Consultation Feedback Report - Microsoft · • VATTENFALL • ENEL S.p.A. • REGULATORY ASSISTANCE PROJECT • EURELECTRIC The final MAF 2016 report, after taking into account stakeholder

ENTSO-E AISBL • Avenue de Cortenbergh 100 • 1000 Brussels • Belgium • Tel + 32 2 741 09 50 • Fax + 32 2 741 09 51 • [email protected] • www. entsoe.eu

MAF 2016 Consultation Feedback Report

March 2017



2

Purpose of this document Mid-Term Adequacy Forecast (MAF) 2016 has been under public consultation during Summer 2016. This report presents an overview of the MAF 2016 consultation process and ENTSO-E answers. The feedback from stakeholders can be classified within the following categories:

• The scope of and relationship between MAF as a Pan-European report and Regional and National adequacy reports

• The involvement of stakeholders regarding the continuous evolution and improvement of

the methodology of ENTSO-E

• Increase in transparency and clarity of the assumptions. Stakeholders welcome the efforts made by ENTSO-E in this regard but expect more

• Sensitivity of results and of the scenarios used with a focus on economic viability of the

generation mix and its impact for adequacy

• Modelling of demand-side management and flexibility needs as well as flow based methods

The following stakeholders provided feedback during the consultation:

• WINDEUROPE • EDF SA • IBERDROLA, S.A. • ENGIE • NORDENERGI • VATTENFALL • ENEL S.p.A. • REGULATORY ASSISTANCE PROJECT • EURELECTRIC

The final MAF 2016 report, after taking into account stakeholder feedback, will be issued together with this report in Q1 2017.

Based on the MAF 2016 consultation feedback and the ongoing MAF 2017 project initiation, the next release of the Mid-Term Adequacy Forecast report (MAF) 2017 will attempt to address stakeholder expectations and feedback.

In the Appendix, the detailed feedback received from each of the stakeholders mentioned above and the corresponding ENTSO-E answers are provided.



3

Background

MAF 2016 was initiated to fulfil the legal obligations of Reg. 714/2009 in the area of mid-term system adequacy.

Within the ongoing political discussions regarding adequacy, there are increasing expectations on the part of relevant stakeholders pertaining to the annual publication of the MAF reports by ENTSO-E.

The MAF report aims to provide stakeholders in the European electricity market with a Pan-European overview of generation, demand and adequacy using different scenarios for the future ENTSO-E power system, focusing on the power balance, margins, indicators and the role of interconnection and cross-border exchanges to achieve adequacy.

The publication of annual adequacy assessments by ENTSO-E serves as an important input regarding the establishment of countermeasures by relevant stakeholders (e.g., member state authorities, policy makers, regulatory agencies, energy producers) in order to ensure the desired adequacy levels. Furthermore, the fast evolution of the energy mix (i.e., growing development of renewable energy sources and increased reduction of conventional power plants) requires a regular evaluation of the adequacy situation for the upcoming ten years at least each year. This is in line with the expectations communicated by relevant stakeholders, namely the EC, during ongoing discussions regarding the Winter Package forthcoming legislative proposal (Market Design and the Risk Preparedness in the Area of Security of Electricity Supply legislation), the recommendations of the Electricity Coordination Group (ECG), the Joint Declaration for Regional Cooperation on Security of Electricity Supply in the Framework of the Internal Energy Market by the Pentalateral Energy Forum and feedback from NRAs and ACER. The MAF presents the first Pan-European probabilistic assessment of adequacy. While market-based probabilistic modelling approaches have already been adopted in a number of national generation adequacy studies and the PLEF regional adequacy assessment, this was the first time such studies have been conducted at the Pan-European level. This represents a significant analytical achievement. Moreover, this has involved extensive collaborative efforts of representatives from TSOs covering the entire Pan-European area under the coordination of ENTSO-E. There is a unique opportunity to work closer with TSOs to garner a better understanding of their timelines for producing national adequacy reports, and, where possible, seek greater harmonisation with the timelines of the annual MAF publication. This will help improve the consistency of data, analysis and key messages between the national reports and the MAF, which will aid stakeholders in realising the benefits from all studies. A market model always presents a simplified representation of the real behaviour of the power system. The results obtained in this report should always be understood under the assumptions of the data used as well as the limitations of each implementation step of ENTSO-E's methodology. Each step towards increasing the level of detail of the data and representation employed in the models significantly increases the complexity of the mathematical problem to be solved.



