – 17:00 Summary (CRU/CEPA)

17:00 End of day 1

Day 1

Page 5

Agenda

09:00 – 09:30 Re-cap (CRU/CEPA)

09:30 – 11:00 Treatment of new entry points (CRU/CEPA)

11:00 – 11:15 Break

11:15 – 12:00 Capacity/Commodity split (CRU/CEPA and GNI/Frontier)

12:00 – 13:00 Optional (contingency time and/or overall summary)

13:00 – 14:00 Lunch

14:00 – 15:00 Overall summary and way forward (CRU/CEPA)

15:00 End of day 2

Day 2

Page 6

METHODOLOGY AND KEY PARAMETERS

Page 7

Reference price methodology (RPM)

• The TAR NC does not specify a RPM that must be applied.

• It only requires that the chosen RPM is compared against the Capacity Weighted Distance (CWD) counterfactual.

• However, the TAR NC includes a number of principles (see Figure 1) that should be considered in determining an RPM.

TAR Network Code requirements

Source: ENTSO-G Implementation Document

Figure 1: TAR NC RPM design principles

ACER comments on other Member State tariff consultations refer to these principles.

Page 8


There are many methodologies that an RPM could be based on. When the TAR NC was in development, five RPM options were considered.

RPM options

• Backward looking

• Tariffs are a function of capacity and distance from Entry to Exit

• Being considered for use in GB (9/10 UNC 0621 variants)Capacity weighted distance

• Forward looking *

• Tariffs are a function of capacity and distance from Entry to Exit

• Currently in use in IrelandMatrix

• Forward looking *

• Tariffs are defined relative to a virtual point on the system

• Existing methodology in GB

Virtual point (with two variants)

• Backward looking

• Tariffs equal regardless of location

• Proposed in consultation for both Netherlands and SwedenPostage stamp

* Unlike CWD and postage stamp, will not recover allowed revenues without secondary adjustments

Page 9


CRU developed and consulted on proposals for choice of an RPM in 2015. The Matrix methodology was considered to have a number of advantages:

• Cost-reflectivity (locational signals): The matrix methodology can be used to provide locational signals through gas transmission tariffs (in contrast to ‘postage stamp’).

• Cost-reflectivity (forward looking signals): Stable/rising gas demand and the potential for new supply sources meant that forward looking signals were considered important.*

• Stability: Modelling identified greater stability of the tariffs across scenarios produced by the Matrix model relative to the CWD and Virtual Point methodologies.

• Transparency and predictability: The Matrix approach was considered more transparent than the Virtual Point methodology.

The Matrix methodology

* This drew on a decision made by CRU in 2012 that a forward-looking signal was most appropriate for the Irish context.

Matrix RPM developed to reflect the objectives and circumstances in Ireland.

Page 10


The Matrix methodology

Initial position is to retain the Matrix RPM going forward.

The Matrix methodology was considered most appropriate for meeting the principles set out by CRU, and in the context of the Irish market.

These principles are broadly aligned with those contained in the TAR NC. The key issues that led to 2015 CRU decision continue to apply:

• Stable / rising gas demand

• New supply sources

• Matrix RPM provides stability of tariffs / tariff differentials

ACER framework guidelines noted that ‘incremental costs may be appropriate in expanding systems, either resulting from an increase in demand, or triggered by a change in the general system sourcing’

Page 11

Entry/Exit split


• The TAR NC does not specify an appropriate Entry/Exit split.

• It only specifies that the CWD counterfactual should have a 50/50 split.

Existing position

• In 2015, CRU originally consulted on a 50/50 split.

• Decided to introduce a 33/67 split based on the following:

• This split better reflected GNI’s asset and revenue split.

• Reduced the potential for redistributive effects on certain network users.

Initial position

Initial position is to retain the status quo.

Page 12

Multipliers and seasonal factors

• TAR NC sets out that multipliers and seasonal factors should consider the following:

• Balance between facilitating short-term gas trade and long-term investment signals.

• The impact on transmission services revenue and recovery.

• The need to avoid cross-subsidisation and enhancement of cost-reflectivity.

• Situations of physical and contractual congestion.

• Facilitating economic and efficient utilisation of infrastructure.

• TAR NC also contains requirements for limits on multipliers for short-term products.


Page 13



Source: ENTSO-G Implementation Document

NB: The ‘Post April 2023’ requirements are dependent on ACER publishing a recommendation

Figure 2: TAR NC multiplier requirements

Page 14


• Existing multipliers and seasonal factors have developed over a number of years.

• They are based on a fundamental methodology which accounts for the likelihood of the top 5 – 25 gas demand days occurring in a given month.

• Daily tariffs are then defined as a percentage of monthly tariffs, with a further disincentive included to encourage longer term bookings.

• Updates over the years which have:

• Reduced the monthly tariffs to encourage use of short-term products.

• Introduced a quarterly capacity product as a simple sum of the respective monthly tariff.

Existing approach

Product Multiplier In bounds of TAR NC?

Daily 2.89 Yes

Monthly and quarterly 1.55 No

Page 15


• Existing methodology is in line with the principles set out within the TAR NC while possible alternatives raise issues.

• At a minimum, we intend to bring the multipliers within the bounds of that required in the TAR NC.

• We seek NTLG views on additional changes that should be made to ensure enduring stability of the arrangements:

Initial position

Does the underlying analysis need to be updated?

Do changes in gas demand profiles need to be incorporated?

Are any other changes necessary?

Still based on 2007 analysis but to what extent does this need to

reflect changing demand patterns?

For example, the use of gas capacity in the summer to meet falls in wind generation output.

Quarterly seasonal factors are compliant but are a simple arithmetic mean. Is change

needed to this?

Page 16


Impact

Figure 3: Current profile Figure 4: Multiplier limits

Source: GNI Source: GNI

The impact of moving the multipliers into the bounds of the TAR NC is relatively small.

35.3%

13.2%

34.2%

12.8%

1.0% 1.0%

17.6%17.1%

25.6%

12.8%

29.9%26.5%

13.2%

30.9%

Page 17


Further changes?

Source: GNI

Figure 5: Potential for flattening of profile

• Further changes to

reflect TAR NC and

changing Irish context

may lead to a

flattening of the

seasonal profile.

• To what extent would

this be desirable?

18 September 2018

Model for NC TAR consultation

NTLG workshop

19frontier economics

Contents

1. Modelling exercise 3

2. Expansion constant and annuitisation factor 10





The model is being prepared to inform the NC TAR consultation

Implementation of the

methodology as outlined in

CER/15/140, i.e. matrix approach

Implementation of counterfactual

methodology outlined in NC TAR,

i.e. capacity weighted distance

approach

Implemented for 5 years (19/20 to

23/24) and different supply

scenarios

Main characteristics of the model

Building on the existing tariff

model, but adjusted for the

purpose of this consultation


This model will be made publicly available for stakeholders

Allows users to tweak or define

their own booking scenariosParameters are flexible

Opportunities and options for the users

Understand and replicate

calculations made as part of the

consultation process

As part of this process we will

provide guidance and a

workshop on how to use the

model


There is a range of inputs needed for the tariff calculation

Inputs on revenue recovery

Allowed revenue

Entry-exit split

Capacity-commodity splitCapacity-commodity split

Inputs on network development and

useInputs on costs

Expansion constant

Annuitisation factor

Network topology

Projected bookings


For today we have applied assumptions for these inputs. Stakeholders

will be able to input their own assumptions (1/3)


Allowed revenue

Entry-exit split

Capacity-commodity split

Assumptions used

Updates to PC4 allowed revenue

33-67

as per CER/15/140

90-10

as per CER/15/140

Stakeholders can input other

assumptions

Additional information

More detailed presentation to follow

as part of NTLG agenda





Expansion constant


Assumptions used

€7782 (per GWh/Km)/€8783 (per

GWh/Km)

Updated inputs into methodology

applied for CER/15/140

9.8%

Updated inputs into methodology

applied for CER/15/140





Stakeholders can input any

assumptions

Additional information





Network topology

Projected bookings

Assumptions used Additional information

Updated topology to include

network expansion and VRF

Updated aggregated bookings Being completed

Stakeholders can input other

assumptions

Testing and verification process

ongoing


We have run the model to allow for an assessment of the preliminary

results using the following three scenarios

Entry from:

▪ Moffat

▪ Bellanaboy (Corrib)

▪ Biogas

▪ Innisfree (from 21/22)

▪ Inch

Scenario 3Scenario 1

Entry from:

▪ Moffat


▪ Biogas*

▪ Inch

Scenario 2

Entry from:

▪ Moffat


▪ Biogas

▪ Foynes (from 21/22)

▪ Inch

*User is able to run sensitivities without biogas, see also next slide


Potential scenarios to be considered in the consultation process and

expected bookings

28

Scenario 1

19/20 - 23/24

Exits

Moffat

Bellanaboy

Inch

Biogas

Foynes

Innisfree

Scenario 2

19/20 - 23/24

Scenario 3

19/20 - 23/24

Projected

bookings 18/19

(for reference)

128

Bookings above represent Annualised GWh’s

82

6

280

152-207

81-45

0-5

4-0

280-300 280-300 280-300

4-0 4-0

0-5 0-5

81-45 81-45

152-101

0-106

152-22

0-185

• Table represents supply range at individual

entry points (Moffat, Bellanaboy, Biogas,

Inch, Foynes, Innisfree)

• Table also represents aggregated demand

range at exits

• Bookings in table represents Annualised

GWh’s

• Range of supply/demand is over 5 year

period (e.g. for Scenario 1, Moffat supply

increases over the 5 year period from 152

GWh’s in 19/20 to 207 GWh’s in 23/24)

• Red circles in table represent entry points

that are not included in certain scenarios

(e.g. Foynes excluded for Scenario 1 & 3,

included in scenario 2)

• Scenarios for modelling and presentation

purposes – users will have flexibility to input

other assumptions within models


The preliminary results based on the working assumptions suggest that

future tariffs are in line with current tariffs, but moving with bookings

Because of its relatively small

volume the introduction of

biogas leads to changes in

the other entry tariffs of less

than €1 in 19/20 and less

than €5 in 23/24

In this scenario the tariffs fall

over time because more gas

is being imported through

Moffat. Moffat has the highest

tariff (excluding the Linkline

element). This means that the

price per booking for all entry

points can go down when

more bookings pay this

higher tariff

LinkLine

element

Current tariffs (17/18) Scenario 1 (19/20) Scenario 1 (23/24)


For the different scenarios we also observe that projected tariffs are close

to current tariffs, with limited variation between scenarios

€0€100€200€300€400€500€600€700€800

Matrix Approach - 2021/22

Scenario 1 Scenario 2 Scenario 3 Current tariff

Difference

between

scenario 1 and 2

just over €1002020/21

Difference

between

scenario 1 and 2

just over €100

Difference

between

scenario 1 and 2

just over €100

Difference

between

scenario 1 and 2

just over €100

Again we observe that a

movement away from Moffat

bookings increases tariff levels

for all points. In scenario 2 more

bookings move away from Moffat

than under scenario 3


Within the scenarios the tariffs remain stable after new entry has occurred

€ 0

€ 100

€ 200

€ 300

€ 400

€ 500

€ 600

€ 700

€ 800

2019/20 2020/21 2021/22 2022/23 2023/24

MA - Scenario 3

Bellanaboy Innisfree Moffat Foynes

Biogas Inch Gormanston Exit

€ 0

€ 100

€ 200

€ 300

€ 400

€ 500

€ 600

€ 700

€ 800

2019/20 2020/21 2021/22 2022/23 2023/24

MA - Scenario 2

Bellanaboy Innisfree Moffat Foynes

Biogas Inch Gormanston Exit

In both of these scenarios we

assume new entry takes

place in 21/22,and a shift in

tariffs occurs. After this shift

the tariffs are stable.

