Otsus

1

APPROVAL PAPER OF THE REGULATORY AUTHORITIES OF THE CONTINENTAL EUROPE

SYNCHRONOUS AREA

ON

THE ASSUMPTIONS AND METHODOLOGY FOR A PROBABILISTIC FCR DIMENSIONING IN THE

CONTINENTAL EUROPE SYNCHRONOUS AREA IN ACCORDANCE WITH ARTICLE 153(2) OF THE

COMMISSION REGULATION (EU) 2017/1485 OF 2 AUGUST 2017 ESTABLISHING A GUIDELINE ON

ELECTRICITY TRANSMISSION SYSTEM OPERATION

15 January 2025

2

I. Introduction and legal context

This document constitutes the agreement of the Regulatory Authorities of Continental Europe synchronous area (hereinafter referred to as: CE NRAs), as voted on 15 January 2025 on the Continental Europe TSOs’ (hereinafter referred to as: CE TSOs) proposal for the assumptions and methodology for probabilistic dimensioning of FCR (hereinafter referred to as: FCR probabilistic dimensioning) in accordance with Article 153(2) of the Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on Electricity Transmission System Operation as amended by Commission Implementing Regulation (EU) 2021/280 of 22 February 2021 (hereinafter referred to as: SO GL).

This agreement of the CE NRAs shall provide evidence that a decision on the FCR probabilistic dimensioning does not, at this stage, need to be adopted by ACER pursuant to Article 6(8) of SO GL. It is intended to constitute the basis on which the Regulatory Authorities will each subsequently approve the above-mentioned methodology pursuant to Article 6 of SO GL.

The legal provisions that lie at the basis of the FCR probabilistic dimensioning, and this CE NRAs agreement on the above-mentioned methodology, can be found in Articles 4, 6, 118, 153 and 156 of SO GL. They are set out here for reference.

SO GL Article 4 Objectives and regulatory aspects 1. This Regulation aims at:

(a) determining common operational security requirements and principles; (b) determining common interconnected system operational planning principles; (c) determining common load-frequency control processes and control structures; (d) ensuring the conditions for maintaining operational security throughout the Union; (e) ensuring the conditions for maintaining a frequency quality level of all synchronous areas

throughout the Union; (f) promoting the coordination of system operation and operational planning; (g) ensuring and enhancing the transparency and reliability of information on transmission

system operation; (h) contributing to the efficient operation and development of the electricity transmission system

and electricity sector in the Union. 2. When applying this Regulation, Member States, competent authorities, and system operators

shall: (a) apply the principles of proportionality and non-discrimination; (b) ensure transparency; (c) […] (d) […] (e) respect the responsibility assigned to the relevant TSO in order to ensure system security,

including as required by national legislation; […]

Article 6 Approval of terms and conditions or methodologies of TSOs 1. Each regulatory authority or where applicable the Agency, as the case may be, shall approve the

terms and conditions or methodologies developed by TSOs under paragraphs 2 and 3. The entity designated by the Member State shall approve the terms and conditions or methodologies developed by TSOs under paragraph 4. The designated entity shall be the regulatory authority unless otherwise provided by the Member State. Before approving the terms and conditions or

3

methodologies, the regulatory authority, the Agency or the designated entity shall revise the proposals where necessary, after consulting the respective TSOs, in order to ensure that they are in line with the purpose of this Regulation and contribute to market integration, non- discrimination, effective competition and the proper functioning of the market. (…)

2. (…) 3. The proposals for the following terms and conditions or methodologies and any amendments

thereof shall be subject to approval by all regulatory authorities of the concerned region, on which a Member State may provide an opinion to the concerned regulatory authority:

(a) (…) (b) (…) (c) (…) (d) methodologies, conditions and values included in the synchronous area operational

agreements in Article 118 concerning: i) (…) ii) the dimensioning rules for FCR in accordance with Article 153;

[…] 6. The proposal for terms and conditions or methodologies shall include a proposed timescale for

their implementation and a description of their expected impact on the objectives of this Regulation. Proposals for terms and conditions or methodologies subject to the approval by several regulatory authorities in accordance with paragraph 3 shall be submitted to the Agency within 1 week of their submission to regulatory authorities. Proposals for terms and conditions or methodologies subject to the approval by a designated entity in accordance with paragraph 4 may be submitted to the Agency within 1 month of their submission at the discretion of the designated entity while they shall be submitted upon the Agency’s request for information purposes in accordance with Article 3(2) of Regulation (EU) 2019/942 if the Agency considers the proposal to have a cross-border impact. Upon request by the competent regulatory authorities, the Agency shall issue an opinion within 3 months on the proposals for terms and conditions or methodologies.

7. Where the approval of the terms and conditions or methodologies in accordance with paragraph 3 or the amendment in accordance with Article 7 requires a decision by more than one regulatory authority pursuant to paragraph 3, the competent regulatory authorities shall consult and closely cooperate and coordinate with each other in order to reach an agreement. Where the Agency issues an opinion, the competent regulatory authorities shall take that opinion into account. Regulatory authorities or, where competent, the Agency shall take decisions concerning the submitted terms and conditions or methodologies in accordance with paragraphs 2 and 3 within 6 months following the receipt of the terms and conditions or methodologies by the Agency or the regulatory authority or, where applicable, by the last regulatory authority concerned. The period shall begin on the day following that on which the proposal was submitted to the Agency in accordance with paragraph 2 or to the last regulatory authority concerned in accordance with paragraph 3.

8. Where the regulatory authorities have not been able to reach an agreement within the period referred to in paragraph 7 or upon their joint request, or upon the Agency’s request according to the third subparagraph of Article 5(3) of Regulation (EU) 2019/942, the Agency shall adopt a decision concerning the submitted proposals for terms and conditions or methodologies within 6 months, in accordance with Article 5(3) and the second subparagraph of Article 6(10) of Regulation (EU) 2019/942.

[…] Article 118 Synchronous area operational agreements 1. By 12 months after entry into force of this Regulation, all TSOs of each synchronous area shall

jointly develop common proposals for: (a) the dimensioning rules for FCR in accordance with Article 153;

4

[…] 2. All TSOs of each synchronous area shall submit the methodologies and conditions listed in

Article 6(3)(d) for approval by all the regulatory authorities of the concerned synchronous area. Within 1 month after the approval of these methodologies and conditions, all TSOs of each synchronous area shall conclude a synchronous area operational agreement which shall enter into force within 3 months after the approval of the methodologies and conditions.

Article 153 FCR dimensioning 1. All TSOs of each synchronous area shall determine, at least annually, the reserve capacity for

FCR required for the synchronous area and the initial FCR obligation of each TSO in accordance with paragraph 2.

2. All TSOs of each synchronous area shall specify dimensioning rules in the synchronous area operational agreement in accordance with the following criteria: (a) the reserve capacity for FCR required for the synchronous area shall cover at least the

reference incident and, for the CE and Nordic synchronous areas, the results of the probabilistic dimensioning approach for FCR carried out pursuant to point (c);

(b) the size of the reference incident shall be determined in accordance with the following conditions:

i) for the CE synchronous area, the reference incident shall be 3 000 MW in positive direction and 3 000 MW in negative direction;

ii) (…) (c) for the CE and Nordic synchronous areas, all TSOs of the synchronous area shall have the

right to define a probabilistic dimensioning approach for FCR taking into account the pattern of load, generation and inertia, including synthetic inertia as well as the available means to deploy minimum inertia in real-time in accordance with the methodology referred to in Article 39, with the aim of reducing the probability of insufficient FCR to below or equal to once in 20 years; and

(d) the shares of the reserve capacity on FCR required for each TSO as initial FCR obligation shall be based on the sum of the net generation and consumption of its control area divided by the sum of net generation and consumption of the synchronous area over a period of 1 year.

Article 156 FCR provision […] 4. An FCR provider shall guarantee the continuous availability of FCR, with the exception of a

forced outage of a FCR providing unit, during the period of time in which it is obliged to provide FCR.

[…] 7. An FCR providing unit or FCR providing group with an energy reservoir that does not limit its

capability to provide FCR shall activate its FCR for as long as the frequency deviation persists. (…)

8. A FCR providing unit or FCR providing group with an energy reservoir that limits its capability to provide FCR shall activate its FCR for as long as the frequency deviation persists, unless its energy reservoir is exhausted in either the positive or negative direction. (…)

9. For the CE and Nordic synchronous areas, each FCR provider shall ensure that the FCR from its FCR providing units or groups with limited energy reservoirs are continuously available during normal state. For the CE and Nordic synchronous areas, as of triggering the alert state and during the alert state, each FCR provider shall ensure that its FCR providing units or groups with limited energy reservoirs are able to fully activate FCR continuously for a time period to be defined pursuant to paragraphs 10 and 11. […]

[…]

5

II. The Continental Europe TSOs’ proposal

The FCR dimensioning methodology is one of the terms & conditions and methodologies to be included in the Continental Europe Synchronous Area Operational Agreement (hereinafter referred to as: CE SAOA), in line with Article 118 of SO GL and to be subject to NRAs approval according to Article 6(3)(d) of SO GL.