4

MAF 2016 Consultation Overview Q1 Scope - From your perspective, how can one ensure consistency between European, regional and national adequacy studies?

Q2 Methodology Improvements - Considering the above, what additional methodological improvements shall ENTSO-E consider for future MAFs? Please justify why through suggestions for methodological improvements and what these improvements will offer.

Q3 Decommissioning/Mothballing - In this respect, how can you or other stakeholders help ENTSO-E (and its members) to obtain more reliable data on power plant availability/decommissioning or mothballing plants?

Q4 Economic Viability - In order to present in the MAF a view on the economic viability of the generation portfolio (at the national level), the TSOs must have a complete perspective of the economic/technical data and assumptions linked to these forecasts. What would you recommend to us in order to improve the quality of the data and assumptions mentioned above?

Q5 Scenarios/Sensitivities - A solution may be to use two different sensitivity-scenarios for each time horizon as described below - what is your view on this possible solution?

i) one scenario linked to the current regulatory framework based on the energy-only market, and;

ii) a second scenario connected to the future regulatory framework (for example reflecting expectations of a national implementation of capacity mechanisms or any other market design instruments).

Q6 Other Comments - Please share with us below other suggestions you may have.

Conclusions of the MAF 2016 Consultation Stakeholder inclusion and external communication still needs to be improved

Stronger engagement via active discussions on scenarios and improved methodologies. Better external expectation management and communication on what will be delivered.

Coordinated MAF & TYNDP

Clear overall common work plan and deliveries for whole period. Use of common/consistent data and assumptions.



5

Make MAF serve both TSO & member states and EC interests MAF report to serve as the basis for discussions on establishment of a framework to

express national security of supply (SoS) targets based upon commonly agreed adequacy standards and discussions by member states’ on regionally coordinated definitions of those standards. Opportunity to improve and coordinate ENTSO-E work as well as delivery of national and regional adequacy reports.

Support of the articulation of national, regional and European adequacy assessments based on a common methodology

The application of consistent methodologies between the national, regional and Pan-European levels would allow for accounting realistically for cross-border support being available under the many different scenarios simulated. The methodology should be common and shared between these three levels taking into account national sensitivities and specificities.

Evolution of methodologies Based on the MAF 2016 consultation feedback, the MAF 2017 report is under preparation with the following improvements:

o Extension of the Pan-European Climatic Database (PECD). o Revision of cross-border interconnector assumptions to account for

seasonality and operational constraints.o Revision of thermal portfolio categories and data details and assumptions

therein.o Considerations of economic viability of scenarios and adequacy results by

sensitivity analysis regarding decommissioning/mothballing of plans, considerations of so-called “system-relevant” assets and impact of units that may be at risk of being mothballed for economic reasons, for example, in the absence of a capacity market and/or unfavourable market conditions.

o Modelling of demand-side response and stronger engagement with stakeholders in terms transparency of data and assumptions.

o Determine feasibility for and implement flow-based methods that are aligned with relevant national and regional forthcoming reports.



6

Appendix – Answer to MAF 2016 Consultation Stakeholders

COMMENTS BY STAKEHOLDERS

Ø ENTSO-E should be entitled to develop and update such methodology, ensuring adequate stakeholder consultations possibly by the creation of a dedicated stakeholders committee as it would pertain to the Ten-Year Network Development Plan (cf. Network Development Stakeholders Group).

Answers from ENTSO-E:

We thank the stakeholder for their constructive comments which are in line with the ENTSO-E Roadmap and Vision. On Stakeholder interactions: The contribution of market participants is very important to develop and apply appropriate models for adequacy assessment, especially in cases of demand-side management (DSM, hydro inflows, pump storage power plants and modelling of reserves.