Another way to interpret this

is to say that in 19/20 and

20/21 all scenarios are the

same

Scenario 2

Scenario 3


€0€100€200€300€400€500€600€700€800

Matrix Approach - 2021/22

Scenario 1 Scenario 2 Scenario 3 Current tariff

The modelling considers a number of anticipated developments. Users

will be able to adjust these as they see fit

For entry potentially

requiring NTS

connections we apply

simplified NTS charges.

For entry potentially

requiring NTS

connections we apply

simplified NTS charges

2020/21


The model allows for a comparison with the CWDA methodology

€0€100€200€300€400€500€600€700€800

CWDA - 2021/22

Scenario 1 Scenario 2 Scenario 3

The level of the tariffs

under CWDA is

comparable to the level

of tariffs using MA when

the same parameters

are used

The user can set the

parameters

Current tariff





We provide a high-level discussion to explain how the concepts relate to

each other and their impact. A detailed discussion of each concept will follow

Expansion constant:

Unit costs of expanding the

system to allow one unit (GWh) of

gas to flow over a distance (km) Onshore expansion constant

Offshore expansion constant

Annuitisation factor :

Factor (%) to convert the

investment costs into annual costs

per GWh/km

Finance costs

Operating costs


The LRMC costs all routes are determined using these concepts

Route A to B

200 km

offshore

100 km

onshore

€7,782

Onshore expansion constant

€8,783

Offshore expansion constant

Costs per GWh/km CAPEX for route per GWh

9.8%


LRMC per GWh for route per

year

€1,756,600

€778,200

Annual costs as % of CAPEX

€2,534,800

€248,410


The annual costs for all routes are the basis for the calculation of the tariff

in the matrix approach

Tariffs cannot be set by route, so the MA

approximates the costs with the

combination of entry and exit tariffs

A C E

B

D

€248,410 €200,000 €300,000

€400,000 €300,000 €200,000


Higher expansions constants and a higher annuitisation factor will increase

the cost differences between routes, and differences between tariffs

Base caseRoute C to B

Route C to D

Higher

expansion

constant

Higher

annuitisation

factor

DistanceExpansion

constant

Higher

annuitisationCost of route

€10,000 10%200 km

300 km

Simple values assumed for

illustration

€200,000

€300,000€100,000

Difference

Route C to B

Route C to D

€20,000 10%200 km

300 km

€400,000

€600,000€200,000

Route C to B

Route C to D

€20,000 20%200 km

300 km

€800,000

€1200,000€400,000

Frontier Economics Ltd is a member of the Frontier Economics network, which consists of two separate companies based in Europe (Frontier

Economics Ltd) and Australia (Frontier Economics Pty Ltd). Both companies are independently owned, and legal commitments entered into by one

company do not impose any obligations on the other company in the network. All views expressed in this document are the views of Frontier

Economics Ltd.

Demand Methodology for Tariff Calculation

18th September 2018

Introduction

• Forecasts of Capacity and Commodity at Entry and Exit used to calculate Transmission Tariffs within the Matrix model

‒ These forecasts represent a commercial view of capacity and not physical peak flow.

• Peak day flows of the relevant exit points within the model are used to apportion demand geographically - these weightings are then applied to a commercial forecast for use in the calculation of the Exit Tariff.

‒ Exit points are identified and peak day flows are assigned based on a combination of actual and forecast data.

‒ Some exit points represent a number of smaller non daily metered exit points.

41

Demand Forecasts

• When forecasting capacity and commodity

demand is split into three sectors, Power,

Industrial/Commercial and Residential

• The following underlying demand

assumptions drive the forecast:

▪ Power driven by assumptions around

electricity demand, renewables, price of

coal, constraints, interconnector flows.

▪ Industrial & Commercial driven by

historic demand, GDP and forecasted

commercial growth

▪ Residential driven by historic demand,

forecasted growth, efficiency

➢ Non Daily Metered (NDM) required to

book for a 1-in-50 gas year.

42

Power58%

I&C29%

Residential13%

2017 Demand (Non Weather Corrected)

Power I&C Residential

Supply Forecast

• Supply is forecast using a merit order

assuming demand is met by indigenous gas

first. It is assumed in the modelling that the

following merit order will exist

‒ Indigenous Renewables (Biogas)

‒ Indigenous Production (Corrib and

Inch)

‒ LNG

‒ Moffat

• Actual flows based on commercial

arrangements between Shippers and

Producers.

• In 2017 61% of demand was met by Corrib43

Inch7%

Corrib61%

Moffat32%

2017 Supply

Inch Corrib Moffat

Annualised Capacity

• Shippers book capacity using Annual, Quarterly, Monthly, Daily and Within-Day products.

• Once demands are forecast GNI next estimate the least cost solution to booking capacity –this will result in a combination of short-term bookings at Entry and Exit

‒ The calculation of capacity considers shippers past behaviour including trades at entry because the price of these products is derived as a multiple of the annual tariff

• Short-term prices are a multiple of the annual tariff – As such, to derive this annual tariff, Short-term bookings are annualised for input into the Matrix model.

‒ This process uses the current multipliers to give the short-term booking a weighting to convert it into an equivalent booking. See following example

▪ Forecast of daily bookings in June = 200,000MWh

▪ Daily Multiplier for June = 0.661765%

▪ Annualised equivalent = 200,000MWh * 0.661765% = 1,324MWh

44

Key Points

• Forecasting the demand for capacity and commodity requires multiple approaches and will depend on the market segment

• Exit forecasts need to consider Power, I&C and NDM

‒ NDM required to book for a 1-in-50

• Entry will take account of what is required at exit but will also look at Trades at the entry points

• Shippers have a variety of products, including Annual, Quarterly, Daily, Within-Day and Entry trading

45

Expansion Constant & Annuitisation Factor Review

Network Analysis

TAR NC Workshop

18th September 2018

Contents

• Expansion Constant Calculation Results

• Expansion Constant – Background and Approach to Calculation

• Expansion Constant – Components and Results

• Changes to GNI Network – Impact on Modelling

• Summary / Conclusion

47

Expansion Constant Calculation Results

• The purpose of the Expansion Constant is to provide a numerical value to the cost of expanding the capacity of the system so that one unit of gas can travel over a specified distance (€/GWh km)

• Requirement for Inputs and Assumptions to be Stable, Transparent and Enduring

• Expansion constants have been updated to account for both inflation and recent changes on the transmission network (2015 to 2018)

• Results in table below;

48

Expansion Constant – Background & Approach toCalculation

• The purpose of the Expansion Constant is to provide a numerical value to the cost of expanding the capacity of the system so that one unit of gas can travel over a specified distance (€/GWh km) 1

‒ Forward-looking approach based on the Matrix Expansion Constant (MEC) methodology2

• The methodology for calculating expansion constants is based around 2 key components:

‒ The pipeline cost of expanding capacity on the network

▪ Pipeline cost is related to the capacity of the pipeline, which in turn is related to the pipeline parameters

‒ The cost of compression (energy) required to move the gas through the pipeline

▪ In order for there to be any flow in a pipeline, the gas must be raised to the head pressure in the first place. The cost of compression is in large part determined by the size (or motive power) of the compressor

• In CER/14/455 the CRU considered it appropriate to consider applying two distinct expansion constants, namely a “wet” and a “dry” expansion constant.

• The expansion constant is calculated for each pipeline size.