The current version of the CE SAOA (included in the wider Continental Europe Synchronous Area Framework Agreement, hereafter referred to as CE SAFA) adopts a deterministic approach, posing the FCR equal to the reference incident that SO GL sets at 3.000 MW for the Continental Europe Synchronous Area, without requiring any probabilistic assessment. Adopting a probabilistic dimensioning approach, in fact, is a right of the CE TSOs, but it is not compulsory.

Due to the significant changes occurring in the electrical system with more and more renewable production with its own volatility, the CE TSOs developed a probabilistic approach for FCR dimensioning to be included in the CE SAOA instead of the current deterministic approach.

The FCR probabilistic dimensioning was consulted between 15 May 2023 and 15 June 20231 and it started being submitted by the CE TSOs to the CE NRAs in December 2023. As confirmed by ACER Decision 10/2024, the last NRA received it on 17 January 2024, which determined the initial deadline for an CE NRA agreement as 17 July 2024.

On 22 May 2024 the CE NRAs asked ACER for a 6-month’s prolongation of the legal deadline to reach an agreement on the FCR probabilistic dimensioning. The CE NRAs asked the CE TSOs for a study about Long Lasting Frequency Deviations (hereafter referred to as: LLFDs) within the scope of the ‘appropriate steps’ according to Article 5(9) of SO GL, defined by the CE NRAs towards the approval of the minimum delivery time period for limited energy reservoir FCR units (hereafter referred to as: Tmin LER). The LLFDs study was delivered in March 2024, but the CE NRAs needed a certain amount of time to review it, in particular to check whether the LLFDs dataset is reliable to be used for the FCR probabilistic dimensioning. Moreover, the proposal by the CE TSOs lacked a number of mathematical details that are needed to better understand the process. Checking the LLFDs and discussing the mathematical details required a certain level of interaction with the CE TSOs, impossible to achieve within the original deadline for reaching a decision on the FCR probabilistic dimensioning. ACER therefore granted an extension with its Decision 10/2024, setting the new deadline to 17 January 2025.

The FCR probabilistic dimensioning contains an iterative process to identify the symmetrical value of FCR. Starting from the reference incident (3.000 MW), several years are simulated applying a Monte Carlo approach, sampling some frequency deviations from three different datasets: deterministic frequency deviations, LLFDs and power imbalances due to outages (reported to a frequency deviation profile by simulating the Frequency Restoration Process with an equivalent single busbar model). By using deterministic formulas, for each frequency deviation the performances of the frequency transient in terms of Rate of Change of Frequency (hereafter referred to as: RoCoF), frequency nadir and frequency zenith are evaluated. These parameters are compared with the admissible thresholds identified by the CE TSOs: if the thresholds are violated no more than once in 20 years (as required by Article 153(2) of SOGL), the considered FCR value is the outcome of the dimensioning process and no further iterations are needed; otherwise FCR is increased and a new iteration is run.

1 The public consultation is available on the ENTSO-e website: https://consultations.entsoe.eu/system- operations/methodology-for-performing-the-probabilistic-dimen/

6

The FCR probabilistic dimensioning will be implemented within 12 months once approved by the CE NRAs.

In addition to the proposed timescale, the proposal includes also a description of its expected impact on the objectives of SO GL, in line with Article 6(6) of SO GL.

III. The Regulatory Authorities’ position

On the content of the TSOs’ proposal

Article 153(2) of SO GL allows the TSOs to adopt a probabilistic dimensioning approach, taking into account the pattern of load, generation and inertia, including synthetic inertia as well as the available means to deploy minimum inertia in real-time with the aim of reducing the probability of insufficient FCR to below or equal to once in 20 years.

The FCR probabilistic dimensioning proposed by the CE TSOs is consistent with the above mentioned provisions. While conducting the FCR probabilistic dimensioning process insufficient FCR is detected by looking at the main parameters (RoCoF, zenith and nadir) of the frequency transients while the ‘once in 20 years’ – criterion of Article 153(2) of SOGL is embedded in the process. The pattern of load, generation and inertia are indirectly considered by looking at frequency deviations datasets: the frequency trend is an effective measure of the volatility of the load and of the production, pointing out whether the system lacks or doesn’t lack inertia.

The CE NRAs had nonetheless the following concerns about specific aspects of the CE TSOs proposal:

Mathematical details

The FCR probabilistic dimensioning lacks all the mathematical details and thus its implementation cannot be properly monitored and, in case needed, enforced. Indeed, the CE TSOs provided all the details in the explanatory note, but this is not a legal document, hence it cannot be considered as a part of the methodology.

Triggering of the alert state and LER depletion

Article 156(9) of SO GL stipulates that the LER FCR providing units shall provide FCR indefinitely during the normal state, and for a limited amount of time (between 15 and 30 minutes to be set by the TSOs by the mean of a cost benefit analysis) as of the triggering of the alert state. The FCR probabilistic dimensioning proposed by the TSOs simulates the LER depletion as the frequency deviation exceeds 50 mHz, even if the alert state has not been triggered yet.

LLFDs classification

LLFDs datasets may include LLFDs that could be effectively mitigated by the CE TSOs and LLFDs that cannot be mitigated due to force majeure or political issues as in the Ukraine and Moldova LFC blocks or from imbalances in Kosovo. Having the roots cause of each LLFD included in the dataset is of utmost importance, in order to understand the positive effect of any mitigation measures to be implemented by the CE TSOs.

Transparency on the parameters

The FCR probabilistic dimensioning relies on different parameters adopted to identify the ‘once in a 20 years insufficiency criterion’ for FCR. These parameters are significant elements of the

7

methodology, but they are neither published nor at least sent to the CE NRAs before any instance of the process.

LER splitting

Article 153(2) of SO GL states that the initial FCR obligation shall be split among the TSOs proportionally to the net generation and consumption. The initial obligations are then adjusted into the binding FCR obligations by taking into account a number of situations as sharing and exchanging of FCR across the different LFC blocks and the outcome of the FCR cooperation process. The latter involves some of the CE TSOs on a voluntary basis in order to procure FCR with a pure market based approach. These aspects are left by the SO GL to the willingness of the TSOs, without any contribution or approval by the NRAs. Having a better knowledge of the process may nonetheless help the CE NRAs in evaluating the proper implementation of the FCR probabilistic dimensioning.

NRAs amendments

The CE NRAs deemed that the FCR probabilistic dimensioning needed to be significantly amended to accommodate the concerns mentioned above. To this extent the CE NRAs have exploited the provisions of Article 6(1) of SOGL, requiring the NRAs to revise the terms and conditions and methodologies, where necessary, after consulting the respective TSOs.

After a fruitful discussion with the CE TSOs, for the hearing phase the CE NRAs proposed the following main changes :

a) including some definitions to improve the quality and the overall readability of the proposal; b) developing a technical annex to include the mathematical details adopting as a basis the

explanatory note already provided by the CE TSOs; c) mandating the TSOs to send the CE NRAs a number of parameters before any instance of the

process and a report once the process is over; d) adding the possibility for NRAs to request a reassessment of the FCR dimensioning e) specifying that the LER depletion shall be simulated as of the triggering of the alert state as

stipulated by Article 156(9) of SO GL; f) including a classification of LLFDs into mitigable and non-mitigable ones; g) including the effect of the mitigation measures in the LLFDs dataset; h) asking the TSOs for a study to better understand how initial and binding FCR obligations are

assigned;

TSOs and stakeholders’ consultation

The CE TSOs expressed a wide acceptance of the changes listed in the letters a), b), c and d), since fruitfully discussed during the interaction with the CE NRAs.

For the other changes suggested by the CE NRAs, the CE TSOs proposed, instead:

i. keeping simulating the LER depletion when the frequency starts exceeding the 50 mHz deviations; the request cannot be accommodated since not consistent with the SO GL provisions; the reservoir amount shall be dimensioned accordingly by the LER providers;

ii. not introducing the classification of LLFDs; the request is accommodated since the classification is not functional to the FCR dimensioning process; nonetheless the possibility of introducing this (o a similar) classification will be rediscussed in the coming months as a part of the CE NRAs duties on monitoring the quality of the frequency;

iii. considering the LLFDs dataset as it is without any corrections to consider the effect of the mitigation measures; these measures are already indirectly taken into account since they are reflected in the most recent frequency samples that have the largest probabilities to be picked

8

up in the simulation process; the request cannot be accommodated, since the CE TSOs proposal would keep in the LLFDs dataset samples that are clearly outdated and unrealistic because of the new mitigation measures in place.

iv. not foreseeing any study on the FCR obligations, but keeping discussing the topic in dedicated meetings with the CE NRAs; the request is accommodated; the CE TSOs are requested to arrange a series of meetings with the CE NRAs to keep discussing the FCR obligation topic.