An in-depth discussion with stakeholders on modelling principles and constraints resulting from technology, regulatory and market frameworks is in the current scope of ENTSO-E in the form of consultations and ad-hoc & expert workshops. Specifically, the focus is on the development of a sound methodology that will allow performance of a diagnosis of the power system with respect to the relevant adequacy risks at the Pan-European level.

Ø Developing a regionalised approach to adequacy assessment by:

• Encouraging bottom-up cooperation in regional for fora, such as the Penta Forum, in order to develop joint assessments in addition to national assessments

• Evolving towards fully regionalised adequacy assessments of which the consistency with EU methodology will be verified by ENTSO-E

Answer from ENTSO-E:

On articulation between national - regional - Pan-European reports -- methodology and report scope.

Q1 Scope - From your perspective, how can one ensure consistency between European, regional and national adequacy studies?

WIND EUROPE



7

ENTSO-E appreciates the positive feedback from stakeholders regarding the ENTSO-E methodology as a common basis. ENTSO-E views this role as follows:

a) ENTSO-E describes and proposes common definitions for the multiple concepts/aspects under Security of Supply (SoS) and adequacy, providing all stakeholders with a common language;

b) ENTSO-E creates or adopts and then standardises methodologies and processes to perform adequacy analyses and market modelling based on its members‘ expertise; and

c) ENTSO-E performs Pan-European adequacy analyses relying on dedicated Pan-European data collection processes.

ENTSO-E’s ability to propose common definitions and methodologies is unchallenged. However, the diversity of SoS concerns across Europe entails more complex modelling and processes, ultimately requiring trade-offs to comply with the many diverse issues, computing power and actual data availability from all 42 TSOs in 35 countries, and more fundamentally with conflicting model features (e.g., hydro-modelling for Nordic and Alpine countries based on different logics, unit commitments against forced outages, thermal sensitivities to demand, etc.). Hence, Pan-European, regional and national studies are an absolute necessity as they complement each other rather than compete - studies covering a larger perimeter (Pan-European compared to regional or regional/local) are limited by potentially contradicting modelling goals and challenging data collection, but provide a consistent appraisal, if not of highest accuracy, of a series of commonly and continuously assessed indicators at the Pan-European level. Conversely, more local studies need the consistent boundary conditions supplied by the larger perimeter studies to address more precisely certain specific issues.

Regional and national studies shall use the common definition and methodology standards compiled by ENTSO-E to enable comparison. For a given country or region, the prevailing conclusions and decision making regarding market design adaptations, to ensure the desired SoS level, shall stem from the study implementing the most advanced modelling for the investigated parameter. ENTSO-E adequacy studies will therefore be needed as a pre-requisite for more local regional and national studies, but they will be the reference to decide whether to implement risk mitigation measures only where no more detailed study is performed. TSO/ENTSO-E processes and bottom-up common methodology along with the use of several TSO tools across the EU perimeter in MAF makes certain there is consistency and articulation between Pan-European (MAF), regional (e.g., PLEF) and national studies in this respect.

• ACER should be endowed with the competence to oversee those regional adequacy assessments and develop guidance on which methodologies to employ for estimating value of lost load (VoLL)

• National Regulatory Authorities (NRAs) and Transmission System Operators (TSOs) should provide stakeholders with a transparent and inclusive framework that will guarantee their involvement during the relevant stages of the process, including by mean of public consultations



8


VoLL: Value of loss load is basically connected to each member state definition of a 'SoS standard' which expresses each member state’s choice regarding the level of security they want to ensure in addition to the value that SoS is given in each country by internalization of ‘SoS costs’ . 'One-size-fits-all' methodologies to define VoLL might not be practical as refined considerations concentrating on particular aspects of each member state power system are needed to properly define VoLL. ACER, NRA, TSOs: The existing framework for cooperation between TSO–NRAs already guarantees that data provided by TSO, in most cases being approved by NRAs first, is well-rooted in the above mention considerations. ACER is welcome to coordinate NRA actions on a regional level to improve consistency when needed. In-depth discussion with stakeholders on modelling principles, constraints resulting from technology, regulatory and market frameworks are planned in the form of continuous consultations and dedicated expert workshops. Focus is on the development of a sound methodology which will allow performing a diagnosis of the power system with respect to the relevant adequacy risks.