• A wet expansion constant is calculated by taking the average of the expansion constants for each pipeline size

• A dry expansion constant is calculated by taking the weighted average of the expansion constants for each pipeline size

49

1 Actual Project Costs need to be determined on a case by case basis

2 CER/15/057 Decision on Future of Gas Entry Tariff Regime

𝐸𝑥𝑝𝑎𝑛𝑠𝑖𝑜𝑛 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡 =𝑃𝑖𝑝𝑒𝑙𝑖𝑛𝑒 𝐶𝑜𝑠𝑡 + 𝐶𝑜𝑠𝑡 𝑜𝑓 𝐶𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛

𝑃𝑖𝑝𝑒𝑙𝑖𝑛𝑒 𝐶𝑎𝑝𝑎𝑐𝑖𝑡𝑦

Expansion Constants - Components and Results

• Calculation of Wet Expansion Constants (2018 results)

50

• Calculation of Dry Expansion Constants (2018 results)

51

Expansion Constants - Components and Results

Impact of Network Changes on Modelling

52

• Changes to the GNI Transmission Network since CER/15/057:

Network Element Impact on Tariff Modelling

Twinning of the South West Scotland Onshore System (PCI 5.2) is

currently in the construction phase

Additional 50 km of 900 mm pipeline will be

accounted for in the weighted average of the ‘dry’

expansion constant

Extension of the gas network to:

• Nenagh, Co Tipperary

• Wexford Town, Co Wexford

• Listowel, Co Kerry

Tx Matrix Model will be updated to incorporate

distance between entry and exit points (offtake points

to new DX networks)

Mungret to Inchmore transmission pipeline replacement Tx Matrix Model will be updated to incorporate

distance between entry and exit points

Gas to Glanbia transmission pipeline Tx Matrix Model will be updated to incorporate

distance between entry and exit points

• Update to ‘Pipeline Cost’ and ‘Cost of Compression’ since CER/15/057:

‒ Project costs have been re-indexed to 2018 value

Summary / Conclusion

53

Questions & Feedback welcome

• Expansion Constant - provides numerical value to the cost of expanding the capacity of the system

so that one unit of gas can travel over a specified distance (€/GWh km)

• Inputs and Assumptions - Stable, Transparent and Enduring

• Retention of 2 specific Expansion Constants (Wet & Dry)

• Exp. Constants updated for inflation and recent changes on the transmission network (2015 to

2018)

54

Annuitisation Factor

• Annuitisation Factor (%) is the annual payment made to remunerate the return of and on capital plus associated

operating costs of delivering gas to the system, while taking account of the depreciation profile of an asset

• Annuitisation Factor formula – WACC * (Pipeline CAPEX & Pipeline OPEX) + WACC * (Compressor CAPEX &

Compressor OPEX) + Fuel Costs + Depreciation

• Annnuitisation Factor updated for PC4 allowed WACC - 4.63% vs 5.2% as per previous calculation

• Decrease in WACC accounts for 0.4% of the total 0.7% reduction in the A.Factor % from previous number

• Further areas of review include updating Pipeline & Compressor CAPEX and Fuel OPEX

• Fuel OPEX – Cost which feeds into overall A.Factor calculation

• Fuel OPEX - Updated cost €9.88m vs €10.8m (based on all known updates to calculation)

• Areas reviewed include - Compression required (MW’s)/Average gas consumption volumes for compressors/Gas prices

Annuitisation Factor calculated at 9.80% (or €9.80 per €100 initial CAPEX) vs 10.50% as per previous calculation

Page 55

VIRTUAL REVERSE FLOW

Page 56

Virtual reverse flow (VRF)

• The TAR NC does not explicitly refer to VRF.

• However, it does contain requirements for the treatment of interruptible products.

• ENTSO-G has subsequently set out its interpretation of VRF as an interruptible product in the context of the TAR NC.

• The TAR NC contains two options for the treatment of interruptible product tariffs:


‘Ex ante’ discounts

• Applied wherever interruption

has occurred in previous year

• Applies a forward looking

adjustment to tariffs

‘Ex post’ discounts

• May be applied where no

interruption has occurred in

previous year

• Provides ex post compensation

where interruption occurs

Page 57


• The Moffat VRF product has been interrupted in the previous year.

• As the Gormanston VRF product has not been utilised, no interruption has technically occurred.

• However, given the likelihood of interruption of a VRF product at Gormanstonthat was used, an ex ante discount approach appears most sensible for both products.

• CRU has also previously set out its intention to move from the existing registration fee to a VRF tariff which reflects the probability of interruption.

• The ex-ante discount should be calculated as follows:

𝐷𝑖𝑒𝑥−𝑎𝑛𝑡𝑒 = 𝑃𝑟𝑜 × 𝐴 × 100%

• Where ‘Pro’ is the probability of interruption (0 ≤ Pro ≤ 1); and

• ‘A’ is an Adjustment factor (1 ≤ A).

Ex ante discounts

Page 58


• There is a formulae for determining ‘Pro’ in the TAR NC:

𝑃𝑟𝑜 =𝑁 × 𝐷𝑖𝑛𝑡

𝐷×𝐶𝐴𝑃𝑎𝑣.𝑖𝑛𝑡𝐶𝐴𝑃

• N is the expectation of the number of interruptions over D

• Dint is the average duration of the expected interruptions expressed in hours

• D is the total duration of the interruptible capacity product in hours

• CAPav.int is the expected average amount of interrupted capacity for each interruption

• CAP is the total amount of interruptible capacity

• The TAR NC does not prescribe what can or should be incorporated into the A factor.

• ENTSO-G give the example of the need for hedging against the probability of interruption which may be necessary to incorporate

The ‘Pro’ and ‘A’ factors

Page 59


• Assuming the definition of VRF as an interruptible product as interpreted by ENTSO-G, there are two key issues:

1. How to determine the probability of interruption ‘Pro’?

a) How should seasonality be taken into account?

b) Is the formulae for calculation proportional in the Irish gas market case?

c) How should this be estimated given recent introduction of the VRF product?

d) How should this be determined at Gormanston where there is no experience of product use and interruption?

2. What should be included in the adjustment factor ‘A’?

a) The need for hedging against interruption?

b) Anything else?

Key issues

Page 60


• Further issues have been raised by stakeholders previously and will need to be considered:

• How do interactions with ‘swaps’ need to be taken into account?

• What impact do different options have on IBP liquidity?

• How should differences between the Moffat and Gormanston products be included?

Previous considerations raised by stakeholders

61

• Data sourced from GNI GTMS system

• Data period assumed to represent most relevant period for

analysis (introduction of Trading Platform Apr’18)

• Probability of interruption based on following formula;

• Above formula assumed as basis for initial application of

interruptible % to VRF tariff

• Initial approach doesn’t currently include an estimated

adjustment (A) factor

Probability of Interruption - Initial calculation

Assumptions

Number of days in period (Apr’18 – Aug’18) 144

Of which VRF allocations were recorded 121

Number of days where interruptions occurred 17

% of days interrupted (17/121) 14%

Potential tariff would be 86% of firm price (VRF Tariff X 86%)

Number of days where interruption(s) were experienced

Number of days where VRF allocations were recorded

Possible method of determining Interruptible %

Page 62

SUMMARY OF DAY 1

Page 63

SummaryDay 1

10:30 – 11:00 Introductions and outline of the day (CRU/CEPA)

11:00 – 12:00 High level model methodology (CRU/CEPA)

12:00 – 13:00 Presentation on tariff modelling (GNI/Frontier)

13:00 – 14:00 Lunch

14:00 – 14:30 Demands (GNI/Frontier)

14:30 – 15:15 Expansion constants and annuity factor (GNI/Frontier)

15:15 – 15:30 Break

15:30 – 16:45 Virtual reverse flow (CRU/CEPA)

16:45 – 17:00 Summary (CRU/CEPA)

17:00 End of day 1

Page 64

DAY 2

Page 65

Agenda

09:00 – 09:30 Re-cap (CRU/CEPA)

09:30 – 11:00 Treatment of new entry points (CRU/CEPA)

11:00 – 11:15 Break

11:15 – 12:00 Capacity/Commodity split (CRU/CEPA and GNI/Frontier)

12:00 – 13:00 Optional (contingency time and/or overall summary)

13:00 – 14:00 Lunch

14:00 – 15:00 Overall summary and way forward (CRU/CEPA)

15:00 End of day 2

Day 2

Page 66

RE-CAP

Page 67

TREATMENT OF NEW ENTRY POINTS

Page 68

Treatment of new entry points

The TAR NC allows for a number of adjustments to tariffs defined by the chosen RPM. Of relevance are:


• At least 50% discount must be applied to storage

• Discounts may be applied to:

• LNG entry points

• Entry points ending isolation of a gas transmission system

Discounts (TAR NC, Article 9)

• Equalisation applies the same reference price to some or all points within a homogeneous group

• It takes place after determination of tariffs under the RPM (‘ex post’)

• E.g. domestic Exit points are currently equalised in Ireland

Equalisation (TAR NC, Article 6, 4(b))

• Allows for a group of homogeneous entry or exit points to be considered as one point for the determination of tariffs

• Can be applied to homogeneous points or to points which are located within the same vicinity

• The same tariff is identified within the RPM (‘ex ante’)

Clustering (TAR NC, Article 3 (19))

Page 69


• Other than storage, discounts may be considered under the TAR NC for:

• LNG entry points

• Entry points ending isolation of a gas transmission system

• The current Matrix RPM has been designed to balance a number of criteria, including transparency, stability, etc.

• One of the key criteria is the provision of locational signals for entry (supply) sources (new and existing) connected to the transmission system.

• Any provision of discounts would impact on these signals.

• For a given Entry/Exit split, it would also require an increase in the tariffs at other entry point tariffs and may therefore place upwards pressure on the wholesale gas price.

Discounts

We are interested in NTLG views on the need and justification for any discounts.

Page 70


• In future, it is possible that a number of small scale sources of gas supply (e.g. biogas) may enter onto different points across the transmission system.

• These supply sources may be in a position to make a choice about whether they connect to the transmission or distribution system.

• The interactions between the two are therefore important:

Small scale entry points

Transmission level

• Forward looking locational signal under the Matrix RPM

Differential between tariffs may lead to sub-optimal connection decisions.

Should there be some form of common treatment for these points at transmission level (e.g. some form of clustering) which allows for consistency with distribution entry tariffs?

Distribution level

• No locational signal

Page 71

CAPACITY/COMMODITY SPLIT

Page 72

Capacity/Commodity split

• The TAR NC specifies that tariffs should be capacity based with the exception of:

a) A ‘flow-based’ charge for covering the costs of gas flow quantities; and

b) A complementary revenue recovery charge for managing under and over recovery

• Many Member States are moving further towards capacity based charges:

• ACER approved the Netherland’s tariff design which recovered all revenues from capacity based tariffs.

• Sweden and Poland also intend to recover 100% of revenues from capacity charges.

• The Utility Regulator (NI) has set out its intention to move to 95% capacity based charging.

• Denmark appears as an outlier within its consultation, with an intention to recover only 52% of revenues from capacity charges. The Danish split is based on capex/opexwith capex recovered from capacity and opex from commodity charges.

TAR Network Code Requirements

Page 73


• We are considering an appropriate capacity/commodity split between 90/10 and 100/0.

• We are interested in NTLG views on the following:

• How important is the direction of travel in Europe?

• …and particularly, interactions with neighbouring transmission systems – i.e. the 95/5 split suggested by the Utility Regulator?

• What might the impact be on the I-SEM for example?

• What are the distributional effects of increasing the proportion recovered from capacity charges?

• Informed by GNI’s impact analysis.