Some further editorial comments were provided by the CE TSOs during the hearing phase and accommodated by the CE NRAs.

IV. Conclusions

The CE NRAs have consulted and closely cooperated and coordinated with each other and with the relevant TSOs and ENTSO-E in order to amend and adopt the FCR probabilistic dimensioning submitted by the CE TSO as outlined above and annexed to this decision paper in both clean (Annex I) and track change (Annex II) versions.

The CE NRAs agree to issue their national approval decisions on the basis of this approval paper.

Assumptions and methodology for a probabilistic FCR

dimensioning in the Continental Europe synchronous area in

accordance with Article 153(2) of the Commission Regulation

(EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

Date: 15 January 2025

Assumptions and methodology for a FCR probabilistic dimensioning in accordance with Article 153(2) of the

Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on electricity transmission system

operation

2

Contents

Whereas ........................................................................................................................................ 3

Article 1 Subject matter and scope ...............................................................................................................3

Article 2 Definitions and interpretation ........................................................................................................4

Article 3 Outcome of the probabilistic FCR dimensioning ..........................................................................5

Article 4 FCR dimensioning criteria and process .........................................................................................5

Article 5 Probabilistic Simulation Model .....................................................................................................5

Article 6 Sources of power imbalances ........................................................................................................6

Article 7 Frequency acceptance criteria .......................................................................................................7

Article 8 Simulation scenarios .....................................................................................................................7

Article 9 Reporting .......................................................................................................................................7

Article 10 Publication and implementation of the probabilistic FCR dimensioning ....................................8

Article 11 Language .....................................................................................................................................8

Assumptions and methodology for a FCR probabilistic dimensioning in accordance with Article 153(2) of the

Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on electricity transmission system

operation

3

Whereas

(1) Article 153(2) of the Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a

guideline on electricity transmission system operation (hereafter referred to as “System Operation

Guideline” or “SO GL”) contains criteria that the Transmission System Operators (hereafter referred

to as “TSOs”) of each synchronous area shall follow when specifying the dimensioning rules for

Frequency Containment Reserve (hereafter referred to as “FCR”).

(2) According to Article 6(3)(d)(ii) of the SO GL, the dimensioning rules for FCR are subject to approval

by all regulatory authorities of the concerned region. Once approved these rules are included in the

synchronous area operational agreement. For the Continental Europe synchronous area this

agreement is part of the wider Synchronous Area Framework Agreement (hereinafter referred to as

“SAFA”) stipulated by the TSOs.

(3) The TSOs of the Continental Europe synchronous area have historically adopted a deterministic

criterion for the dimensioning of FCR. Such criterion considersconsider that the FCR shall be able

to contain a frequency deviation due to the worst expected outages combination in the system,

reflected by the ‘reference incident’ being equal to 3000 MW in both positive and negative direction,

pursuant to Article 153(2)(b) of the System Operation Guideline.

(4) For the Continental Europe synchronous area, Article 153(2)(c) of the SO GL states that the TSOs

of the Continental Europe synchronous area have the right to define a probabilistic dimensioning

approach for FCR, taking into account the pattern of load, generation and inertia, including synthetic

inertia as well as the available means to deploy minimum inertia in real-time in accordance with the

methodology referred to in Article 39 of the SO GL, with the aim of reducing the probability of

insufficient FCR to below or equal to once in 20 years.

(5) The probabilistic FCR dimensioning generally contributes to the achievement of the objectives of

Article 4(1) of the SO GL. Specifically, the probabilistic FCR dimensioning provides the TSOs of

the Continental Europe synchronous area with a methodology to evaluate the needs of FCR

considering all the relevant contributing factors. Such methodology contributes to the determination

of common operational security requirements and principles as set in Article 4(1)(a) of the SO GL.

It furthermore contributes to ensuring the conditions for maintaining operational security throughout

the Union as set in Article 4(1)(d) of the SO GL. Finally it contributes to ensuring the conditions for

maintaining a frequency quality level of all synchronous areas throughout the Union as set in Article

4(1)(e) of the SO GL. The probabilistic FCR dimensioning does not impact on the other objectives

listed in Article 4(1) of the SO GL.

(6) The probabilistic methodology for FCR dimensioning contributes to pursuing the general objectives

of the SO GL of safeguarding operational security by defining the proper FCR dimensioning needs.

Article 1 Subject matter and scope

1. The assumptions and methodology for the probabilistic FCR dimensioning represent the dimensioning

rules for FCR for Continental Europe synchronous area in accordance with Article 153(2) of the SO GL.

Assumptions and methodology for a FCR probabilistic dimensioning in accordance with Article 153(2) of the

Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on electricity transmission system

operation

4

Article 2 Definitions and interpretation

2. For the purposes of the probabilistic FCR dimensioning, terms used in this document shall have the

meaning of the definitions included in Article 3 of the SO GL.

3. Further, in theprobabilistic FCR dimensioning, unless the context requires otherwise, the following

definitions shall apply:

a) Critical Condition’ is a serie of minutes meeting one or more of the citeria for not acceptable minute

and spaced each other not more than a parametrical number of minute. b) ‘Deterministic frequency deviation’ or ‘DFD’ means regular deviations of the grid frequency that

occur around the hourly or sub-hourly intervals.

c) ‘Equivalent reservoir energy capacity’ means the energy requirement for LER associated to the Time

Period and shall amount to twice the energy provided by the full activation of LER for the Time

Period.

d) ‘FAT’ means ‘automatic FRR Full Activation Time’ as defined in Article 3 (101) of SO GL.

e) ‘frequency nadir’ is the minimum instantaneous frequency reached during an underfrequency

transient.

f) ‘frequency zenith’ is the maximum instantaneous frequency reached during an overfrequency

transient.

g) ‘Initial RoCoF’, is the RoCoF calculated at the time in which a disturbance happens.

h) ‘LER’ means ‘FCR providing units or groups with limited energy reservoirs’: FCR providing units

or FCR providing groups are deemed as with limited energy reservoirs in case a full continuous

activation for a period of 2 hours in either positive or negative direction might, without consideration

of the effect of an active energy reservoir management, lead to a limitation of its capability to provide

the full FCR activation.

i) ‘LER Share’ means the amount of LER in MW.

j) ‘Long lasting frequency deviation’ or ‘LLFD’ means an ‘event with an average steady state frequency

deviation larger than the long-lasting frequency threshold over a period longer than the time to restore

frequency.

k) ‘Long-lasting frequency threshold’ means a parameter used to identify Long lasting frequency

deviation.

l) ‘Market induced imbalances’ means the ‘generation-load imbalance caused by the change in

generation set points according to the results of the market scheduling’.

m) ‘Maximum Transient Frequency Deviation’ is the difference in absolute value between the frequency

at the time in which the disturbance happens and the frequency nadir for under-frequency or the

frequency zenith for over-frequency phenomena. It represents the maximum frequency excursion

before frequency starts to recover.

n) ‘Maximum Initial RoCoF’ is maximum RoCoF acceptable during a transient.

o) ‘RoCoF’, means Rate of Change of Frequency, is the derivative of the frequency.

p) ‘System droop’ means ‘the ratio between frequency deviation and steady state power response

provided by FCP’.

q) ‘Time Period’, means ‘the time for which each FCR provider shall ensure that its FCR providing

units or groups with limited energy reservoirs are able to fully activate FCR continuously, as of

triggering the alert state and during the alert state’ as determined according to Article 156(9) of the

System Operation Guideline.

Assumptions and methodology for a FCR probabilistic dimensioning in accordance with Article 153(2) of the

Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on electricity transmission system

operation

5

4. In this document, unless the context requires otherwise:

a) the singular indicates the plural and vice versa;

b) references to an “Article” are, unless otherwise stated, references to an Article of this document;

c) the table of contents and headings are inserted for convenience only and do not affect the

interpretation of the probabilistic FCR dimensioning; and

d) any reference to legislation, regulation, directive, order, instrument, code or any other enactment shall

include any modification, extension or re-enactment of it then in force.

Article 3 Outcome of the probabilistic FCR dimensioning

1. The outcomes of the probabilistic FCR dimensioning is a symmetrical value in MW for FCR for the

entire Continental Europe synchronous area in accordance with Article 153 of the System Operation

Guideline, computed according to the process described in Article 4.

Article 4 FCR dimensioning criteria and process

1. The symmetrical value for FCR for the entire Continental Europe synchronous area represents the

minimum amount of FCR needed in accordance with Article 153 of the SO GL, taking into account the

pattern of load, generation and inertia, including synthetic inertia as well as the available means to deploy

minimum inertia in real-time in accordance with the methodology referred to in Article 39 of the SO GL,

with the aim of reducing the probability of insufficient FCR to below or equal to once in 20 years.

2. The symmetrical value for FCR for the entire Continental Europe synchronous area is computed by the

mean of an iterative procedure as follows:

a) the process starts by setting a FCR value equal to the reference incident;

b) the FCR value is tested by the mean of the Probabilistic Simulation Model referred to in Article 5;

c) if the FCR is deemed sufficient according to the criteria in Article 7, the procedure stops, otherwise

the FCR value is increased by 100 MW and a new iteration is run;

d) the process continues until a sufficient FCR value is detected.