• That being said, NRAs/governments should remain responsible for setting binding

reliability standards that they compare against the outcome of adequacy assessment in order to inform the need for and size of any necessary remedial actions

• The implementation of these remedial actions should be scrutinised by the EC, ideally ex ante (if not ex officio procedures should be considered) in order to determine their consistency with European State Aid Guidelines

Answer by ENTSO-E:

A framework to express national targets for SoS on the basis of commonly agreed adequacy standards at the regional or Pan-European level is necessary. ENTSO-E methodology provides the basis for such a framework.

The adequacy levels should be nationally set but understood within a coordinated regional approach and based on the principles of subsidiarity regarding the definition of measures necessary to fulfil these standards by member states. Risks associated with these standards must be commonly understood across Europe through common definitions.

Indeed, decisions to implement measures to ensure SoS at the national level is directly connected to the responsibility for SoS. Only the entity responsible for SoS can make the decision to implement or not measures to guarantee SoS. This however should be carried out in coordination with neighbouring member states and TSOs as a result of the cross-border impact that these decision might have on neighbouring areas.



9


Ø Developing a holistic approach that systematically and realistically includes renewables, demand-side response, storage and interconnection contributions to adequacy. In particular:

• Wind capacity credit • Accurate modelling of interconnection contributions using flow-based methodology • Properly taking into account and promoting more transparency on the economic

viability of power plants and the level of operating reserves (including must-run arrangements)

• Evolve towards more granular time resolution while extending the size of the climatic data set


We thank the stakeholder for their constructive comments that are in line with the scope of the methodological evolution foreseen for forthcoming MAFs.

The purpose of the ENTSO-E adequacy methodology and MAF reports under evolution is the development of a robust methodology which will allow conducting regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Although these studies will not encompass every potential issue of future power system, ENTSO-E strongly believes that they:

i) Will permit the development of a model able to capture all key features and risks regarding adequacy for the Pan-European power system. This is a main objective for ENTSO-E;

ii) Will act as a common basis for methodologies and definitions; and iii) Will greatly facilitate consistent articulation between Pan-European, regional and national adequacy studies.

ENTSO-E is also aware that the choice of the mathematical approach could significantly affect the indicators that may be assessed through the simulation as well as the structure and complexity of the input data and modelling assumptions used.

Q2 Methodology Improvements – Considering the above, what additional methodological improvements should ENTSO-E consider for future MAFs? Please justify why through suggestions for methodological improvements.



10

ENTSO-E is additionally cognizant of the impact that designing an optimal and economically feasible set of investments plans in generation portfolio has on adequacy levels. We refer to the answers below. Regarding Planned Methodological Improvements

Demand: ENTSO-E agrees with stakeholders that demand forecasting (levels, but also profiles) as one of the main items affecting the future adequacy outlook. Therefore, ENTSO-E is working on improving modelling and the reliability of demand forecasting through enhancing methodologies. In order to ensure consistency and transparency, ENTSO-E has set up a process to define, in a centralized manner, demand forecast levels and profiles. TSOs are requested to provide national figures for the main parameters to model demand: temperature sensitivity of load, electric vehicle (EV) penetration, heat pumps (HPs), energy efficiency (EE), DSM/demand-side response (DSR), expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. In this way, consistent regional/Pan-European demand figures and profiles are defined that are also in line with national expectations and capture national specificities properly.

Hydro: Hydro modelling details can have a significant influence on adequacy indicators, especially of regions, including countries, with large shares of hydro production. Dry and wet hydro conditions (inflow and reservoir) have been considered in MAF 2016. In addition, the ENTSO-E PECD is currently being extended to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the probability of occurrence of these dry-normal and wet conditions is being investigated systematically in order to define so-called hydrological regions for which the same probability of similar conditions can be accounted for in Monte Carlo probabilistic simulations.

Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices. New Methodological Improvements:

Economic Viability: Sensitivity analysis that will take a more ‘conservative’ view of installed capacity might be needed in future MAFs. To enable this, ENTSO-E seeks to include in the data collection process for future MAF reports estimates of the volume of capacity that could be at risk of being mothballed for economic reasons, like, for example, in the absence of a capacity market and/or unfavourable market conditions – ‘missing money’ problem. These data will be analysed and sensitivities will be defined for the simulations.

Visibility of Data: ENTSO-E strives toward a high level of transparency. The so-called 'MAF2016_market_modelling_data' package, including all input net generating capacity (NGC), demand and net transfer capacity (NTC) figures, has been made available upon publication of the report. ENTSO-E strives to enhance transparency regarding other key parameters used for the simulations. Transparency also implies responsibility not only of the publisher, but also of the recipient. Improvement of these key parameters would be possible by discussion with relevant



11

market parties/actors that can provide detailed technical data and feedback regarding the different generation technologies available. Such exchange could also make better the available data by TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as it concurs that the results should always be understood within the context of the data and assumptions used. Transparency of Algorithms:

Different market modelling tools available, owned or procured by TSOs are utilised during the MAF exercise. These tools are proven to capture nationally and regionally relevant specificities. MAF work relies on this know-how from TSOs in terms of the strong bottom-up basis of all knowledge and detailed data. Tool dependency and reliability is a major focus for MAF. Use of exclusively open-source tools might be beneficial from a purely ‘academic’ point of view, though on the other hand, does not guarantee a strong link to bottom-up national details – know-how of TSOs and national studies. Full transparency is not always possible because these tools are, in certain cases, proprietary materials from different tool providers in a similar way that generation and market data is confidential to market parties as communicated via the stakeholder feedback and answers throughout this document.

Role and Impact that the Results of the MAF Reports: The MAF 2016 report features an ‘Overview Table’ on page 3 complemented by an appendix (Appendix 2) presenting the comments from each national TSO on the results in MAF 2016 in relation to each country’s own assessment of adequacy levels, its national adequacy standards and the measures taken to maintain them in the case of problems. The choice of LOLE = 1 hour as a threshold in such an overview table was chosen only because of technical reasons related to the presentation of the results. Only values of LOLE > 1 hour after averaging the results of the whole probabilistic Monte Carlo exercise were deemed significant enough to be highlighted, i.e.. the occurrence of adequacy problems across all the Monte Carlo ensemble of situations analysed was deemed significant rather than marginal.

This choice of 1 hour should NOT be interpreted as a statement by ENTSO-E or its member TSOs on any proposals of harmonized SoS standards for EU of LoLE = 1 hour (!).

We know that such a choice of threshold for the previously referred to Overview Table could be nonetheless confusing and leads to misinterpretation. ENTSO-E will carefully consider the role and impact of the presentation of the results in future MAF reports. ENTSO-E is aware of the impact that the MAF results might have on discussions by member states on regionally coordinated definitions of those standards. Total Need for Reliable/Dispatchable Capacity (MW):

ENTSO-E will carefully consider the possibility of reporting the aforementioned Figure in the presentation of the results in future MAF reports. However, providing a view on specific generation technologies is not necessarily the role of TSOs as regulated entities. Capacity Surplus/Deficit (MW):

ENTSO-E will cautiously account for the possibility of reporting the aforementioned Figure in the presentation of the results in future MAF reports. Providing views on specific generation technologies is not necessarily the role of TSOs as regulated entities. This indicator seems related to the deterministic indicator, remaining capacity margin (RCM), employed by ENTSO-E in SO&AF 2015 and previous reports. This indicator was not chosen in MAF 2016 in order to focus



12

attention to the new probabilistic method deployed and LOLE and ENS indicators used. Reporting the RCM again in forthcoming MAF reports will be considered. Transparency of Key Parameters, Assumptions and Data:

ENTSO-E must fulfil obligations of Reg. 714/2009 and contribute to the overall European/national debates on adequacy concerns. ENTSO-E is committed to transparency of data and products delivered while respecting national legislation and confidentiality agreements between TSOs and national stakeholders. Furthermore, confidentiality issues might also require data to be publically released in an aggregated manner. ENTSO-E endeavours towards a high level of transparency on key parameters and assumptions. Improvement of key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the various generation technologies available. Such exchange could enhance the available data from TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as ENTSO-E agrees that the results should always be understood within the context of the data and assumptions used.