Considerations

Justifiable bounds somewhere between 90/10 and

100/0 for Ireland

90% 100%0%

Customer Impact Analysis Capacity Commodity Split

19th September 2018

• Tariffs currently designed to collect 90% of revenue through capacity tariffs and 10% through commodity tariffs

• High level analysis on impact of alternative split such as 95:5 and 100:0 split as outlined in the Tariff Network Code

• Conducted high level analysis using generic examples.

• Effect and result will vary depending on booking strategy

• Welcome discussion and feedback on this analysis

75

Review of Capacity/Commodity Spilt

• Assuming 18/19 demands the 18/19 tariff was re-run using both 95:5 and 100:0 capacity commodity split, all other things being equal.

• Load factor represents the relationship between a users average daily consumption and their peak day consumption. The higher the load factor the flatter the profile i.e. average usage closer to peak

• GNI looked at a user with an Estimated Annual Consumption (EAC) of 5,000,000 KWh and derived their profile using three different load factors

‒ 90% load factor

‒ 50% load factor

‒ 20% load Factor

• Looked at Transmission Transportation costs only.

• Assume Entry and Exit bookings match

‒ In reality Shippers can vary their bookings using aggregate profiles, trades etc.

• Repeated the analysis where the Shipper is assumed to book only daily

76

Analysis - Assumptions

Initial Findings Assuming Annual Bookings…..

95:5• 90% LF = -3.0%

• 50% LF = +0.5%

• 20% LF = +3.0%

100:0

• 90% LF = 5.0%

• 50% LF = 0.9%

• 20% LF = 6.0%

77

• In this analysis a customer with a 90%

load factor is likely to see a fall in

overall cost.

‒ Moving from 3% reduction under a 95:5

split to a 5% reduction with a 100:0

split.

• The higher the load factor the greater

weighting commodity has relative to

capacity

‒ As you move towards a zero

commodity charge the increase in

capacity charges will not offset the

reduction in commodity charge.

• Conversely the lower the load factor

the higher the cost will be as you move

towards a zero commodity charge

Initial Findings Assuming Daily Bookings…..

95:5• 90% LF = +2.0%

• 50% LF = +3.3%

• 20% LF = +4.3%

100:0• 90% LF = 4.0%

• 50% LF = 6.6%

• 20% LF = 8.6%

78

• In this analysis because the shipper is

booking all daily the commodity

weighting is not as large as in the

previous example

‒ In practice users are likely to book

a combination of annual and short-

term

• A move towards a zero commodity

charge suggests the cost increases

under all scenarios

‒ And also increases as the load

factor reduces

• Effect is dependent on individual booking behaviour and load factors of the Shipper/Customer

• Results will vary depending on individual analysis

• Can utilise their portfolio to reduce capacity costs through utilising a combination of capacity

products and trades at entry.

• However, from this analysis trends have emerged based on;

‒ the impacts that a users load factor has,

‒ the commodity used relative to booked capacity.

• Each individual Shipper/Customer will need to asses its own effects/impact

• Welcomes views and feedback on this analysis

79

Conclusion

Page 80

WAY FORWARD

Page 81

Way forward

• 9th October 2018: NTLG 3 – Discussion of key issues and tariff models

• Early November 2018: Consultation published

• Early January 2019: Consultation closes

• Early February 2019: Summary of responses published

• March 2019: Final decision published

• May 2019: Final tariffs published

Key dates

Thank you for your input

1

Networks Tariff Liaison Group Meeting

9th October 2018

Location: Crowne Plaza, Northwood, Dublin

1. Background & Context

The CRU (with CEPA consultancy support) and GNI (with Frontier consultancy

support) presented on the slides attached below.

Slides below discussed.

181009 NTLG 3 Slidedeck.pdf

List of attendees

181010 NTLG Attendee List 09Oct18.pdf

2. Review of minutes from NTLG 1&2

GNI presented an overview of the actions from NTLG 1&2.

IOOA noted that at the last NTLG it had queried whether GNI could calculate the dry

expansion constant to reflect the actual flow on the network for comparison against the

theoretical flow calculation. This was not reflected in the actions.

GNI responded that this will be covered off in the compressibility calculation.

3. Information on compression calculation

GNI presented on the compression calculation.

Vermilion highlighted that Corrib pay for their own upstream compression costs to

bring gas to 86 bara and that compression of entry points up to this pressure should

be reflected in the compressibility calculation.

GNI responded that this issue was debated through the 2015 process, noting that the

summary from that time was to maintain stability of expansion constants. GNI don’t

see a basis to revisit unless any new information is presented. The CRU believes

principles laid out in 2015 still holds and this will be further detailed in the

consultation paper.

2

Pardus questioned whether a new entrant would be expected to deliver gas at 86 bar

and whether the theoretical and actual costs could be compared.

GNI stated that it is a forward-looking approach (calculates the projected costs) and

any new costs are only known after a full tender has been completed. The purpose of

the calculation is to send economic signals.

SLNG stated that they take the view of Corrib. Would need more technical

background on the matter.

Nephin stated that it seems intuitively there is a mis-match between people putting

gas in at different pressures.

Pardus queried whether we can compare the historical costs that have been used for

the calculation and the actual costs.

GNI stated that to inform costs, GNI take the most recent historical project costs to

inform calculations.

Pardus highlighted that from an LNG perspective they will be providing compression.

GNI stated that even if it is dry expansion – there is a compressibility cost to ensure

operability of the gas transmission system. They also emphasised the importance of

considering a single transmission system for the purposes of designing the tariff

methodologies.

GNI restated that the issue is about marginal costs and the creation of locational

signals. They stated their view that no new information has been evidenced to

change the decision.

Pardus noted that there must be an opex saving from Corrib carrying out its own

upstream compression and that this should be reflected in the expansion constant

calculations.

CEPA stated that the tariff methodology is designed to reflect the economic costs of

the GNI system. The important consideration is about the differentials between

Moffat and the other entry points, including compression and how this impacts

competition between entry points.

Pardus requested that compression at LNG is taken into consideration going forward.

The CRU requested that participants respond in detail through the consultation

process if with any views and evidence that they consider should be taken into

consideration. On this matter, the CRU requested stakeholders to focus contributions

on the purpose of the dry expansion constants i.e. the theoretical expansion of the

onshore transmission network.

Frontier presented their review of the demand profiles used in the model, as actioned

by IOOA. Refer to Slide 12.

Vermilion requested that the Bellanaboy Entry Tariff be broken-down into the linkline

tariff element and the Cappagh South tariff element this should also be presented in

the consultation.

IOOA queried difference in capacity bookings (almost 20%) between this year and

that assumed for 2019-20.

GNI agreed with the need to consider the capacity bookings further and stated that

they are liaising with their demand team to finalise capacity booking estimates.

3

4. LNG and small-scale entry discounts

Frontier presented an explanation and impact assessment of a possible discount for

new entry points.

SLNG questioned whether there has been an analysis on the impact of cost to gas

consumer of a large-scale entry point such as an LNG entry point.

GNI stated that they would not have sight of gas sales agreements to carry out such

an impact analysis.

CEPA highlighted that as Moffat is the marginal source of gas it sets the wholesale

cost of gas. The analysis therefore implicitly includes consideration of the impact on

the gas market by applying the increased tariff to the full volume of gas bookings (i.e.

237 GWh), not only bookings on the Moffat interconnector.

Nephin pointed out that for an LNG project to alter the marginal source of gas the

assumption would have to be made that Ireland is disconnected from the GB market

which could only be the case if there were high levels of friction between IBP and

NBP.

IBEC noted that if LNG needed to be incentivised at some future time due to

concerns over security of gas supply, providing an entry point tariff discount would

potentially be more costly than a straightforward PSO support because it would raise

the marginal cost of all delivered gas. IBEC pointed out that there are no such

concerns at present, hence it is premature to consider instituting either option.

SLNG of the view that now is an appropriate time for the discussion of SoS benefits

of LNG.

RGFI discussed the PSO levy that is being collected from energy consumers. A PSO

fund should be used to support renewable gas projects as a lot of industries have

targets to be carbon neutral by 2030.

5. Multiplicative versus additive approach

Frontier presented an impact assessment of multiplicative rescaling versus additive

rescaling.

SLNG welcomed calculation of figures noting they are similar to figures they have

calculated.

SLNG stated that there are precedents in Europe of the multiplicative approach and

that these need to be considered in light of the principles of the TAR-NC.

6. Treatment of small scale entry points

CEPA presented a slide on key takeaways from last NTLG and key remaining issues.

Ormonde Organics (OO) questioned the purpose of having an interruptible product for

Biogas. The incentive should be for producers to get the maximum volume on the

system. There shouldn’t be a dis-incentive to connect to the Tx network which provides

firmer access. One also doesn’t want to be in the position of paying people to flare the

gas. Suggested there should be a cost on Distribution injection to counter

4

Transmission compression costs and to reflect the potentially non-firm nature of

capacity.

RGFI stated that any such problems should be identified beforehand through the

connections application process (i.e. either pipeline overloading or lack of demand

within a local network).

GNI responded that this was the case, however, there are two separate issues, i.e.

production and injection.

Ceres (Nephin’s advisor) indicated that, from its GB experience biogas producers are

very unlikely to want to or be able to build a business case around an interruptible

connection point. In the UK biogas producers receive a rebate for bringing on gas close

to customers (for avoiding transportation investments).

OO stated that if the proposed notional point is used it takes away the incentive for

producers to locate close to their customers whilst undermining the principle of sending

locational signals, adding that this applies to both Transmission and Distribution.

7. Virtual Reverse Flow (VRF)

CEPA presented on VRF, outlining considerations which may apply to a decision of

how high to set any discount to reflect the interruptible nature of the product.

GNI presented update on probability of interruption data which is set out in two ranges.

IOOA queried why there isn’t one seamless 12-month range of data.

GNI responded that they are waiting to retrieve data from old database which is time

consuming.

BGE questioned whether M&SF apply to the VRF based tariff.

GNI stated that the VRF product will be interrupted for technical reasons very rarely

when the new algorithm is put in place.

Nephin stated need to consider the market objectives of changing VRF product.

IBEC stated that the efficient use of the market is to the benefit of the consumer and

questioned what the effect would be on the consumer.

CEPA stated that currently VRF is a relatively cheaply priced product. If it moves to a

more expensively priced product – there needs to be an understanding of the trade-

offs and impacts on the market.