Article 5 Probabilistic Simulation Model

1. The Probabilistic Simulation Model simulates the behaviour of the whole Continental Europe

synchronous area in terms of frequency trends, testing the efficiency of the value of FCR in ensuring a

proper frequency quality according to the frequency acceptance criteria in Article 6. .

2. The Probabilistic Simulation Model shall implement a function to calculate the dynamic frequency

response consequent to a disturbance. Such function shall consider the variation in power imbalance

between two following calculation steps and calculate the key parameters of the frequency transient:

(frequency nadir, frequency zenith and RoCoF), along with the steady state frequency deviation

considering the system droop. The models parameters are tuned to provide the best equivalent behaviour

of the power system.

Assumptions and methodology for a FCR probabilistic dimensioning in accordance with Article 153(2) of the

Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on electricity transmission system

operation

6

3. The Probabilistic Simulation Model uses a Probabilistic Simulation Process in order to simulate several

years of operation conditions of the synchronous area by means of random draws of power imbalances

associated to DFDs, LLFDs, and outages of relevant grid elements. For each simulated year a power

imbalance trend is determined and the corresponding frequency deviation and relevant parameters are

computed according to the function described in paragraph 2.

The operation period to be simulated shall be estimated to generate statistically significant results and to

provide the best compromise among the desired level of accuracy and computational time efforts; in any

case at least 200 years shall be simulated.

The time discretization adopted by the Probabilistic Simulation Process shall be 1 minute. Each variable

shall thus be calculated on a 1-minute basis.

4. Input power imbalances deriving from DFDs and LLFDs are computed by the mean of an algebraic

relation simulating the steady state behaviour of the system.

5. Power imbalances associated to outages of relevant grid elements are determined simulating the FRP

with a single FRP controller without FRR limitations. The single FRP controller shall use a FAT

calculated as an average of the FAT of all the LFC areas belonging to the synchronous area weighted by

the FRR K-factors, until the FAT will be harmonized.

6. The annual review of FRR K-factors can be neglected as long as the review does not affect signicantly

the average FAT as defined in paragraph 5.

7. The Probabilistic Simulation Process can neglect the entire Cross-Border Load-Frequency Control

Process.

8. The Probabilistic Simulation Process shall be able to simulate the depletion of LER and its effects on the

frequency deviation, taking into account the LER Share and the Time Period. If an alert state is detected,

as of triggering the alert state and during the alert state, the depletion of the LER is simulated considering

that the energy content in the reservoir as of triggering the alert state allow the LER to fully activate FCR

continuously for a period equal to the Time Period.

9. More details about the Probabilistic Simulation Model are reported in the Annex.

Article 6 Sources of power imbalances

1. As detailed in the Annex and mentioned in Article 5(3), the Probabilistic Simulation Model shall take

into account:

a) Outages of relevant grid elements,

b) Deterministic frequency deviations (DFDs),

c) Long lasting frequency deviations (LLFDs).

2. For DFDs and LLFDs, the TSOs shall consider the market induced imbalances and analyse frequency

historical trends of the synchronous area over a number of years, as defined by the Continental Europe

TSOs according to Article 9.

Assumptions and methodology for a FCR probabilistic dimensioning in accordance with Article 153(2) of the

Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on electricity transmission system

operation

7

3. For outages of relevant grid elements the TSOs shall define a list of all the grid elements whose outages

lead to relevant power imbalances and indeed to relevant FCR activation.

Article 7 Frequency acceptance criteria

1. At each iteration, all Critical Conditions occurring in each simulated year are identified by checking

whether a serie of minutes, spaced each other not more than a parametrical number of minutes meets one

or more of the following criteria:

a) The Steady State Frequency Deviation exceeds the steady state maximum frequency deviation.

b) The frequency nadir or frequency zenith during a frequency transient exceeds the admissible

thresholds, as defined by the Continental Europe TSOs according to Article 9.

c) The absolute value of RoCoF exceeds the Maximum Initial RoCoF, as defined by the Continental

Europe TSOs according to Article 9.

2. The FCR considered is deemed sufficient when the number of identified Critical Conditions is less than

or equal to 1/20 of the number of simulated years. Such condition shall be fulfilled by the final

dimensioned FCR

Article 8 Simulation scenarios

1. The symmetrical value for FCR for the entire Continental Europe synchronous area is determined every

two years considering the best estimations of the input data regarding the evolution of sources of

frequency disturbances (taking into account the frequency management procedures implemented in the

meantime by the Continental Europe TSOs), the expected LER shares, their respective Time Period and

any other factor impacting the calculation and dimensioning of FCR.

2. In case there are significant changes in the input datasets, the TSOs may, on their own initiative,

redetermine the symmetrical value for FCR for the entire Continental Europe synchronous area even

before the two years period foreseen in paragraph 1.

3. The national regulatory authorities of the Continental Europe synchronous area have the right to send the

TSOs a coordinated request for the redetermination of the The symmetrical value for FCR for the entire

Continental Europe synchronous area.

Article 9 Reporting

1. Before each run of the FCR dimensioning process pursuant to Article 4, the TSOs shall provide the

national regulatory authorities of the Continental Europe synchronous area with the values, and

justifications for each value, of all the relevant thresholds adopted to assess the frequency acceptance

criteria of Article 7, and all the parameters described in the Annex.

2. The TSOs shall send to the national regulatory authorities of the Continental Europe synchronous area at

the end of each run of the FCR dimensioning process pursuant to Article 4 a report listing:

i. the mitigation measures considered in the LLFDs dataset and how they were taken into account;

Assumptions and methodology for a FCR probabilistic dimensioning in accordance with Article 153(2) of the

Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on electricity transmission system

operation

8

ii. the main parameters adopted to assess the frequency acceptance criteria and the reasons behind

their choice;

iii. the symmetrical value for FCR;

iv. the reasonings behind the choice to redetermine the symmetrical value for FCR in case such

redetermination occurs on initiative of the TSOs according to Article 8(2);

Article 10 Publication and implementation of the probabilistic FCR dimensioning

1. Each Continental Europe TSO shall publish the probabilistic FCR dimensioning without undue delay

after all the national regulatory authorities of the Continental Europe synchronous area have approved

the document, in accordance with Article 8 of the SO GL.

2. The Continental Europe TSOs shall have implemented the probabilistic FCR dimensioning within 12

months after the national regulatory authorities of the Continental Europe synchronous area have

approved the document.

3. Within 1 month from the approval of the FCR dimensioning by the national regulatory authorities of the

Continental Europe synchronous area, the Continental Europe TSOs shall organize a series of meeting

with the above mentioned regulatory authorities in order to keep discussing how the FCR obligation may

be identified in order to allocate more responsibilities to the LFC blocks causing the most significant

LLFDs.

Article 11 Language

1. The reference language for this methodology shall be English. For the avoidance of doubt, where TSOs

need to translate this methodology into their national language(s), in the event of inconsistencies between

the English version published by TSOs in accordance with Article 8(1) of the SO GL and any version in

another language, the relevant TSOs shall, in accordance with national legislation, provide the relevant

national regulatory authorities with an updated translation of the methodology.

Technical Annex of the methodology for performing the

probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU)

2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

Date: 15 January 2025

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

2

Contents

1 Acronyms and references ...................................................................................................... 3

2 Methodology for probabilistic approach for FCR dimensioning ............................................... 4

2.1 Overview and description of the methodology.................................................................................4

2.2 Functionality for DFDs statistics and DFDs random extractions .....................................................7

2.3 Functionality for LLFDs statistics and LLFDs random extractions .................................................7

2.4 Functionality of outages random extractions and calculation of associated power imbalances .......8

2.5 Functionality of combination of extracted DFDs, LLFDs and outages to generate global power

imbalance trends ...........................................................................................................................................9

2.6 Model to calculate the steady state frequency deviation in every minute. .....................................10

2.7 Model to calculate the dynamics of the frequency deviation in each minute. ................................12

2.8 Assessment of the acceptability criteria on the resulting simulated frequency deviation ..............14

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

3

1 Acronyms and references

ACE Area Control Error

CE Continental Europe

LER FCR providing units or groups with limited energy reservoir

FCR Frequency Containment Reserve

FCP Frequency Containment Process

FRR Frequency Restoration Reserve

FRP Frequency Restoration Process

FSM Frequency Sensitive Mode

Non-LER FCR providing units or groups without limited energy reservoir

NP RES Non Programmable Renewable Energy Sources

RES Renewable Energy Sources

SO GL System Operation Guideline

SA Synchronous Area

Tmin LER As of triggering the alert state and during the alert state, time for which each FCR

provider shall ensure that its FCR providing units with limited energy reservoirs are

able to fully activate FCR continuously.