Thermal Outage Correlation with Climatic Conditions: Extensions of the methodology might be considered to capture correlations of thermal outages, thermal production and climate conditions, notably temperature conditions. It is worth noting that this effect is very pronounced in certain countries, e.g., PL, but might not require a Pan-European deployment, and instead require local sensitivities with respect to the MAF results. Sensitivity of Scenarios:

ENTSO-E must fulfil obligations of Reg. 714/2009. The European generation adequacy outlook shall build upon national generation adequacy outlooks prepared by each individual transmission system operator. The first goal of ENTSO-E is to collect and set up a contiguous (central) best estimate ‘Pan-European’ scenario based on data collected from TSOs as meaningful sensitivities are defined around this. ENTSO-E is therefore revising its data collection and scenario-building process in order to consider relevant sensitivities based on data provided by TSOs on, e.g., units at risk of being mothballed for economic reasons, information regarding decommissioning of units for purely technical reasons and/or shut downs for legal/policy reasons within the storyline assumptions of the scenarios considered. Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios.

Considering the above, what additional methodological improvements should ENTSO-E consider for future MAFs? Please justify why through suggestions and what these improvements will offer.




13

Ø Move towards a systemic approach, including RES, DSR, interconnections, etc.:

As of today, most national adequacy assessments concentrate on the contributions of firm generation units, with little or no consideration for the contributions of other energy resources, such as DSR, storage, imports/exports (see next section) or renewables.

In WindEurope’s view, however, their contribution should be systematically and realistically reflected in future adequacy assessments in order to properly reflect how the EU power system operates.

In particular, regarding wind energy’s contribution, the aggregated capacity credit (annual availability factor, measure on % of a year) of the wind farms in a system depends on many factors. Among them, the characteristics of the power system in question (reliability level, geographical area, flexibility and composition of the total generation mix, correlation between low electricity price and demand) and the penetration level of wind power in the system should be highlighted. It is also reliant on a range of wind and wind technology-specific factors, like the capacity factor or location of wind farms in the system.

Despite the real physical capacity value of wind power, it is not yet regularly used for capacity planning and is frequently not given a value in power markets using deterministic adequacy assessments. This is partially because of the diversity of methods available for calculating the capacity credit, but also stems from a lack of assessing adequacy at the European level beyond control zones. Firm capacity from wind power has neither been thoroughly analysed in an integrated EU system nor has its interplay with other renewables, including photovoltaic, been considered. Such analysis would aid in mitigating the variability from both, increasing their firm capacity share. The TradeWind4 study found that the effect of aggregating wind energy across multiple countries increased the average capacity credit by a factor 1.7 compared with the capacity credit averaged over separate countries.

Clearly, the completion of the Internal Energy Market is instrumental for this benefit to be exploited. The wider the control zones are geographically, the higher the resulting capacity credit of wind.

Additionally, the operational performance of wind turbines has improved over time thanks to continuous technological innovation. Those future improvements should somehow be reflected.


Firm Capacity from Renewables: The PECD will be extended to 35 years for the whole Pan-European perimeter. This is a significant enhancement of Pan-European accuracy and the scope of MAF. Such an improved data set allows even more robust probabilistic assessments. ENSTO-E will carefully consider whether to incorporate firm capacity, either from renewables or other generation technologies. Notice that it is not the role of TSOs to provide statements with respect to specific technologies and the MAF reports must stay technology neutral in this respect.

Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction



14

should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices. Flow-Based Methods

Market studies provide input to detailed grid studies. Some countries are more sensitive than others to commercial exchanges and their differences with so-called “physical exchanges”.