Energia stated that Shippers have had 3-4 years of a benefit from a low cost of VRF.

The regulator gave a commitment to address this within one year at that time and failed

to do so.

IOOA stated that it might be more applicable to use the Entry tariff as a basis when

calculating VRF.

Vermilion stated that the VRF product is completely different from a standard exit

product. For VRF it can be the case that on one day 40 GWh is available and next day

only 10 GWh, based on available forward flow. If initial 10 GWh is available and 5 GWh

is booked and in a later stage only 6 GWh is available, GNI stated that this is not an

interruption. These characteristics should be reflected in the A-factor.

5

IOOA added, that if the full forward capacity would have been offered, then there would

be an interruption every day.

GNI questioned what type of VRF charge would encourage a more liquid IBP.

Nephin stated that as a general principle, the more trading you can do, the more

liquidity you will have, and hence the more integrity in the price.

Energia stated that where there is a low level of liquidity, this shouldn’t be split across

markets.

IBEC stated that the only divergence from IBP and NBP is the wedge of transport costs

so whatever keeps that as low as is possible for customers is best.

IOOA stated that from the perspective of new entrants, VRF is a helpful product in

terms of signing up with new counterparties. Also, helpful from a balancing perspective.

Nephin asked in the context of the presentation, whether the objective of any change

was revenue recovery or development of the market. CEPA agreed with views that

the revenue recovery impact of any decision was likely to be low given the low volume

of use of VRF.

8. Capacity/Commodity split and shrinkage

CEPA presented slides on the capacity/commodity split and shrinkage.

Energia stated that the only market with bidding rules is the non-energy balancing

market. The capacity/commodity split is less relevant now for power generators in I-

SEM due to the fact that there are no bidding rules for most products (in multiple

markets). Constrained generators are moving to mix of products.

Unclear what incentives are if you compete in a number of markets.

No daily products in NI so has always been a different treatment. This is perhaps a

bigger potential source of distortion than the potentially different capacity/commodity

splits in Northern Ireland in comparison to RoI.

Energia stated that there would need to be a good reason to move away from 90:10

split.

Vermilion stated that there should be a distinction within shrinkage for compression

and unaccounted for gas. Vermilion’s opinion is that shrinkage is not applicable at the

Corrib Entry Point.

GNI stated that if shrinkage is paid for through tariffs this may increase tariff volatility.

Nephin questioned how much capacity/commodity split would change if shrinkage was

included in allowed revenues.

GNI stated that if you bring in shrinkage into allowed revenues could see an increase

in the capacity element as much of the commodity element would be accounted for by

shrinkage.

GNI highlighted that a change in the general principle of how shrinkage is charged

would require a change to the Code of Operations.

6

Vermillion noted that shrinkage was discussed recently at the Code Modification

Forum in terms of invoice issues. The slides shown at the Code Modification Forum

show that since Corrib came on stream, shrinkage volumes purchased have reduced

due to Corrib compressing gas upstream of its entry point, reducing flows at the Moffat

entry point and associated Moffat shrinkage volumes. Moreover, all non Bellanaboy

shippers have benefited from lower GNI shrinkage charges due to Corrib supplying its

gas at 86bara.There is a question as to whether this is fair, equitable and cost

reflective.

9. Multipliers & Seasonal Factors

GNI presented slides on the historic use of short-term products.

Energia pointed out that we may start to see more of a blend between annual and

short-term products in the power sector within I-SEM.

Energia pointed out that constrained power generators are centrally dispatched by

TSO. There is uncertainty and lack of information. Could be dispatched for long period

of time.

Energia questioned whether there would be any adjustment if you were to book

monthly product each month for 12 months i.e. a rebate because if you had full

information you would have booked yearly.

CEPA stated not necessarily covered under TAR NC and that implications would need

to be considered before identifying a position.

GNI stated that this could add to tariff volatility i.e. could discourage the booking of

annual products.

Vermilion questioned how this would impact secondary market trading of capacity and

that could introduce challenges of under-recovery for GNI.

Energia responded stating that on the first point regarding secondary market trading

the market would determine the outcome.

Vermilion stated that they would think a booking rebate would be a distortion of the

market.

IOOA pointed out that action regarding indexation of expansion constants had not been

covered off.

GNI stated their view was that they would be reviewed as part of periodic review every

five years.

IOOA stated that the gas prices range that is being used is giving a particular gas price.

If we were to use 16/17 data could be an increased price.

IOOA questioned whether the gas prices should be forward looking.

BGE of the view that a broader range makes more sense.

GNI and CRU stated that the price volatility and uncertainty to both the upside and

downside would need to be considered.

7

10. Workshop model and methodologies

Frontier presented a workshop on model and methodologies.

11. Next Meetings

An open stakeholder forum will be convened during the consultation period to

discuss the policy positions outlined. The date is yet to be decided and will be

dependent on the date of release of the consultation paper to allow sufficient time

for paper review but also to allow time to consider consultation responses after the

session.

GNI will organise a support teleconference for detailed use of the tariff model in the

consultation phase. This will be arranged after the model is released to allow users

time to work with the model and gather questions to be addressed.

www.cru.ie

Network Tariff Liaison Group 3

October 9th 2018

TAR Network Code Workshop

www.cru.ie

IntroductionUpdate and goals for NTLG 3

1

Following NTLG 1&2 on 18/19 September.

GNI carried out further analysis of issues and added additional functionality to models.o These models are now being reviewed in detail by CRU/CEPA.

CRU Gas Team highlighted to Commission its initial views, which incorporated feedback received from participants.o Their initial guidance and further CRU/CEPA analysis has informed the slides for today.

NTLG 3 high-level goals

CRU/CEPA to further highlight its minded to position on some items.

NTLG participants’ feedback is essential to informing the CRU’s proposed approaches within the TAR NC Consultation document.

www.cru.ie

Agenda NTLG Day 3

2

09:30 – 09:45 Introductions (CRU/CEPA)

09:45 – 10:15 Review of actions from NTLG 1&2 (GNI/Frontier)

10:15 – 11:15 Entry Points (All)

11:15 – 11:30 Break

11:30 – 12:15 VRF (All)

12:15 – 12:30 Capacity/Commodity Split & Shrinkage (CRU/CEPA)

12:30 – 13:00 Multipliers & Seasonal Factors (All)

13:00 – 13:45 Lunch

13:45 – 14:00 Wrap up on first half of day

14:00 – 14:15 Introduction to Training (GNI/Frontier)

14:15 - 16:00 Tariff Model –Training Session (GNI/Frontier), including Break at 15.00

16:00 Close – (Contingency to 17:00 if required)

www.cru.ie

NTLG 1&2 Actions

3

4

Table of Actions from NTLG 1&2

Topic Action Status

Reference Price MethodologyFrontier to examine effect of the rescaling approach for next NTLG and provide an optional LNG discount functionality so

that the effects can be modelled.To be presented at NTLG #3

Multipliers & Seasonal Factors GNI to provide information on the level of short term

bookings. Quarterly multipliers to be considered. Transition to full compliance by 2023 to be considered

To be presented at NTLG #3

Demands GNI to review demand profiles used in model To be presented at NTLG #3

VRFGNI to examine probability of interruption over 12-month

period. To be presented at NTLG #3

Entry Point Discounts (LNG)GNI to include an optional discount for LNG entry point in the

model with a view to informing results for consultationCompleted

Small scale entry CRU/GNI committed to analysing notional point for biogas

entryIdentification of notional point ongoing - potential approach

to be presented at NTLG #3

Capacity/commodity splitGNI to circulate capacity/commodity impact analysis to

participantsCirculated with NTLG #3 material

Expansion Constant & Annuitisation Factor

GNI to update SWSOS costs to reflect the gross estimated cost (i.e. including grant) as outlined in PC4 decision

Review completed - confirmed that total estimated costs were included in model as communicated at NTLG #1/2.

GNI to circulate 2015 expansion constant and annuitisation factor calculations.

Links to the 2015 Exp.Constant (CER15060) & A.Factor (CER15059) circulated

GNI/CRU to further consider as part of this NTLG process updating the annuitisation factor methodology to take

account of 2018 gas prices.Inclusion of 2018 Gas Prices (Actuals to Sep'18)

GNI to provide information with respect to the compressibility calculation.

To be presented at NTLG #3

GNI to consider annual indexation of expansion constants. To discuss at NTLG #3 as part of overall presentations

Information on Compression Calculation

Network Analysis

TAR NC Workshop

9th October 2018

Calculation of Expansion Constants

• The purpose of the Expansion Constant is to provide a numerical value to the cost of expanding the capacity of the system so that one unit of gas can travel over a specified distance (€ / GWh km)

‒ Forward-looking approach based on the Matrix Expansion Constant (MEC) methodology1

• The methodology for calculating expansion constants is based around 2 key components:

‒ The pipeline cost of expanding capacity on the network

▪ Pipeline cost is related to the capacity of the pipeline, in turn related to the pipeline parameters

‒ The cost of compression (energy) required to move the gas through the pipeline

▪ In order for there to be any flow in a pipeline, the gas must be raised to the head pressure in the first place. The cost of compression is in large part determined by the size (or motive power) of the compressor

• In CER/14/455 the CRU considered it appropriate to consider applying two distinct expansion constants, namely a “wet” and a “dry” expansion constant.

• The expansion constant is calculated for each pipeline size.