FAT Full Activation Time of FRR

ROCOF Rate of Change of Frequency

[1] COMMISSION REGULATION (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation.

[2] ENTSO-E, SPD, “Frequency Stability Evaluation Criteria for the Synchronous Zone of Continental

Europe”, 2016.

[3] ENTSO-E, SPD – Inertia TF, “Inertia and Rate of Change of Frequency (RoCoF)”, 2020.

[4] ENTSO-E, “ENTSO-E HVDC Utilization and Unavailability Statistics 2021”, 2022.

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

4

2 Methodology for probabilistic approach for FCR dimensioning

2.1 Overview and description of the methodology

The methodology to perform the FCR probabilistic dimensioning required in the Continental Europe

Synchronous Area is based on a probabilistic model which randomly combines the most important source of

power imbalance in the system and simulate the resulting frequency deviation.

The model works on a large set of simulated years, in order to reach probabilistic significant results.

According to Article 153(2)(c) of SO GL, the probabilistic approach to FCR dimensioning shall be aimed at

reducing the probability of insufficient FCR to below or equal to once in 20 years.

Namely, whenever a power imbalance exceeds the available FCR, the FCR is considered insufficient. In

terms of frequency deviation, such condition results in a steady state frequency deviation larger than the

Maximum Steady State Frequency Deviation (at which FCR shall be fully deployed).

Moreover, since the available FCR impacts also the frequency transient following a sudden change in power

imbalance. “insufficient FCR” conditions are also those conditions where the frequency dynamic

performances are severely degraded. , i.e. those conditions whenthe frequency exceeds specific thresolds in

terms of Δf peak or ROCOF.

The purpose of the model is thus to determine the minimum amount of FCR which allows to ensure that the

insufficient FCR conditions (i.e., a Critical Conditions) occur not more often than once in 20 years.

A Critical Condition is a serie of minutes spaced each other not more than a parametrical number of minutes

and meeting one or more of following criteria:

a. The absolute value of Steady State Frequency Deviation (SSΔf) as simulated by the Probabilistic

Simulation Model exceeds the steady state maximum frequency deviation (200 mHz fro CE)

b. The absolute value of frequency peak reached during a transient exceeds the admissible tresholds .

c. The absolute value of ROCOF exceeds the Maximum Initial ROCOF.

Maximum Transient Frequency Deviation and Maximum Initial ROCOF are parameters defined by TSOs

and made publicly available before the execution of the methodology.

The model starts with the current deterministic FCR. The model then iterates increasing step by step the FCR

until the number of Critical Conditions in the simulated frequency deviation is such that they occur not more

often than once in 20 years.

The model takes into account the potential presence of LER (Limited Energy Reservoir FCR providers) in

the calculation of the results.

• The Probabilistic Simulation Model shall take into account:Outages on generation plants and HVDC

connections. Details on how the power imbalance is calculated are provided in Section 2.4.

• Power imbalance associated with Deterministic Frequency Deviations (DFDs). Details on how the

DFDs are calculated are provided in Section 2.2.

• Power imbalance associated with Long-Lasting Frequency Deviations (LLFDs). Details on how the

LLFDs are calculated are provided in Section 2.3.

DFDs and LLFDs are calculated starting from historical data of frequency deviations while the power

imbalances due to outages are derived from outages statistics.

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

5

An overall power imbalance is randomly generated from these three different sources of disturbance. Such

power disturbance is used to calculate simulated frequency deviation trends which are then analysed to verify

whether they fulfill the minimum acceptance criteria.

The whole model operates with a time granularity of one minute. Hence the power input power imbalance as

well as the simulated frequency deviations are trends with 525600 minutes a year (leap-year presence is

neglected).

The overview of the process is shown in the following Figure 1.

Figure 1: General overview of the model adopted for the probabilistic approach for FCR dimensioning

Figure 2 provides a more detailed depiction of how the input statistics (frequency, outages) are exploited.

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

6

Figure 2: Detailed overview of the model adopted for the probabilistic approach for FCR dimensioning

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

7

2.2 Functionality for DFDs statistics and DFDs random extractions

DFDs are market-induced frequency deviations which regularly occur around the change of the market time

unit (usually the change of hour).

In the model, the statistic of DFDs are directly calculated from the historical frequency trends with 1-minute

granularity. The model extracts the frequency around the change of hour: are considered as DFDs all the

frequency samples in a minutes range around the minute 0 (DFD interval).

For each simulated year, this functionality is aimed at calculating a trend of the frequency deviation due to

DFDs.

This trend is equal to 0 for all the minutes m which do not belong to the DFD interval. The minutes m

belonging to the DFD interval are taken from the input historical frequency trends.

DFDs are randomly selected for the input of the model, looking at homologous days in past years. For

example the DFDs to be assigned to 1st January of a simulated year are directly taken from the DFDs which

actually occurred in the system during the 1st January of a randomly selected past year (e.g. 2018). This

mechanism allows to keep the daily pattern of occurrence for DFDs: for example, the DFDs occurring around

6 am are taken from the same hour in the same day of another year.

The random choice of the year is biased towards most recent years. The probability of the past year is

indeed calculated with the following formula:

= 1

− −

Where:

is the most recent year for which data are available;

is the number of years for which historical trends are available.

The functionality results in a frequency trends composed by randomly extracted DFDs.

2.3 Functionality for LLFDs statistics and LLFDs random extractions

For the purpose of FCR dimensioning, the definition of long-lasting frequency deviations (LLFDs) is a

“condition with an average steady state frequency deviation larger than a share of the Standard Frequency

Range over a period longer than the Time To Restore Frequency”.

The tool scans the frequency trends acquired as input to detect all such conditions.

The scan operates following these rules:

• A moving average (with a width equal to Time to Restore Frequency) scans the data of a whole year.

• If the moving average frequency deviation exceeds a threshold equal to a share of the Standard

Frequency Range, a LLFD is detected.

• The LLFD length is calculated looking at its average frequency. The LLFD lasts as long as its average

frequency exceeds a share of the Standard Frequency Range. This average is calculated from the

beginning of the LLFD).

A list of all the detected LLFDs is created. Each LLFD is associated with the following information:

• Year of occurrence;

• Minutes in which it started;

• Duration;

• Frequency trend (vector of df characterizing the event)

These statistics are than exploited to generate a random extraction of LLFDs to be used as input by the model.

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

8

The procedure iterates on all the minutes of the year, asfollows:

1. It decides whether or not a LLFD starts at minute m.

This choice depends on the probability that a LLFD starts at the generic minute m of a day (e.g., at

02.15 PM). The latter probability is equal to the ratio between the number of LLFD starting in the

minute m (in the whole frequency dataset) and the number of days in the frequency dataset (365 ∗ ).

If no LLFD occurrence is extracted, the procedure proceeds analysing the following minute (m+1).

If a LLFD occurrence is extracted, the procedure proceeds at the step 2.

2. The year y from which to select a LLFD starting at minute m is randomly extracted. For this the

following probability is used:

, = 1

,

− −

,

Where , is the number of years for which at least one LLFD starting at minute m has been

detected and is the most recent year for which data are available;

3. The specific LLFD to be used is then chosen from the set of all LLFDs started at minute m and

occurred in the year y (chosen in the step 2). The random choice of the specific LLFD to be used is

based on an uniform distribution: all LLFDs in the set have the same probability to be chosen.

4. The selected LLFD is assigned to the trend. If the LLFD lasts for k minutes, the LLFD frequency

trend is assigned to the interval between minute m and minute m+k-1.

5. The procedure returns to step 1 for minute m+k.

The functionality results in a frequency trends composed by randomly extracted LLFDs.

2.4 Functionality of outages random extractions and calculation of associated

power imbalances

The outages are provided as input already in a statistical form: each potential event is associated with its:

• power loss: power change as of the event occurs;

• probability of occurrence: average number of events in a year.

The random extraction of outages uses as input the list of possible events

The extraction operates cycling on all the minutes of the year. For each minute m, all the possible events are

tested to verify whether they occur or not.

For each possible event v, a random value in [0, 1] is generated and it’s compared with the probability that

the event occurs in the minute (,):

, = 1 − −

365∗24∗60

Where : is the average number of occurrence in a year for a specific outage.

If the random generated value is below , the outage occurs. It means that the system must cope with the

power imbalance associated with the event.

The total amount of power imbalance in each minute is equal to the sum of the power imbalances of all the

events which are extracted in that minute.

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

9

The result of the calculation is a yearly power imbalance trend due to extracted outages.

FRR effects are applied to such calculated power imbalance yearly trend. The FRR is modelled as a simplified

1st order dynamic system. The power imbalances are brought to zero by FRR with a time constant equal to

1/3 of the FRR FAT.

After roughly 3 time constants the transient is ended, this condition simulates the restoring effects of FRR in

balancing the power imbalance due to outages within FRR FAT.

The following Figure 3 shows an example of the FRR effects on the power imbalances due to outages.

