Gas transmission pricing models for entry-exit systems 6th Annual CRNI Conference, Brussels November...

Gas transmission pricing models for entry-exit systems6th Annual CRNI Conference, BrusselsNovember 22 – 2013Bert Kiewiet

ContentsThis paper discusses the major gas transmission pricing models. It explains how they allocate costs to entry/exit points. It assesses the applicability to network topologies.

Application of

four models to

two networks

Conclusions

Introduction to

entry-exit

systems

Four cost

allocation

models

Main objectives

of tariff design

Gas Transmission Pricing Models and Implementation Options

3

Introduction / The Entry-Exit SystemThe functionality of the entry-exit system is explained using a schematical representation.

Production

Storage

LNG

N X

X

Cross border

N

Directlyconnected customers

Storage

Cross border

Trading

VP

Local Local

TS

O l

evel

DS

O l

eve

l

X

N Physical entry point X Physical exit point Contractual flow of gas System boundary

Source: DNV KEMA 2013 – Entry-exit regimes in gasavailable on: http://ec.europa.eu/energy/gas_electricity/studies/gas_en.htm

4

Introduction / The Entry-Exit SystemOne of the main features is that network users contract entry and exit capacity separately.

Production

Storage

LNG

N X

X

Cross border

N


Storage

Cross border

Trading

VP

Local Local

TS

O l

evel

DS

O l

eve

l

X

Network users can contract entry and exit capacity separately.

5

Introduction / The Entry-Exit SystemAnother feature is that gas which has entered the system can be nominated to any off-take point.

Production

Storage

LNG

N X

X

Cross border

N


Storage

Cross border

Trading

VP

Local Local

TS

O l

evel

DS

O l

eve

l

X

Gas brought into the system at any entry point can be made available for off-take at any exit point within the system on a fully independent basis, without any restrictions

6

Introduction / The Entry-Exit SystemThe virtual point is a fundamental feature of the entry-exit model. It facilitates the bilateral title transfer of gas between network users.

Production

Storage

LNG

N X

X

Cross border

N


Storage

Cross border

Trading

VP

Local Local

TS

O l

evel

DS

O l

eve

l

X

The virtual trading point offers the users the possibility to bilaterally transfer title of gas and/or swap imbalances between network users.

7

Introduction / The Entry-Exit SystemIdeally the shipper books the exit capacity only at the network level where final exit takes place.

Production

Storage

LNG

N X

X

Cross border

N


Storage

Cross border

Trading

VP

Local Local

TS

O l

evel

DS

O l

eve

l

X

Network users only book exit capacity on the level where the final exit takes place. Imbalances between injections and withdrawals are aggregated across all entry and exit points in a network user’s portfolio, regardless of the network level.

8

Introduction / The Entry-Exit SystemThe functionality of the entry-exit system is explained using a schematical representation.

Production

Storage

LNG

N X

X

Cross border

N


Storage

Cross border

Trading

VP

Local Local

TS

O l

evel

DS

O l

eve

l

X

N Physical entry point X Physical exit point Contractual flow of gas System boundary

Source: DNV KEMA 2013 – Entry-exit regimes in gasavailable on: http://ec.europa.eu/energy/gas_electricity/studies/gas_en.htm

Introduction / EU Regulations

Regulation (EC) no. 715/2009:

Implementation of the entry-exit system is mandatory in the European Union.

Tariff setting in the entry-exit model:

- Tariffs should be set separately for every entry and exit point

- Tariffs should not be calculation on the basis of contract paths

- Non-discrimination between domestic transport and transit

Member States developed different solutions.

Harmonisation:

- ACER: Framework guidelines regarding harmonised transmission tariffs structures

- ENTSOG: Development of a network code on harmonised tariff structures (to do)

The entry-exit systems is the mandatory access model for gas transmission system operators in the European Union. Regulation provides overall requirements for tariffs.

Main Objectives Of Tariff Setting

Third Energy Package objectives:

Cost reflectivity

Non-discrimination

Avoid cross-subsidisation

Economic efficiency

Cost recovery

Transparency

The Third Energy Package specifies objectives for the setting of tariffs. Experience shows that other, more practical, aspects are important also.

Practical requirements:

Stability and predictability

Stakeholder acceptance

Efficient regulation

Macro-economic constraints

Transit network

Cost Allocation Models Applied To Two NetworksCost allocation models address the issue of different ways to allocate the allowed revenue to the specific entry and exit points. Four models are applied to two networks.