• A wet expansion constant is calculated by taking the average of the expansion constants for each pipeline size

• A dry expansion constant is calculated by taking the weighted average of the expansion constants for each pipeline size

61 CER/15/057 Decision on Future of Gas Entry Tariff Regime

𝐸𝑥𝑝𝑎𝑛𝑠𝑖𝑜𝑛 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡 =𝑃𝑖𝑝𝑒𝑙𝑖𝑛𝑒 𝐶𝑜𝑠𝑡 + 𝐶𝑜𝑠𝑡 𝑜𝑓 𝐶𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛

𝑃𝑖𝑝𝑒𝑙𝑖𝑛𝑒 𝐶𝑎𝑝𝑎𝑐𝑖𝑡𝑦

Calculating ‘Cost of Compression’

• Calculation of Wet Expansion Constants:

7

Compression

Required (MW) Cost of

Compression

per km

(€ / km)Cost of

Compression / MW

(€/MW)

Isentropic

Head

Eqation

GNI Compressor

Station

Costs

Calculating ‘Compression Required’

8

• Gas Throughput

• Inlet Pressure

• Outlet Pressure

• Other constants

Isentropic

Head

Equation

Compressor

Power

Requirement

1. Isentropic Head

Equation

2. Model Inputs and

Output

23.7 MW

3. Worked Example

43.5 bara 131 bara 43.1 MW43.5 bara 131 bara

174 GWh/day 316 GWh/day

600 mm 750 mm

Calculating ‘Cost of Compression’

• Calculation of Wet Expansion Constants:

9

• Calculation of Dry Expansion Constants:

Impact of Pressure Assumptions

10

1. Wet Expansion Constant: €8,783 /GWhd/km

23.7 MW43.5 bara 131 bara

174 GWh/day

600 mm

43.1 MW43.5 bara 131 bara

316 GWh/day

750 mm

2. Dry Expansion Constant: €7,810 /GWhd/km 1

10.4 MW43.5 bara 86 bara 13.2 MW43.5 bara 86 bara


600 mm 650 mm

19.0 MW43.5 bara 86 bara 30.6 MW43.5 bara 86 bara


750 mm 900 mm

1 Value updated since NTLG 2 (18th September 2018)


1. Sensitivity around entry bookings

2. LNG and small scale entry discounts – explanation and impact assessment

3. Multiplicative versus additive revenue recovery – impact assessment


5. VRF tariff results

6. CWDA counterfactual

7. CWDA and MA: intuition and examples


We have reviewed a sensitivity in which Moffat bookings are in line with

18/19 projections

In this sensitivity, projected bookings for 2019/20 are in line with 2018/19 projected bookings and an increase over time

is applied to the 19/20 base that is in line with 2018/19 projected bookings

Reduced entry bookings sensitivity – Scenario 1 Base case – Scenario 1

The reduction in entry bookings leads to a tariff increase as revenue is being recovered over fewer bookings.

In scenario 1, the reduction leads to tariff increases of €28 in 2019/20 for all entry points. This reduces to €24 in 2023/24

as the volume of bookings increases over the years, allowing the tariff increase to be spread over more bookings.

The tariff increase ranges from €19 to €33 across all years and scenarios.

13frontier economics www.cru.ie

Entry points

13










Introduction to a possible approach to discounts for new entry points

Article 9 of the TAR NC provides for the use of discounts in tariff setting.

It indicates:

▪ a discount of at least 50% must be applied to storage; and

▪ a discount may be applied to the transmission tariffs of:

LNG facilities; and

entry points developed to end the isolation/increase security of supply.

On the following slides, we will cover:

1. how a discount could be applied;

2. the results of applying a discount to LNG and small scale entry on these points as

well as the Moffat entry point.

TAR NC

Implementation Example


Discount leads to an initial under-recovery, and therefore an adjustment

needs to be implemented – indicative calc. as per TAR NC example

LNG

Pre-discount

Ta

riff in

€

LNG

Indicative Post- discount

Adjustment only to non-discounted point

Imagine a world with two entry points with the

same tariff, e.g. two entry points at exactly the

same location, but one is supplied by LNG

whereas the other is supplied by a piped gas

Simply taking 50% from the LNG tariff pre-

discount and recovering it from the non-

discounted points leads to the following

situation:

• The LNG point has a 50% discount

compared to the situation pre-discount

• However, post-discount the LNG facility has

a 67% discount compared to the non-

discounted point. The tariff of the non-

discounted point is now adjusted for all tariff

revenue removed from the LNG point. This

adjusted tariff will be higher than in the pre-

discount situation, creating a greater wedge

between the two points

LNG

Recovering the discount-affected revenue from

all points proportionally (also applying the

discount to this revenue for discounted points)

creates a situation in which:

• The LNG discount relative to a situation

without discounts is less than 50%

• But in a world with LNG discounts it creates a

50% discount relative to an entry point not

supplying LNG at that same location

Indicative Post- discount

Adjustment to both points

67%

dis

count

50%

dis

count

50

%

33 %


Discounts of 50% on LNG and small scale entry points drive an increase in the Moffat tariff,

leading to increased costs for consumers

*We have estimated the total impact of offering a discount on LNG and small scale entry points as the

difference between the Moffat tariff with and without discounts multiplied by the annual bookings

Scenario 1

Scenario 2

Scenario 3>€1

€43

19/20

23/24

237 GWh

256 GWh

€14k p.a.

€11m p.a.

Increase in

Moffat tariff

(€/MWh)

Annual

bookings

Impact of

discount per

annum*

>€1

>€1

19/20

23/24

237 GWh

256 GWh

€14k p.a.

€122k p.a.

>€1

€126

19/20

23/24

237 GWh

256 GWh

€14k p.a.

€32m p.a.

Moffat tariff

w/out

discounts

Moffat tariff

with

discounts

€291

€265

€291

€392

€291

€342

€291

€266

€291

€518

€291

€385

There are no LNG

bookings and low small

scale entry bookings under

scenario 1, leading to a

relatively low total impact

of the discount.

The larger the proportion

of LNG and small scale

entry bookings, the larger

the increase in the Moffat

tariff as there is more

revenue to be recovered.

This in turn leads to a

higher overall impact on

gas consumers.

For this reason, the total

impact of the discount is

highest under Scenario 2.


The total impact of a discount for LNG is significant because of its scale

The table below shows the impact of a 50% discount only offered to LNG

50% discount – LNG only

Scenario 119/20

23/24

Scenario 219/20

23/24

Scenario 319/20

23/24

Increase in

Moffat tariff

(€/MWh)

Moffat tariff

with

discounts

Impact of

discount per annum

NA

NA

NA

€122

NA

€41

NA

NA

NA

€31m p.a.

NA

€10m p.a

Moffat tariff

w/out

discounts

NA

NA

NA

€514

NA

€383

NA

NA

NA

€392

NA

€342

In scenario 2 the tariff for LNG

decreases from €254 to €207, the

differential therefore changes from

€138 to €307, an increase of €169

In scenario 3 the tariff for LNG





The total impact of a discount for small scale entry is significantly smaller than for LNG due

to the volume of bookings

The table below shows the impact of a 50% discount only offered to small scale entry

▪ In addition to the much smaller scale, the projected tariff for small scale entry is lower than for LNG so the discount also

has a smaller effect

50% discount – small scale entry only

Scenario 119/20

23/24

Scenario 219/20

23/24

Scenario 319/20

23/24

>€1

>€1

€14k p.a.

€122k p.a.

>€1

>€2

€14k p.a.

€426k p.a.

>€1

>€2

€14k p.a.

€308k p.a.

Increase in

Moffat tariff

(€/MWh)

Moffat tariff

with

discounts

€291

€265

€291

€392

€291

€342

Moffat tariff

w/out

discounts

Impact of

discount per annum

€291

€266

€291

€393

€291

€344

In this scenario the tariff for biogas













The use of multiplicative rescaling leads to higher Moffat tariffs than the use of additive

rescaling, leading to higher costs for consumers

We have estimated the total impact of multiplicative rescaling as the increase in the Moffat tariff resulting from switching

away from additive rescaling multiplied by the annual bookings. This assumes that Moffat will remain the marginal source of

gas.

Moffat tariff –

multiplicative

rescaling

Increase in

Moffat tariff

(€/MWh)

Impact of

multiplicative

rescaling p.a.

Scenario 1€65

€21

19/20

23/24

237 GWh

256 GWh

€15m p.a.

€5m p.a.

Scenario 2€65

€437

19/20

23/24

237 GWh

256 GWh

€15m p.a.

€112m p.a.

Scenario 3€65

€210

19/20

23/24

237 GWh

256 GWh

€15m p.a.

€54m p.a.

Annual

bookings

Moffat tariff –

additive

rescaling

€291

€265

€291

€392

€291

€342

€356

€286

€356

€829

€356

€553

Most significant impact

under scenario two

▪ Under multiplicative rescaling, the absolute tariff differential between entry points varies while the proportional difference

remains the same

▪ As a result, the higher an entry point’s primary tariff, the higher the absolute value of the rescaling is for that point.

▪ Moffat has the highest primary entry tariff and therefore faces a significantly larger increase under multiplicative than

additive rescaling









Page 23

Treatment of small-scale entry points

• There was support for a notional point approach as a pragmatic solution for small-scale entry points.

• We are interested in your views on the following:

• Should there be one or a small number of notional points and tariffs to retain some locational signal?

• Where should (the) notional point(s) be located?• Close to demand centre?

• Based on location of expected biogas entry?

• Some other approach?

• Implications for distribution tariffs may need to be considered, e.g. in relation to the treatment/ probability of interruption. However, this is not in scope of TAR NC.

Takeaways from NTLG and key remaining issues

Action: development of detailed arrangements for notional point(s) following feedback from today.


Treatment of small scale entry points

Transmission level charges

▪ At transmission, the NTLG feedback was that locational

signals are less important for small scale entry, and that

it would be better to avoid the administrative complexity

associated with setting multiple entry tariffs for each

individual site

▪ The following arrangements could follow from this:

Small scale entry tariffs should be based on a single

notional entry point, and all plant, regardless of

location, pay that single charge

The notional point will be chosen on the assumption

that small scale producers will choose to inject into

the grid close to demand

Distribution level charges

▪ The current tariff work does not need to establish exact

distribution charges. However, the following principles

could be applied:

Where the product offered to small scale producers is

identical – i.e. zero probability of interruption – then

the entry tariff should be the same across

transmission and distribution

Where the product is not the same – i.e. there is a >0

probability of interruption, a discount could be offered

for distribution entry tariffs.

There are two issues considered below:

▪ The treatment of small scale entry for the purposes of setting a transmission level entry tariff; and

▪ The treatment of small scale entry for the purposes of setting distribution level entry tariffs


Moving the notional small scale entry point from Corracunna to Gormanston lowers the

tariff by a small amount

Scenario 1

Scenario 2

Scenario 3€74

€125

19/20

23/24

Small scale entry

tariff at Gormanston

€74

€48

19/20

23/24

€74

€175

19/20

23/24

Small scale entry

tariff at Corracunna

€77

€51

€77

€178

€77

€128

The table below shows the impact of moving small scale entry from Corracunna to Gormanston – holding all other inputs

constant.