Figure 3: Example of FRR effects on power imbalance due to outages (15 min FAT is merely illustrative)

The functionality results thus in power imbalance trends due to outages and consequenct FRR activation.

2.5 Functionality of combination of extracted DFDs, LLFDs and outages to

generate global power imbalance trends

The combination of the input due to different sources takes place in terms of power imbalance: the power

imbalance due to outages is combined with the power imbalance corresponding to DFDs and LLFDs.

In order to convert the frequency deviation trends into equivalent yearly power imbalance trends, a converting

module is used. The module operates the conversion using a MW/Hz curve (given as input). In other words,

the frequency deviations due to DFDs and LLFDs are converted into power imbalances assuming the

conversion factor which was in place at the moment of their real occurrence. Such converting factor is the

MW/Hz dependency with a FCR equal to the value present in the year the data are referred to (e.g., 3000

MW for years up to 2024).. Such MW/Hz dependency doesn’t change during the iteration since it is related

to historical data trends.

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

10

The global power imbalance is obtained by summing the three power imbalances (due to LLFDs, DFDs, and

outages).

To avoid overlaps between DFDs and LLFDs the priority is given to LLFDs. LLFDs and DFDs are not

summed each other, but - on each minute - the presence of a LLFD overrides the presence of a DFDs.

2.6 Model to calculate the steady state frequency deviation in every minute.

This functionality progressively simulates system operation (in terms of frequency control) over the 525600

minutes of a year.

For each minute m it calculates the stady state simulated frequency deviation (∆ ) considering as input:

• The global power imbalance: ∆

• current regulating energy: . .

The regulating energy depends on the FCR amount in the current interation and on the possible exhaustion

of LER present in the FCR provision.

The output of the functionality is the simulated steady state frequency deviation trend (∆).

Such variable is modelled through a MW/Hz curve as shown in the example of Figure 4.

Figure 4: Example of using a MW/Hz curve of 3000 MW/Hz to calculate SSΔf from ΔPtot

A change in the regulating energy (. . ) leads to a different frequency deviation, starting from the same

power imbalance.

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

11

The standard regulating energy depends on the procured FCR. For instance, if in the current iteration a

condition with FCR = 3000 MW is considered, the standard regulating energy (. ) is equal to

15000 MW/Hz (i.e., 3000 MW of FCR with full activation at 0.2 Hz).

Should a LER depletion be detected, the regulating energy (. . ) decreases and the modelled curve has

to be rescaled.

When LER reservoirs are depleted, their FCR contribution is indeed considered as instantaneously lost (they

cannot provide anymore upward/downward regulation power).

Only the non-LER providers are still available to regulate the system. Given an input power imbalance, the

resulting frequency deviation is thus greater than in the situation with all the LER available.

This condition is modelled with a reduction of regulating energy (i.e., a rescale of the MW/Hz curve) equal

to the proportion of FCR lost due to the LER depletion. This proportion is the LER share.

For instance, if the LER share is 50%, once the LER are depleted the regulating energy is reduced by a factor

2 (the MW/Hz is rescaled by a factor 2). It means that the frequency deviation associated with a power

imbalance is doubled if compared to the standard condition.

The following Figure 5 shows the reduction in such example.

Figure 5: Example of rescale of MW/Hz curve by a factor 2

The model update at each minute m the current regulating energy (. .). The formula is:

. . = { . ∙ (1 − ℎ),

. . , (1)

To check whether the LER are depleted or not, the energy content of LER reservoir is calculated in each

minute.

The Figure 6 schematically shows the process by which the regulating energy is rescaled as a consequence

of the a LER depletion.

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

12

Figure 6: Schematic procedure for rescaling MW/Hz curve as a result of LER depletion

The combined effects of the recharging strategy and of the simulated frequency deviation can lead the LER

to recover from a depletion condition. As this occurs, the regulating energy returns to its standard condition

(e.g., 15000 MW/Hz if FCR = 3000 MW).

The LER are considered without energy limitations while frequency remains inside the Standard Frequency

Range.

If a continuous exceeding of the Standard Frequency Range includes the triggering of an alert state1, the

activated energy and the residual energy in the reservoir is calculated from the triggering of the alert state..

LER deplete as their reservoir reaches the maximum or minimum energy level. The capacity of the reservoir

depends on the minimum activation time period the LER are subject to.

2.7 Model to calculate the dynamics of the frequency deviation in each minute.

The characteristics of the frequency during a transient - such as the frequency peak (nadir or zenit) and the

ROCOF – need to be considered for the FCR dimensioning process (Figure 8).

1 An alert state is triggered if at least one of the following conditions occurs:

• The absolute value of simulated steady state frequency deviation exceeds for 5 consecutive minutes half of

the Maximum Steady State Frequency Deviation.

• The absolute value of simulated steady state frequency deviation exceeds for 15 consecutive minutes the

Standard Frequency Range.

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

13

Figure 7: Example of frequency transient characteristics and main performance indicators: zenit, nadir and RoCoF

Given the wide number of transient to be calculated for the dimensioning exercise, it is unfeasible to perform

an actual dynamic simulation in each single minute. There is therefore the need to adopt an algebraic

calculation of zenit/nadir and ROCOF starting from the aggregated single-busbar model depicted in Figure

8, based on considerations from [2].

Figure 8: Simplified single-busbar dynamic model of the CE power system

Where:

• Equation of motion: represents the response of the power systems in terms of inertia and self

regulation of load;

• Droop: represents the static response of the FCR (see Figure 4);

• Equivalent dynamic of FCR provision: represents the average combined effects of the dynamic

responses provided by all FCR providers.

The parameters presented in Figure 8 are thus:

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

14

=

1

. . ∙

[/∆] Droop in pu (. . is associated with a

certain MW/Hz curve and it’s expressed in [MW/Hz])

1 [] Pole time constant of average FCR dynamics

2 [] Zero time constant of average FCR dynamics

=

[/∆] Self-regulation of load ( is expressed in

[pu/Hz])

= 2 ∙ [] System equivalent angular momentum

(2*Inertia)

[] Load at SA level

Nominal frequency (50 Hz)

The output of the diagram (∆) is the frequency deviation in pu.

Considering the actual and complex dynamics of the SA, with this model significant approximations are

introduced, since each provider (and each technology) has its own peculiarities when it comes to the FCR

deployment dynamic. Such variety of responses is simplified with a single 2nd order dynamic model in order

to derive the algebraic formulas for Zenit/Nadir and ROCOF. The ROCOF is evaluated as the initial ROCOF.

Such formulas are derived assuming that a stepwise disturbance is applied on the model presented in Figure

8.

In this way an algebraic relationship between the disturbance and the system parameters can be used within

the iterative probabilistic model.

The calculation of dynamic performances of frequency deviations are based on the same 1-minute granularity

adopted for the steady state calculations. It means that all the variables (e.g., power imbalance and steady-

state frequency deviation) continue to change minute-by-minute.

Both the transient frequency peak (zenit/nadir) and the ROCOF are therefore calculated on a 1-minute basis.

The input of such calculation is the difference of power imbalance between two following minutes.

2.8 Assessment of the acceptability criteria on the resulting simulated frequency

deviation

A FCR dimensioning shall be considered acceptable if ensures that the FCR is insufficient not more often

than once every 20 years.

The first step is to assess whether a specific minute is considered an acceptable minute. A minute is considered

an acceptable minute if it fulfills all the following three criteria:

• The absolute value of the simulated steady state frequency deviation does not exceed the steady state

maximum frequency deviation;

Technical Annex to the methodology for performing the probabilistic dimensioning of FCR in CE synchronous area

according to Article 153(2) of Commission Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on

electricity transmission system operation

15

• The absolute value of the maximum/minimum instantaneous frequency deviation during transients

doesn’t exceed the thresholds defined by the TSOs;

• The absolute value of the ROCOF does not exceed the Maximum Initial ROCOF as defined by the

TSOs?

A minute is considered a not acceptable minute if at least one criterion is not fulfilled.

To interpret the «once in 20 years» criterion, the concept of “Critical Condition” is then introduced: a Critical

Condition is a series of not acceptable minutes spaced each other not more than a parametrical number of

minutes (e.g., 15 minutes).

A single Critical Condition could then be made by several following minutes with one or more criteria not

fulfilled.

The choice of such approach is related to the fact that the combination of disturbances causing a condition

where one or more criterion (SSΔf / zenit/nadir / ROCOF) are not fulfilled could persists for several minutes.

The «once in 20 years» criterion is applied on the number of Critical Conditions rather than on single minutes.

The FCR dimensioning is thus aimed at ensuring that the number of detected Critical Conditions is less or

equal to 1/20 of the number of simulated years.

E.g., if 200 years are simulated by the model, no more than 10 (200/20) Critical Conditions shall occur.