ENTSO-E is aware that the current approach to carrying out market studies may yield certain unrealistic results corresponding to copper plates connected with HVDCs. Approaches, such as PTDF methods, facilitate the establishment of the link between market studies and grid studies by translating market results into approximated physical flows for each hour of a Monte Carlo year. Once this translation is complete, sanity checks can be performed to ensure that the bilateral transfer capacities (BTCs) between countries or price zones are not over- or underestimated. However, closing the loop is then a difficult matter. The integration of physical grid concepts (capacities, impedances) directly into market models, usually referred to as “flow-based methods”, avoid the unrealistic results pitfall. ENTSO-E is currently determining how to define and put in place flow-based methods within long-term Pan-European market studies, the main purpose being to make commercial exchanges more realistic and faithful to actual physical flows versus the current approach.

It should be noted that flow-based methods are currently deployed by a number of TSOs in national studies (Belgium and France) and are in the scope of regional studies, e.g., PLEF. Pan-European studies, like MAF, will profit from these evolutions and will integrate them at the EU level once mature at the national and regional levels.

Moreover, in a Pan-European electricity market, problems (e.g., bottlenecks, blackouts, etc.) do not stop at national borders - experience has largely proved that. In addition, interconnectors play an essential role in ensuring security of supply as they can support efficient utilisation of electricity resources across Europe, in particular for renewables producers with variable output. There is a strong case to take the benefits of grid exchanges, as well as the benefits of shared operational reserves when relevant, into account. Analysis of France on a standalone basis (source: RTE), i.e., not considering cross-border exchanges, shows how vital imports are to ensuring domestic security of supply. In that case, exchanges contribute about 8 to 10 GW on average during peak periods over the coming years.

Last but not least, most national adequacy assessments fail to capture the contribution of DSR. In 2014, only three countries reported that they include demand response as a separate factor in their load forecast. Alternatively, DSR may be indirectly included in the projections through the effects it has had on the historical load curves. Yet, there is neither a common approach across the EU to factor in its impact on adequacy (see box below) nor is there a common understanding regarding what DSR actually covers. ENTSO-E SOAF only takes into account interruptible demand schemes where they exist, and starts factoring in DSR capabilities that are emerging in specific markets (Belgium, France). Similarly, a minority of member states take storage contributions into account



15

in their national adequacy outlooks. When they do, it is mainly pumped hydro storage. The next editions of the MAF should concentrate on bolstering this, as well.

Answers from ENTSO-E: Seeabove.

Ø Transparency on economic viability of power plants and the level of operating reserves:

Regarding forecasts for installed capacity in particular, there is often very limited information available about the commercial plans of individual operators. Information on/analysis of the economic viability of existing and new plants should somehow be improved and considered.

In the future, daily gaps between maximum and minimum residual demand is likely to increase, thereby increasing the related flexibility requirements. This raises the question of whether or not the amount of operational reserves (for system security) should be taken into account as “available capacity” for adequacy assessments.

According to CEER, in at least four countries (France, Sweden, Italy and Hungary), the volumes procured by TSOs in terms of ancillary services and balancing reserves is treated as “available” capacity. On the contrary, many member states, as along with ENTSO-E, subtract it from their calculations. In that case, national TSOs should make certain the level of operating reserves is set transparently and considering the potential contribution of pooling and sharing reserves across borders as foreseen by the Target Model. Besides, the inclusion or not of operational reserves should be considered together with the chosen time resolution. If they are deemed “available capacity”, time resolution must be reduced to 15 minutes as shortage can still occur within an hour (not solved by day-ahead trading). In addition, more transparency is necessary with regards to the level of must-run obligations. As there is little or no access to those data, it grants TSOs the opportunity to set conservative figures that might influence the result of the simulation.


Frequency Containment Reserves (FCR): Operating reserves necessary for constant containment of frequency deviations (fluctuations) from nominal value in order to constantly maintain the power balance in the whole synchronously interconnected system. Activation of these reserves results in a restored power balance at a frequency deviating from the nominal value. This category typically includes operating reserves with an activation time of up to 30 seconds. Operating reserves of this category are usually activated automatically and locally.

Frequency Restauration Reserves (FRR): Operating reserves are there to restore the imbalance of a control area after an outage or forecast error for that area. They should be available within 7.5 minutes (automatic) or 15 minutes (manual). These can be contracted or procured by TSOs through the market but the TSO must guarantee their availability in accordance with the reserve requirements of each control area. FRR should be always available.