Distribution level charges

Transmission Distribution

Small scale tariffSmall scale tariff

50% Interruptibility discount

Scenario 2

19/20

23/24

€38

€89

€77

€178

Scenario 3

19/20

23/24

€38

€64

€77

€128

Scenario 1

19/20

23/24

€38

€25

€77

€51

An interruptibility

discount of 50% is

used in this

example.

The final discount

will depend on the

calculation of the

probability of

interruption.

The application of distribution level charges equal to the transmission tariff, but discounted for the probability of

interruption, would result in the tariffs in the table below.


Virtual Reverse Flow

27

Page 28

Virtual reverse flow

• Alignment with treatment of VRF as an interruptible product within the TAR NC.

• Some debate regarding whether the interruptible discount should be applied to the Moffat Entry tariff or Moffat Exit tariff.

• Stakeholders requested that the P-factor is analysed over 12 month period.

• Also debate regarding the level of discount. Arguments made for and against a relatively large discount were as follows:

Takeaways from NTLG 1-2

For Against

A high tariff (i.e. low discount) would lead the market to use swaps resulting in lower reverse and forward bookings

Value of VRF to shippers is high giving them access to a liquid NBP

Marginal costs of providing VRF service may be considered low

Ability to use VRF product is dependent on forward flow product existing

Page 29


• We intend to interpret VRF as an interruptible product.

• Treated as an exit product with tariff defined relative to the equivalent firm Exit tariff. Reflects fact that VRF represents (virtual) gas existing the system.

• GNI are developing analysis to propose the P-factor.

• We will consider an A-factor which reflects:

• the ‘economic value’ of the product;

• the impact on the market; and

• the impacts on other tariffs.

• The same principles will be applied to define an interruptible discount for Gormanston as for Moffat.

Intended way forward

Page 30


Consideration of absolute discount

Impact of relatively high tariff (low discount) Impact of relatively low tariff (high discount)

Could increase Exit revenue recovery, reducing Exit tariffs*

Little revenue recovery from the VRF product so Entry and Exit tariffs similar to under existing registration fee

But, if a high tariff incentivises a shift to swaps may reduce Entry revenue recovery (given reduction in forward flows), increasing Moffat Entry tariff

Encourages use of VRF product, and hence retains forward flow bookings relative to use of swaps

Greater use of swaps may increase IBP liquidity?

VRF product may help manage risk in the presence of low liquidity on the IBP – may attract new market participants

Provides a mechanism for market participants to self-balance reducing GNI balancing requirements

* Although current level of use of VRF product means impact could be small

Page 31

• Increasing liquidity of IBP may lead to a shift from VRF to swaps regardless of the level of the VRF tariff?

• However, in the short term, VRF product supports integration with the GB market?


Consideration of absolute discount









33

• Available data sourced from GNI GTMS system

• Periods used represent available data as per GTMS system

• Probability of interruption based on following formula:

• Two periods of analysis show a similar % of interruptible ‘days’

Probability of Interruption - Initial calculation

Assumptions

Number of days in period (Apr’18 – Aug’18) 144




Number of days where interruption(s) were experienced

Number of days where VRF allocations were recorded

Possible method of determining Interruptible %

Probability of Interruption - Additional Date Range

Number of days in period (May’17 – Nov’17) 210





VRF tariffs with no discount applied are significantly higher than

current VRF charges



& Shrinkage

35

Page 36


• On balance, stakeholders had concerns with the re-distributive impacts of moving away from the existing 90/10 split.

• GNI analysis demonstrated that customers with low load factors (e.g. electricity generators and residential consumers) in particular could face increased costs from a move to a higher capacity share of charges.

• An increase in the capacity element may also discourage certain consumers from shifting to gas use from other, less environmentally friendly fuels.


Justifiable bounds somewhere between 90/10 and

100/0 for Ireland

90% 100%0%

Page 37


• There was good discussion of interactions between the gas network tariff regime and the all-island electricity market at the last NTLG.

• For example, we heard from participants that may be will be more difficult for power generators to recover short-run marginal costs. Thus, a lower capacity element may be preferred by some stakeholders.

• We would like to understand implications of the move from SEM to I-SEM and interactions between capacity/commodity split and the electricity market further.

• Under I-SEM, how might gas network tariffs, in particular the capacity/commodity split, interact with:

• Bidding behaviour;

• Merit orders; and

• Market outcomes?

Interactions with electricity market

Page 38


• The Utility Regulator in Northern Ireland has consulted on a 95/5 capacity/commodity split.

• What impact might this have on the I-SEM and the merit order if Ireland adopts an different capacity/commodity split?

Interactions with electricity market – further discussion

Page 39

Shrinkage

• Shrinkage is currently not included within GNI’s allowed revenues which are recovered via transmission network capacity and commodity tariffs.

• Instead, GNI recover the costs of Shrinkage from Shippers on a monthly throughput basis (set out in the Code of Operations).

• Use of compressors on the transmission system and unaccounted for gas are both inherent costs of operating the gas transmission system. It may therefore be considered appropriate that all those who make use of the transmission system contribute to these costs.

• However, the approach for cost recovery of Shrinkage, and the interactions with transmission tariffs need to be considered.

Current arrangement

Action: Explore further how Shrinkage may be captured within the TAR NC

Page 40

Shrinkage

Option 1: Include shrinkage within commodity element of network tariff structure:

• Define as a transmission service under the TAR NC.

• Shrinkage recovered through existing commodity charge.

Consideration of role within TAR NC

Pros Cons

• Likely to be compliant with requirements of the TAR NC

• Would support justification of 90/10 capacity/commodity split.

• May introduce potential for under/over-recovery for GNI based on volatility of Shrinkage costs. *

• A large chunk of the commodity element may be driven by Shrinkage, effectively reducing the commodity element relative to the status quo.

* Also potential to increase the volatility of gas network tariffs through annual K-factor adjustments

Page 41

Shrinkage

Option 2: Have separate Shrinkage, Commodity and Capacity charges:

• Define as a transmission service under the TAR NC.

• Separate flow based charge for shrinkage.

• May be possible to retain monthly charging structure?


Pros Cons

• Dedicated flow-based Shrinkage charge which may allow status quo to broadly be retained.

• Dedicated (e.g. monthly) charge administered so as to mitigate volatility issues with Option 1.

• Need to consider whether approach strictly complies with TAR NC (e.g. Article 4).

• Role of 90:10 capacity/commodity split in wider tariff structure?

Page 42

Shrinkage

Option 3: Retain a separate charging approach as non-transmission service

• Define shrinkage as a non-transmission service under the TAR-NC.

• Separate flow based charge for shrinkage.

• Possible to retain monthly charging structure?


Pros Cons

• May allow status quo to be broadly retained.

• Unclear whether approach would comply with TAR NC (e.g. Article 4).

Page 43

Shrinkage

Option 4: Consider outside of allowed revenues

• There is a question about whether and how shrinkage should be captured within the TAR NC tariff arrangements.

• Another option might be to contend that in Ireland context shrinkage should not be contained within TAR NC transmission tariff requirements.


Pros Cons

• Would allow the status quo to retained

• Potentially highest risk of non-compliance if shrinkage is considered to be a transmission service

Page 44

Shrinkage

• Some of the options discussed above could lead to redistribution / incidence effects relative to current arrangements.

• For example, option 1 may lead to a greater allocation of network operation costs to high capacity / low throughput users compared to how GNI currently recover costs of Shrinkage from Shippers (monthly throughput basis).

• A number of the options above would also change the basis on which shrinkage costs are recovered from the market – how will this impact upstream / downstream contract arrangements?

Wider impacts of including shrinkage in gas network tariffs

We are interested in the views of the NTLG on the importance of these impacts

Page 45www.cru.ie

Multipliers & Seasonal Factors

45

Short-Term ProductsHistoric Usage

03/10/2018

• GNI were asked to present on the level of utilisation of short-term products in recent years

‒ The following slides outline the usage from 2014-15 to 2017-18

• GNI Looked at the absolute level of bookings and the weighted average (weighted by price)

‒ Also looked at the Power sector and Industrial and Commercial sector separately.

• The analysis will show a high level of usage over the period reviewed.

• The impact on different customer segments will need to be considered before transitioning to a new set of multipliers.

47

Introduction

Total Short-term usage October 2014 – Sep 2018

48

• Significant amounts of short-term

usage over that last four years

• Highest utilisation in the summer while

highest revenue generated in the

winter -see dotted lined

‒ This curve is influenced by the price of

short-term

• Next slides look at the Power and I&C

sectors separately -

2,000

4,000

6,000

8,000

10,000

12,000

-

200,000

400,000

600,000

800,000

1,000,000

1,200,000

1,400,000

1,600,000

1,800,000

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Tx Exit Cap - Total Annualised Short Term MWh

2015/16 2014/15 2016/17

2017/18 Average Weighted Average

Short-term usage _ POWER

• Significant use of short-term by power

all year round

‒ Summer and Shoulder periods showing

heavier usage in this analysis

• Short-term usage is dominated by

“Constrained On” plants.

• Variability of wind, Interconnector flows,

money point outages and weather will

have an impact on usage

• I-SEM may affect booking pattern.

49

-

200,000

400,000

600,000

800,000

1,000,000

1,200,000

1,400,000

1,600,000

1,800,000

2,000,000


Total Annualised Short Term at EXIT _ POWER MWh

2014/15 2015/16 2016/17 2017/18 Average

Short-term usage _Industrial & Commercial

• Average curve of I&C shows significant

usage of Short-term in the summer

• Key principle of the short-term

multipliers was to encourage summer

time usage

• Usage in the summer is driven by

particular load types e.g. Dairy Industry

‒ a number of dairy processing units

have connected to the gas networks in

recent years.

50

-

10,000

20,000

30,000

40,000

50,000

60,000

70,000


Total Annualised Short Term at EXIT _ LDM MWh

2014/15 2015/16 2016/17 2017/18 Average

• Significant amount of short-term used which can be seen from the analysis

• Any movement away from the current set of multipliers will need to occur on a phased basis

• Customer impact analysis will need to be undertaken to avoid price shocks

• I-SEM and the impacts on short-term bookings will need to be considered.

51

Summary

Page 52


• Stakeholders generally agreed with limiting the extent of change from the current approach.

• They queried the transition path towards the potential limits which may be set by ACER in 2023.

• They also suggested that quarterly multipliers may be the most important to consider.