Tatari 39 / 10134 Tallinn / 667 2400 / [email protected] / Registrikood 70000303

OTSUS

01.08.2025 nr 7-26/2025-003

Mandri-Euroopa sünkroonala põhivõrguettevõtjate metoodika muudatuste

kooskõlastamise kohta kooskõlas Komisjoni määruse (EL) 2017/1485, 2. august 2017,

millega kehtestatakse elektri ülekandesüsteemi käidueeskiri, artikli 153 lõikega 2

1. Haldusmenetluse alustamine

Euroopa Komisjoni Määrusega (EL) 2017/1485 kehtestati elektri ülekandesüsteemi

käidueeskiri (edaspidi ka SOGL või SOGLi määrus)1.

SOGLi määruse artikli 6 lõike 1 alusel kiidab iga reguleeriv asutus või vajaduse korral

koostööamet heaks tingimused või meetodid, mis põhivõrguettevõtjad on lõigete 2 ja 3 kohaselt

koosatanud. Liikmesriigi määratud asutus kiidab heaks lõike 4 kohased põhivõrguettevõtjate

väljatöötatud tingimused või metoodika. Kui liikmesriik ei ole sätestanud teisiti, on määratud

asutus reguleeriv asutus. Enne tingimuste ja meetodite heakskiitmist vaatab reguleeriv asutus,

koostööamet või määratud asutus pärast asjaomaste põhivõrguettevõtjatega konsulteerimist

vajaduse korral ettepanekud läbi, et tagada nende kooskõla käesoleva määruse eesmärgiga ning

aidata kaasa turgude lõimimisele, mittediskrimineerimisele, tõhusale konkurentsile ja turu

nõuetekohasele toimimisele.

SOGLi määruse artikli 6 lõike 3 punkti d) alapunkti ii) alusel peab artikli 118 kohastes

sünkroonala talitluslepingutes sisalduv sageduse hoidmise reservi (ingl frequency containment

reserves ehk FCR) suuruse määramise eeskirjade ettepanek või muudatus kooskõlas SOGL

määruse artikliga 153 saama kõigi vaatlusaluse piirkonna2 reguleerivate asutuste heakskiidu.

SOGLi määruse artikli 6 lõike 6 järgi peab tingimuste või meetodite ettepanek hõlmama nende

rakendamise esialgset ajakava ning kirjeldust selle kohta, kuidas need võiksid mõjutada SOGLi

määruse eesmärkide saavutamist. Mitme reguleeriva asutuse poolt lõike 3 kohaselt heaks

kiidetavate tingimuste või meetodite ettepanekud esitatakse koostööametile ühe nädala jooksul

pärast nende esitamist reguleerivatele asutustele. Määratud asutuse poolt lõike 4 kohaselt heaks

kiidetavate tingimuste või meetodite ettepanekud võib esitada koostööametile ühe kuu jooksul

pärast nende esitamist määratud asutuse äranägemisel, kuid juhul, kui koostööamet leiab, et

ettepanekul on piiriülene mõju, esitatakse need koostööameti teavitamiseks vastavalt määruse

(EL) 2019/942 artikli 3 lõikele 2. Pädevate reguleerivate asutuste taotluse korral väljastab

koostööamet kolme kuu jooksul arvamuse tingimuste või meetodite alaste ettepanekute kohta.

1 Kättesaadav: https://eur-lex.europa.eu/legal-content/ET/TXT/HTML/?uri=CELEX:02017R1485-20210315 2 Antud juhul kõigi Mandri-Euroopa sünkroonala riiklike reguleerivate asutuste heakskiidu.

2 (5)

SOGLi määruse artikli 6 lõike 7 järgi, kui tingimuste või metoodika heakskiitmiseks vastavalt

lõikele 3 või nende muutmiseks vastavalt artiklile 7 on vaja rohkem kui ühe reguleeriva asutuse

otsust vastavalt lõikele 3, peavad pädevad reguleerivad asutused kokkuleppele jõudmiseks

omavahel konsulteerima, tegema tihedat koostööd ja kooskõlastama oma tegevuse. Kui

koostööamet esitab oma arvamuse, võtavad pädevad reguleerivad asutused seda arvesse.

Reguleerivad asutused või vajaduse korral koostööamet teevad lõigete 2 ja 3 kohaselt esitatud

tingimuste ja meetodite kohta otsuse kuue kuu jooksul pärast seda, kui koostööamet või

reguleeriv asutus või asjakohasel juhul viimane asjaomane reguleeriv asutus on tingimused või

meetodid kätte saanud. Ajavahemik algab järgmisel päeval pärast seda, kui ettepanek esitati

koostööametile vastavalt lõikele 2 või viimasele asjaomasele reguleerivale asutusele vastavalt

lõikele 3.

SOGLi määruse artikli 6 lõike 8 järgi, kui reguleerivad asutused ei ole jõudnud kokkuleppele

lõikes 7 osutatud ajavahemiku jooksul või nende ühise taotluse korral või koostööameti

taotlusel vastavalt määruse (EL) 2019/942 artikli 5 lõike 3 kolmandale lõigule, võtab

koostööamet kuue kuu jooksul vastavalt määruse (EL) 2019/942 artikli 5 lõikele 3 ja artikli 6

lõike 10 teisele lõigule vastu otsuse tingimuste või meetodite ettepanekute kohta.

SOGLi määruse artikli 153 lõike 1 järgi määravad iga koordineeritud võimsusarvutusala kõik

põhivõrguettevõtjad vähemalt üks kord aastas sünkroonalal vajatava FCRi reservvõimsuse ning

iga põhivõrguettevõtja esialgse FCRi kohustuse kooskõlas lõikega 2.

SOGLi määruse artikli 153 lõike 2 alusel kehtestavad iga sünkroonala kõik põhivõrguettevõtjad

sünkroonala talitluslepingus suuruse määramise juhised vastavalt järgmistele kriteeriumidele:

a) sünkroonala FCRi reservvõimsus peab katma vähemalt võrdlusjuhtumi vajaduse ning

Mandri-Euroopa sünkroonala ja Põhjamaade sünkroonala puhul tõenäosuslikul meetodil

vastavalt punktile c määratud FCRi suuruse;

b) võrdlusjuhtum tuleb sätestada vastavalt järgmistele tingimustele:

i. Mandri-Euroopa sünkroonala puhul vastab võrdlusjuhtumile 3 000 MW positiivses

suunas ja 3 000 MW negatiivses suunas;

c) Mandri-Euroopa sünkroonala ja Põhjamaade sünkroonala kõik põhivõrguettevõtjad

kehtestavad tõenäosusliku meetodi FCRi jaoks, võttes arvesse tarbimise ja tootmise

suundumusi, inertsi ja tehisinertsi muutumist ning seda, millised on võimalused reaalajas

kasutada väikseimat inertsi vastavalt metoodikale, millele on osutatud artiklis 39, et

vähendada FCRi mittepiisavuse tõenäosust kuni ühe juhtumini 20 aasta jooksul, ja

d) see osa FCRi reservvõimsusest, mis on nõutav igalt põhivõrguettevõtjalt esialgse

kohustusena, peab põhinema selle põhivõrguettevõtja ala netotootmise ja -tarbimise

summal, mis on jagatud vastava sünkroonala aastase ajavahemiku netotootmise ja -

tarbimise summaga.

2. Menetlusosaline

Elering AS, äriregistri kood 11022625, asukoht Kadaka tee 42, Tallinn, 12915, e-post:

[email protected].

3. Asjaolud ja menetluse käik

17.06.2025 esitas Elering AS (Elering) Konkurentsiametile kooskõlastamiseks taotluse kõikide

Mandri-Euroopa sünkroonala põhivõrguettevõtjate (edaspidi ka põhivõrguettevõtjad) ühise

3 (5)

metoodika muutmist käsitleva ettepaneku FCRi suuruse määramise eeskirjade kohta Mandri-

Euroopa sünkroonala jaoks, koos tehnilise lisaga, kooskõlas SOGLi määruse artikli 153 lõikega

2 (metoodika).

Elering toob taotluses välja, et metoodika muutmist käsitleva ettepanek on Mandri-Euroopa

riiklike reguleerivate asutuste poolt heaks kiidetud 15.01.2025.

Seoses asjaoluga, et Eesti elektrisüsteemi sünkroniseerimine Mandri-Euroopa

elektrisüsteemiga toimus 09.02.2025, ei olnud Konkurentsiamet kaasatud sünkroniseerimisele

eelnenud Mandri-Euroopa riiklike reguleerivate asutuste metoodikate heakskiitmisprotsessi.

Siiski on Mandri-Euroopa sünkroonalaga liitudes vajalik kooskõlastada Mandri-Euroopa

sünkroonala metoodikad, mis kehtisid sünkroniseerimise kuupäeval, kuna, kui

põhivõrguettevõtja liitub sünkroonalaga, muutub ta vastava sünkroonala talitluslepingu

osaliseks ning peab lähtuma sünkroonalas kehtivatest metoodikatest.