Cost allocation

Capacity Weighted Distance

Postage stamp

Matrix

Distance to Virtual Point

Entry 1

Entry 2

Exit 1

Exit 2

Exit 3

Entry 1

Entry 2

Entry 3

Exit 1

Exit 2

Exit 3

Ring shaped network

Entry 1

Entry 2

Entry 3

Exit 1

Exit 2

Exit 3

Some features

Different Keys For Allocation Of Costs To Entry-Exit PointsCost allocation models address the issue of different ways to allocate the allowed revenue to the specific entry and exit points. Four models are applied to two networks.

Cost allocation

Postage stamp

Matrix

· Most straightforward. · Single uniform tariff, does not provide locational signals· Likely to involve cross subsidies between network users.· Not particularly suitable for networks with longer

distances

· Uses replacement cost as a key to allocate revenue· Optimisation problem: additional constraints can easily be

applied.

Some features

Different Keys For Allocation Of Costs To Entry-Exit PointsCost allocation models address the issue of different ways to allocate the allowed revenue to the specific entry and exit points. Four models are applied to two networks.

Cost allocation

· Uses distance as a key to allocate costs· Based on the assumption that tariffs should reflect the

costs of bringing gas to the virtual point.· Takes into account the direction of the flow under peak

conditions. Negative distances = cost savings.· Resulting tariffs provide locational signals.

· Uses distance as a key to allocate costs as well. · Additional weighing by technical capacities of the entry

and exit points.



Postage Stamp – calculation steps

Features: Single uniform tariff applied to either the entry or exit points Does not provide any locational signals Not particularly suitable for transmission networks with longer distances Likely to involve cross subsidies between network users due to the uniform tariff level

The postage stamp is the most straightforward of all cost allocation methodologies. It does not provide any locational signals. It is less suitable for long distance networks.

2. Allocation of

Cost to Entry or

Exit points

1. Setting of

Allowed Revenue

· Starting point of the tariff calculation

· Set by regulator

· Costs are allocated to entry and exit points in proportion to the booked capacity

· Results in a uniform tariff for all points

Steps:

Matrix approach – calculation stepsThe matrix approach used replacement costs of sections as a key for the allocation of revenues.

4. Supplementary

Adjustments

3. Derivation of

Entry-Exit Tariffs

2. Allocation of

Cost to Pipeline

Sections

1. Setting of

Allowed Revenue


· Set by regulator

· Allocation of allowed revenue to different pipeline sections (external key for allocation: replacement costs of sections)

· Derivation of unit cost by incorporating chargeable capacity

· Construction of unit cost matrix.

· Calculation of tariffs by minimizing differences between unit costs and entry-exit tariffs.

Entry AB

Entry CB

Exit BD Exit EF Exit EG

UCAB+UCBD

UCCB+UCBD

UCAB+UCBE + UCEF

UCCB+UCBE + UCEF

UCAB+UCBE + UCEG

UCCB+UCBE + UCEG

Entry AB

Entry CB


TariffAB+ TariffBD

TariffCB+ TariffBD

TariffAB + TariffEF

TariffCB+ TariffEF

TariffAB+ TariffEG

TariffCB + TariffEG

· Tariffs adjustments to meet the requirements of being competitive, sustainable and affordable to network users and ensuring a successful transition.

y

x

Steps:

General calculation approach

Cost Allocation Models Applied To Two NetworksCost allocation models address the issue of different ways to allocate the allowed revenue to the specific entry and exit points. The models are applied to two networks.

Cost allocation


Postage stamp

Matrix


· Basic starting point is the allowed revenue· Arbitrarily chosen value for the allowed revenue· Focus is on the revenue allocation question· Only capacity charges are considered

(EUR/kWh/day/year)

Transit network


Entry 1 = 360 GWh/day

Entry 2 = 5 GWh/day

Exit 1 = 3 GWh/day

Exit 2 = 2 GWh/day

Exit 3 = 360 GWh/day

Entry 1

Entry 2

Entry 3

Exit 1

Exit 2

Exit 3

Ring shaped network


Entry 2 =

170 GWh/day


Exit 1

Exit 2

Exit 3

Results Transit NetworkThe four cost allocation models were applied to the transit network. The resulting (capacity) tariffs for the entry and exit points are presented.

1 2 1 2 3Entry Exit

0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

The main entry and exit point are priced


1 2 1 2 3Entry Exit

0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

The main entry and exit point are priced similarly for each model.


1 2 1 2 3Entry Exit

0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

In the distance to virtual point the exits closest to the main entry point is priced lowest:

• Value of chosen cost driver (total distance) is lower.


1 2 1 2 3Entry Exit

0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

This effect is also observed in the capacity weighted distance model (though less pronounced):

• In CWD this effect is lessened by the capacity as additional cost driver.