Action: Intend to retain the existing approach but consider the points raised at the NTLG

Page 53


Transition

• Intend to align with immediate multiplier limits. I.e.

• 1.5 for monthly and quarterly product.

• 3.0 for daily product.

• 2023 multiplier limits will be dependent on an ACER recommendation in 2021.

• Multipliers can be revised annually under Article 28.

• Therefore, intention is to keep daily product multiplier limits under review and consider gradual transition to the potential 1.5 limits.

• Before further change, it may be preferable to monitor impacts of broader TAR NC change on the market.

Consideration of issues

Page 54


Quarterly multipliers

• The existing quarterly multipliers are simply an arithmetic mean of the relevant monthly multipliers within that quarter.

• They do not provide a strong incentive for quarterly bookings.

• We are therefore interested in your views on whether a stronger incentive for use of the quarterly product is needed.

• For example, a multiplier around 1.4 could perhaps be justified by the benefits of encouraging quarterly as opposed to monthly bookings.

Consideration of issues

There are currently very low bookings for the quarterly product

Page 55

WRAP UP ON FIRST HALF OF DAY

Page 56

Way forward

• Early November 2018: Consultation published

• Early January 2019: Consultation closes

• Early February 2019: Summary of responses published

• March 2019: Final decision published

• May 2019: Final tariffs published

Key dates

Thank you for your input










The TAR NC requires the chosen reference price methodology to be

compared to the CWDA with an entry/exit split of 50/50, as shown below


A comparison of the reference price methodology and the CWDA with

an entry/exit split of 33/67 can also be done, as shown below








7. Workshop model and methodologies


Disclaimer

This workshop will be organised as follows

Home page model (on screen)

Model inputs (on screen)

MA explained and worked example

MA applied in model (on screen)

CWDA explained and worked example

CWDA applied in model (on screen)


Disclaimer

This Consultation Transmission Tariff Model (including the enabling logic and input data contained therein)

is made available for the sole benefit of parties who are engaged in the Tariff Network Code consultation

process. GNI has and shall retain exclusive ownership of the Consultation Transmission Tariff Model and all

intellectual property rights in or relating to it, including the copyright and all other protected intellectual

property rights. The information contained in the Consultation Transmission Tariff Model is provided by

GNI “as is” and GNI reserves the right to amend the information at any time at its discretion. GNI makes

no representations or warranties of any kind, express or implied, in relation to the information and hereby

excludes all such representations or warranties, express or implied, to the fullest extent permitted by

law. GNI does not accept any responsibility, liability or duty of care to you or to any other person in

respect of the information, and any reliance you or any other person places on such information is

therefore strictly at your own or their own risk. In no event will GNI be liable for any loss or damage

including, without limitation, indirect or consequential loss or damage of any nature, arising out of or in

connection with the use of the Consultation Transmission Tariff Model or any information contained in it.

Model Disclaimer


1km 1km

1km

Entry point A

Forecasted cap : 50

Forecasted cap: 100

Entry point B

Exit point 1

Exit point 2

Forecasted cap: 90

Forecasted cap: 60

1.4 km

1km

We will use the following simplified network in our worked example

Shortest path Exit 1 Exit 2

Entry A 1 2

Entry B 1.4 1


We will go through the following steps of the calculation in turn:

The matrix approach captures the specific costs of each path (as much

as possible) but does not directly recover all allowed revenue

Calculate the costs of each path

Calculate entry and exit tariffs minimising the difference between the costs of the paths and the

sum of the entry and exit tariff. This results in primary tariffs

Adjust the tariff to recover all allowed revenue with secondary adjustments (and discounts)

Matrix


Calculate the costs of each path using Long Run Marginal Costs (Expansion constant and

annuitisation factor)

1km 1km

1km

Entry point A

Forecasted cap : 50

Forecasted cap: 100

Entry point B

Exit point 1

Exit point 2

Forecasted cap: 90

Forecasted cap: 60

1.4 km

1km

Costs of shortest path Exit 1 Exit 2

Entry A 1*€5 = €5 2*€5 = €10

Entry B 1.4*€5 = €7 1*€5 = €5

In this example, assume

every km of pipeline has a

cost of €5/GWh capacity per

annum

Matrix


Calculate entry and exit tariffs minimising the difference between the costs

of paths and the sum of both tariffs, i.e. best approximation of costs

Costs of shortest path Exit 1 Exit 2

Entry A 1*€5 = €5 2*€5 = €10

Entry B 1.4*€5 = €7 1*€5 = €5

Determine tariffs Costs Entry tariff Exit tariff Difference

costs and

tariffs

Squared

difference

Entry A to Exit 1 €5 €4.13 €2.62 €-1.75 €3.06

Entry A to Exit 2 €10 €4.13 €4.13 €1.75 €3.06

Entry B to Exit 1 €7 €2.62 €2.62 €1.75 €3.06

Entry B to Exit 2 €5 €2.62 €4.13 €-1.75 €3.06

Total €12.25

By using the squared

differences, positive and

negative deviations are

both considered and

large deviations are more

important than small

ones

Matrix

Let the computer find a

stable and unique

solution mimizinfg the

squared differences

Let the computer find a

unique solution

minimizing the squared

differences


Adjust the tariff to recover all allowed revenue

Point Primary tariff

Collected from

primary tariff Adder Tariff after adjustment

Collected revenue

after adder

Entry A € 4.13 € 412.50 € 3.04 € 7.17 € 716.67

Entry B € 2.62 € 131.25 € 3.04 € 5.67 € 283.33

Exit 1 € 2.62 € 157.50 € 3.14 € 5.77 € 346.00

Exit 2 € 4.13 € 371.25 € 3.14 € 7.27 € 654.00

Total € 1,072.50 Total € 2,000.00

It is unlikely that the primary tariffs (the tariffs calculated earlier) recover all allowed revenue. By adding a

fixed value to each of the entry (exit) tariffs, the differential between points is maintained but also all allowed

revenue to be recovered

Matrix

The regulator sets the amount of revenue recovered from entry and exit, assume €1000 from entry, €1000

from exit


The CWDA allocates allowed revenue across entry/exit points based on:

▪ the average distance gas travels if gas flows from an entry (exit) point to (from) all exit (entry) points proportionally to the demand (supply) at the exit

(entry) points

▪ the forecasted capacities at those points

We will go through the following steps of the calculation in turn:

The CWDA allocates allowed revenue across entry/exit points. It does not

directly account for costs of different routes but for the length of routes

Calculate capacity weighted average distances for all points using the forecasted booking

and the routes to the points

Calculate the proportion of distance travelled by gas to (from) a point relative to all

travelled distance (as assumed in this methodology)

Determine the revenue to be recovered from all exit and all entry points (50/50 is the

default counterfactual)

Use the proportion calculated in (2) to determine the revenue to be recovered from a

specific point. Tariffs are revenue per point divided by the bookings at that point

CWDA


Calculate capacity weighted average distances for all points using the

forecasted booking and the routes to the point

1km 1km

1km

Entry point A

Forecasted cap : 50

Forecasted cap: 100

Entry point B

Exit point 1

Exit point 2

Forecasted cap: 90

Forecasted cap: 60

1.4 km

1km

▪ Average distance entry point A = 1 km*(60/150) + 2 km * (90/150) = 1.60 km

▪ Average distance entry point B= 1.4 km*(60/150) +1 km *(90/150) = 1.16 km

▪ Average distance exit point 1 = 1 km *(100/150) +1.4 km *(50/150) = 1.13 km

▪ Average distance exit point 2 = 2 km *(100/150) +1 km *(50/150) = 1.66 km

The weights for entry are

based on exit point

bookings and visa versa

If injected gas at entry point A

were to serve the exit points in

proportion to their demands,

gas from entry point A would

travel 1.6 km on average

CWDA


Calculate the proportion of distance travelled by gas to (from) a point

relative to all travelled distance (as assumed in this methodology)

1km 1km

1km

Entry point A

Forecasted cap : 50

Forecasted cap: 100

Entry point B

Exit point 1

Exit point 2

Forecasted cap: 90

Forecasted cap: 60

1.4 km

1km

▪ AD EpA = 1*(60/150) + 2* (90/150) = 1.60 * (100/150) / ( 1.60* (100/150) + 1.16 * (50/150)) = 73.4%

▪ AD EpB = 1.4*(60/150) +1*(90/150) = 1.16 * (50/150) / ( 1.60* (100/150) + 1.16 * (50/150)) = 26.6%

▪ AD Xp1= 1*(100/150) +1.4*(50/150) = 1.13 * (60/150) / ( 1.13* (60/150) + 1.66 * (90/150)) = 31.2%

▪ AD Xp2 = 2*(100/150) +1*(50/150) = 1.66 * (90/150) / ( 1.13* (60/150) + 1.66 * (90/150)) = 68.8%

Average travel of gas from point A

and the share of supply from point A

Travel of gas from all entry

points weighted by supply

Proportion of

travel of gas

from point A

relative to all

travel

CWDA


Determine the revenue split between entry and exit and calculate a tariff for

each of the points

▪ Proportion EpA = 73.4% * €1000 = €734 so €734/100 bookings = €7.34 per booking

▪ Proportion EpB = 26.6%* €1000 = €266 so €266/50 bookings = €5.32 per booking

▪ Proportion Xp1 = 31.2%* €1000 = €312 so €312/60 bookings = €5.02 per booking

▪ Proportion Xp2 = 68.8%* €1000 = €688 so €688/90 bookings = €7.64 per booking

The regulator sets the amount of revenue recovered from entry and exit, assume €1000 from entry, €1000

from exit

1km 1km

1km

Entry point A

Forecasted cap : 50

Forecasted cap: 100

Entry point B

Exit point 1

Exit point 2

Forecasted cap: 90

Forecasted cap: 60

1.4 km

1km

Total revenue collected from point Tariff per booking

CWDA

Frontier Economics Ltd is a member of the Frontier Economics network, which consists of two separate companies based in Europe (Frontier Economics Ltd) and Australia (Frontier

Economics Pty Ltd). Both companies are independently owned, and legal commitments entered into by one company do not impose any obligations on the other company in the

network. All views expressed in this document are the views of Frontier Economics Ltd.

Networks Tariff Liaison Group Meeting€¦ · Bord Gáis Energy stated that they were supportive of the RPM as tariffs have been stable to date and worked as expected. SSE echoed

Documents