Põhivõrguettevõtjad on kooskõlas SOGLi määruse artikliga 11 metoodika muutmist käsitlevat

ettepanekut avalikult konsulteerinud Euroopa põhivõrguettevõtjate koostööorganisatsiooni

(ENTSO-E) konsultatsioonikeskkonnas perioodil 15.05.2023 kuni 15.06.20233.

25.07.2025 pöördus Konkurentsiamet Itaalia reguleeriva asutuse esindaja poole, kes vastutab

Mandri-Euroopa sünkroonala talitluslepingutes sisalduvate meetodikate ja tingimuste

koordineerimise protsessi eest. Pöördumises palus Konkurentsiamet edastada Mandri-Euroopa

sünkroonala riiklike reguleerivate asutuste ühise seisukoha, millega kiideti heaks metoodika

muutmist käsitlev ettepanek.

25.07.2025 vastas Itaalia reguleeriva asutuse esindaja, kinnitades, et metoodika muutmist

käsitlev ettepanek on asjaomaste riiklike reguleerivate asutuste poolt 15.01.2025 muudetud

kujul (Lisa 1), koos tehnilise lisaga (Lisa 2) heaks kiidetud ning edastas Mandri-Euroopa

sünkroonala riiklike reguleerivate asutuste ühise seisukoha metoodika muutmist käsitleva

ettepaneku heakskiitmise kohta (Lisa 3).

4. Asjaomaste reguleerivate asutuste ühine seisukoht

SOGLi artikkel 153 lõige 2 lubab põhivõrguettevõtjatel rakendada tõenäosuslikku meetodit

FCRi suuruse määramise jaoks võttes arvesse tarbimise ja tootmise suundumusi, inertsi ja

tehisinertsi muutumist ning seda, millised on võimalused, et vähendada FCRi mittepiisavuse

tõenäosust kuni ühe juhtumini 20 aasta jooksul.

Mandri-Euroopa sünkroonala riiklikud reguleerivad asutused on ühisel seisukohal, et Mandri-

Euroopa põhivõrguettevõtjate poolt esitatud ettepanek FCRi tõenäosusliku meetodi kohta on

kooskõlas ülaltoodud SOGLi määruse sätetega. FCRi suuruse määramise tõenäosuslikus

meetodis tuvastatakse FCRi mittepiisavus, vaadeldes sagedusüleminekute põhiparameetreid

ning muuhulgas on protsessi sisse viidud SOGLi määruse artikli 153 lõikest 2 tulenev

kriteerium FCRi mittepiisavuse tõenäosuse kohta vähendada seda kuni ühe juhtumini 20 aasta

jooksul. Koormuse, tootmise ja inertsuse mustreid käsitletakse kaudselt sagedushälvete

andmekogumite kaudu. Sageduse trend on tõhus meede koormuse ja tootmise volatiilsuse

hindamiseks, näidates, kas süsteemis esineb või ei esine inertsuse puudust.

3 Kättesaadav: https://consultations.entsoe.eu/system-operations/methodology-for-performing-the-probabilistic-

dimen/

4 (5)

Siiski esines Mandri-Euroopa sünkroonala riiklike reguleerivate asutuste hinnangul murekohti

ja puuduseid põhivõrguettevõtjate metoodika muutmist käsitlevas ettepanekus matemaatiliste

üksikasjadega, häireseisundi käivitamisega ning piiravate energiasalvestite tühjenemisega ja

jagamisega, pikaajaliste sagedushälvete klassifikatsiooniga ning parameetrite läbipaistvusega.

Mandri-Euroopa sünkroonala riiklikud reguleerivad asutused leidsid, et FCRi suuruse

määramise tõenäosuslik meetod vajab märkimisväärseid muudatusi, et kõrvaldada metoodika

põhimõtetes eelnevalt viidatud puudused. Seega otsustasid asjaomased riiklikud reguleerivad

asutused enne metoodika ettepaneku heaks kiitmist seda SOGLi määruse artikli 6 lõike 1 alusel

muuta.

Pärast arutelu asjaomaste põhivõrguettevõtjatega otsustasid Mandri-Euroopa sünkroonala

riiklikud reguleerivad asutused viia metoodikasse sisse (kokkuvõtlikult) järgmised

muudatused:

a) mõnede definitsioonide lisamine, et parandada ettepaneku kvaliteeti ja üldist loetavust;

b) tehnilise lisa koostamine, mis sisaldab matemaatilisi üksikasju, kasutades aluseks

Mandri-Euroopa põhivõrguettevõtjate poolt juba esitatud selgitavat dokumenti;

c) lisada nõude, et põhivõrguettevõtjad saadaksid enne iga protsessi läbiviimist Mandri-

Euroopa sünkroonala riiklikele reguleerivatele asutustele teatud parameetrid ning

esitaksid protsessi lõppedes aruande;

d) lisada võimaluse, et Mandri-Euroopa riiklikud reguleerivad asutused saavad vajaduse

korral taotleda FCRi suuruse määramise ümberhindamist;

e) täpsustus, et piiravate energiasalvestite tühjenemist tuleb simuleerida alates

häireseisundi käivitumisest vastavalt SOGLi määruse artikli 156 lõikele 9;

f) pikaajaliste sagedushälvete liigitamine leevendatavateks ja mitteleevendatavateks

g) leevendusmeetmete mõju kaasamine pikaajaliste sagedushälvete andmekogumisse;

h) lisada nõude, et põhivõrguvõtjad viiksid läbi uuringu, et paremini mõista, kuidas

esialgsed ja siduvad FCRi kohustused on määratud.

Detailsemalt on puuduste ja muudatuste sisu välja toodud Mandri-Euroopa sünkroonala riiklike

reguleerivate asutuste ühise seisukoha punktis III.

Kokkuvõte

Konkurentsiamet töötas läbi 17.06.2025 Eleringi poolt ametile kooskõlastamiseks esitatud

Mandri-Euroopa sünkroonala põhivõrguettevõtjate ühise metoodika muutmist käsitleva

ettepaneku FCRi suuruse määramise eeskirjade kohta Mandri-Euroopa sünkroonala jaoks, koos

tehnilise lisaga, kooskõlas SOGLi määruse artikli 153 lõikega 2. Esitatud metoodika versioon

ja tehniline lisa juba sisaldab punktis 4 toodud Mandri-Euroopa sünkroonala riiklike

reguleerivate asutuste poolt sisse viidud muudatusi ning arvestab kehtivast seadusandlusest

tulenevate alusetega ning ei ole vastuolus SOGList tuleneva regulatsiooniga.

SOGLi artikli 8 lõike 1 alusel SOGLi määruse kohaste tingimuste või meetodite väljatöötamise

eest vastutavad põhivõrguettevõtjad avaldavad need tingimused ja meetodid internetis pärast

koostööametilt või pädevatelt reguleerivatelt asutustelt saadud heakskiitu, või kui heakskiitu ei

ole vaja, siis pärast nende tingimuste või meetodite väljatöötamist, välja arvatud juhul, kui

selline teave on artikli 12 kohaselt konfidentsiaalne.

Arvestades eeltoodut ja tuginedes SOGLi artikli 6 lõike 3 punkti d) alapunktile ii)

5 (5)

otsustan:

kooskõlastada Mandri-Euroopa sünkroonala põhivõrguettevõtjate ühise metoodika muutmist

käsitleva ettepaneku FCRi suuruse määramise eeskirjade kohta Mandri-Euroopa sünkroonala

jaoks, koos tehnilise lisaga, kooskõlas Komisjoni määruse (EL) 2017/1485 artikli 153 lõikega

2.

Käesoleva otsusega mittenõustumise korral on õigus esitada kaebus otsuse tühistamiseks

Tallinna Halduskohtule. Kaebuse halduskohtule võib esitada 30 päeva jooksul arvates

käesoleva otsuse teatavaks tegemisest.

(allkirjastatud digitaalselt)

Evelin Pärn-Lee

peadirektor

Lisad:

1. Assumptions and methodology for a probabilistic FCR dimensioning in the Continental

Europe synchronous area in accordance with Article 153(2) of the Commission

Regulation (EU) 2017/1485 of 2 August 2017 establishing a guideline on electricity

transmission system operation

2. Technical Annex of the methodology for performing the probabilistic dimensioning of

FCR in CE synchronous area according to Article 153(2) of Commission Regulation

(EU) 2017/1485 of 2 August 2017 establishing a guideline on electricity transmission

system operation

3. APPROVAL PAPER OF THE REGULATORY AUTHORITIES OF THE

CONTINENTAL EUROPE SYNCHRONOUS AREA ON THE ASSUMPTIONS

AND METHODOLOGY FOR A PROBABILISTIC FCR DIMENSIONING IN THE

CONTINENTAL EUROPE SYNCHRONOUS AREA IN ACCORDANCE WITH

ARTICLE 153(2) OF THE COMMISSION REGULATION (EU) 2017/1485 OF 2

AUGUST 2017 ESTABLISHING A GUIDELINE ON ELECTRICITY

TRANSMISSION SYSTEM OPERATION

Armin Ilisson

+372 667 2437

[email protected]