1 2 1 2 3Entry Exit

0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

The trend is different in the matrix approach:

• Both spur lines have same dimensions, but exit 2 has lower bookings unit cost higher.

Transit network



Entry 2 = 5 GWh/day

Exit 1 = 3 GWh/day

Exit 2 = 2 GWh/day

Exit 3 = 360 GWh/day

Entry 1

Entry 2

Entry 3

Exit 1

Exit 2

Exit 3

Ring shaped network


Entry 2 =

170 GWh/day


Exit 1

Exit 2

Exit 3

Results Ring Shaped NetworkThe four cost allocation models were applied to the ring shaped network. The resulting (capacity) tariffs for the entry and exit points are presented

1 2 3 1 2 3Entry Exit

0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]



0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

Tariffs of distance to virtual point = postage stamp:

• Due to symmetry in network topology.

• Distance to virtual point does not take into account differences in booked capacities.



0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

On the entry side the capacity weighted distance model = distance to virtual point model:

• Booked capacities at exit points are equal to each other.

• Network symmetry and thus equal distances.



0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

Exit 3 is priced highest:

• Furthest away from entry with highest booked capacity



0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

Matrix approach results in opposite outcomes:

• Exit tariffs are similar (booked capacity is same)

• Thus difference in booked capacity is now solely reflected in entry tariffs.



0

0.1

0.2

0.3

0.4

0.5

D2VPMatrixCWDPS

Ta

riff

[€

/kW

h/d

ay

/ye

ar]

Tariff entry 2 is lower:

• Due to higher booked capacity at entry 2 (lower unit costs).

• This effect discourages the use of the other entry points (undesired effect).

Concluding Remarks

For all four models the basic starting point is the allowed revenue of the network operator.

Postage stamp model:

Generally, the postage stamp model may be suitable in highly meshed networks where models resulting in locationally different tariffs might be too cumbersome

Application of equal tariffs does not affect cost-reflectivity too much Additionally relevant when tariff equality is prioritized over other principles

Distance to virtual point:

Distance as key for cost allocation Takes into account direction of the flow

Capacity weighted distance model:

Distance as key for cost allocation Capacity weighing

We have demonstrated four cost allocation models for calculating network tariffs. Results differ due to algorithms applied and the chosen cost drivers.

Concluding Remarks

Matrix approach:

Replacement cost as key to allocate revenues Allows for an easy incorporation of additional pricing considerations, for example:

- Equity requirements- Transition needs, and - Price stability

General:

The models are just different mathematical ways to describe reality and largely aim to achieve the same.

Results differ due to differences in the algorithms applied and in particularly the chosen cost drivers.

Trade-offs between the complexity of the model and its outcome- More complex networks may require more detailled modelling- Unilateral networks could be represented by less sophisticated modelling efforts

We have demonstrated four cost allocation models for calculating network tariffs. Results differ due to algorithms applied and the chosen cost drivers.

Contact

KEMA Nederland B.V. Energieweg 17, 9743 AN GroningenP.O. Box 2029, 9704 CA Groningen

The Netherlandswww.dnvkema.com

Bert KIEWIETSenior Consultant Markets & RegulationManagement & Operating Consulting

[email protected]: +31 50 700 98 69Fax: +31 50 700 98 59

33

Back-up slides

34

Derivation of Entry-Exit Tariffs

Example of applying the replacement value of pipeline sections to distribute the Allowed Revenue to be recovered by the various sections

Pipeline section

Length [km]

Diameter[inch]

Replac. value[mil. €]

AB 50 36 76,25

CB 100 36 152,50

BD 125 30 153,75

BE 200 30 246,00

Etc… Etc… Etc… Etc…

Entry

Entry Exit

Exit

Exit

Each pipeline section has its own replacement value

Stylized network:

AB

C

D

E

F

G

Unit cost[€/(Nm3/d)]

0,15

0,20

0,23

0,40

…

Capacity.[Nm3/d]

50 000 000

75 000 000

65 000 000

60 000 000

…

Allocation of Revenu [%/M€]

5% / 7,5

10% / 15

10,1% / 15,1

16,1% / 24,2

…

35


Unit cost matrix can be constructed by applying the shortest-path method

Tariffs from one network point to another (entry tariff + exit tariff) should equal (as much as possible) the unit costs of this route

Entry A

Entry C

Exit D Exit F Exit G

UCAB+UCBD

UCCB+UCBD

UCAB+UCBE + UCEF

UCCB+UCBE + UCEF

UCAB+UCBE + UCEG

UCCB+UCBE + UCEG

TariffAB+ TariffBD

TariffCB+ TariffBD

TariffAB + TariffEF

TariffCB+ TariffEF

TariffAB+ TariffEG

TariffCB + TariffEG

Unit cost matrix:

Tariffs:

Entry A

Entry C

Exit D Exit F Exit G

36


Values for sum of the entry and exit tariffs need to the same as the corresponding values of the unit cost matrix (cannot be solved algebraically)

This can be achieved by applying an Least Squares approach which results in the following minimization task:

- min ∑ij (Cij – (TNi + TXj))2

This minimization problem can be solved by using a numerical solver (for example using Excel’s Solver)

Usually, additional adjustments are necessary to attain the eventual tariffs:- Scaling to the required revenue- Incorporation of the main objectives in tariff setting (equity goals, stability and predictability,

transparency, etc.)

Distance To Virtual PointThe distance to the virtual point is based on the assumption that the entry and exit tariffs should reflect the costs of bringing gas to the virtual point.

4. Calculating

tariff

3. Calculate

distance to

reference node

2. Calculate flows

and direction

in network

1. Setting of

Allowed Revenue


· Set by regulator

· Definition of network sections

· Calculate flows at peak demand situation.

· Determine flow directions

· Reference node can be arbitrarily chosen

· Calculate distance from each point to reference node

· Distances may be negative (cost savings)

Entry AB

Entry CB


UCAB+UCBD

UCCB+UCBD

UCAB+UCBE + UCEF

UCCB+UCBE + UCEF

UCAB+UCBE + UCEG

UCCB+UCBE + UCEG

Entry AB

Entry CB


TariffAB+ TariffBD

TariffCB+ TariffBD

TariffAB + TariffEF

TariffCB+ TariffEF

TariffAB+ TariffEG

TariffCB + TariffEG

· Tariff = distance * constant unit cost of infrastructure

· Scaling tariffs in order to reach allowed revenue

y

x

Steps:

Distance To Virtual PointThe distance to the virtual point is based on the assumption that the entry and exit tariffs should reflect the costs of bringing gas to the virtual point.

Features: Costs are allocated to the different entry and exit points based on the distance to the virtual

point (reference node). The reference node can be arbitrarily chosen. Resulting tariffs provide locational signals.

Capacity Weighted DistanceUses distance as a key to allocate costs as well, but weighs them with the technical capacities of the demand/supply node.

4. Calculate tariffs

3. Calculate

capacity weighted

distance

2. Create a

distance matrix

1. Setting of

Allowed Revenue


· Set by regulator

· Create a matrix with the distances between every entry point and exit point.

· Calculate the proportion of the capacity of each entry/exit point relative to the total capacity

· Calculate capacity weighted distance for each entry point and exit point.

· Multiply distance by the share of capacity exit j in reltaion to total exit capacity

· For each point the revenue recovered is calculated.

· Dividing the revenue per point and the booked capacity yields the tariff

Steps:

40

Templates

41

Entry

Entry Exit

Exit

Exit

AB

C

D

E

F

G

Operational costs € / yearCapital costs € / yearReturn on assets € / year+

Allowed revenue € / year

42

Entry

Entry Exit

Exit

Exit

AB

C

D

E

F

G

Allowed revenue € / year

Booked capacity kWh/day/year

43

D =

D11

Dl1

D1j

DIJ

AD

Different keys for allocation of costs to entry-exit points

Postage stamp model:

Single uniform tariff applied to either the entry or exit points Does not provide any locational signals Likely to involve cross subsidies between network users due to the uniform tariff level Not particularly suitable for transmission networks with longer distances

Distance to Virtual Point Model: Uses distance as a key to allocate costs Takes into account the direction of the flow under peak conditions

Capacity Weighted Distance Model: Uses distance as a key to allocate costs as well, but weighs them with the technical capacities

of the demand/supply node

Matrix Approach Uses replacement cost as a key (and thus implicitly assumes length/diameter as major cost

drivers

x

Matrix approachThe matrix approach used replacement costs of sections as a key for the allocation of revenues.

Features: Replacement costs of sections used as key for allocation of revenue bears some similarity

with marginal cost pricing. Calculation of unit costs per segment:

- Low utilisation of sections may lead to higher unit costs for that section. May result in signals opposite to what would be desired.

- Alternatively a marginal cost concept can be applied. This approach will require additional adjustments t ensure revenue recovery.

Since this is an optimisation problem, additional constraints can be applied.

Gas transmission pricing models for entry-exit systems 6th Annual CRNI Conference, Brussels November...

Documents

Transcript of Gas transmission pricing models for entry-exit systems 6th Annual CRNI Conference, Brussels November...