16

With MAF, the main imbalances addressed in the simulations relate to planned outages (maintenance schedules) and forced (unplanned) outages. Forced outages are modelled in MAF as occurring randomly at a given hour, though can have different durations in reality. If a forced outage happens and lasts for > 24 hours, it is visible in the day-ahead (D-1) market of the next x day(s). This unit is then not available neither for the market nor for the balancing reserves (FCR, FRR). So, in the end, capacity that has to be provided for the D-1 is the sum of what is required to cover the load + FCR + FRR requirements. This is what is modelled in MAF when modelling FCR + FRR as an extra load.

If a forced outage is of a short duration, lasting a few hours (a “failed to start”, for example, which is very common), then the balancing responsible party (BRP) might not have time to recover this capacity in the market (intra-day market gate closing being longer than the outage). In this case, TSO will need to restore such an imbalance with the ‘remaining’ available balancing reserves. Under these circumstances, the modelling FCR + FRR as an extra load might represent a double-counting during the assessment as reserves might be used to cover the FOR fault.

Even if those short outages are able to have a hourly instantaneous high impact to the adequacy level, their weight in the total forced outage rate calculated by TSOs and used in MAF is low (longer outages have much more weight), so the forced outage rates deployed in MAF account mostly for long-lasting outages impacting the adequacy situation significantly in a way that it is visible to the day-ahead market for subsequent days. Other arguments favour the modelling of FCR + FRR as an extra load - BRPs are/will be more and more incentivized to restore their balances in real-time, so if an outage takes place within their portfolio, they need to recover it as soon as possible without the intervention of TSOs employing FCR and FRR. These reserves are not designed to cover shortfalls in capacity or energy in the market but to ensure frequency stability.

Furthermore, not only imbalances because of forced outages are covered by balancing reserves. Forecasting errors for load, wind and solar production are also addressed. With the increase of RES capacity, it is occasionally difficult to forecast hour by hour (even if improvements are made year after year), so this is more and more important.

ENTSO-E aims to achieve a high level of transparency for key parameters and assumptions. ENTSO-E will consider making available the values of FCR + FRR assigned as an extra load in the simulations.

Ø Time resolution: More granularity in time resolution, such as a 30 or even 15 minute time window may better capture the variable RES coverage of load. Further, the length of the historical climate data set is key to ensuring it is sufficiently statistically relevant.

Answers from ENTSO-E: See previous answers.

The purpose of the ENTSO-E target adequacy methodology being developed is the generation of a sound methodology that will foster performing regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Considering any modelling implies simplifications, so ENTSO-E adequacy target modelling focuses on hourly power balance



17

modelling. Although these studies do not encompass every potential issue involved in the state of the future power system, ENTSO-E strongly believes that this methodology will permit the development of a model able to capture all the key features and risks regarding adequacy for the Pan-European power system - this is a main objective for ENTSO-E.

The effect of imbalances based on forced outages occurring close to real-time and the effect of forecasting errors of load, wind and solar production is accounted for in MAF by consideration of operational reserve requirements as an extra constraint while modelling. The detailed modelling of these aspects is outside of the scope of MAF as these imbalances are typically covered by the intra-day and primarily balancing markets and are not observed systematically in the day-ahead market during several days of a week.

Flexibility: Residual load analysis, as considered in SO&AF2015, will be appraised for future MAF reports in order to inform them with respect to the need for flexibility necessary in power systems, typically within intra-day and balancing markets. One of the main goals is to be able to integrate 'need for flexibility' into the system. Flexibility in adequacy assessments is based on weather-dependent effects related to load variation, generation patterns of wind and solar power plants with a one-hour resolution and the consideration of the resources for flexibility. The existing PECD will be used for the adequacy assessments.


Such reporting should be organised at the national level by the NRA, and then scaled up to the regional or EU level.

WindEurope agrees that further reporting and monitoring rules at the EU level should be developed. This could possibly be accomplished with the upcoming legislation on Energy Union's governance. It would also make certain that future adequacy assessments take into account the objectives in terms of RES deployment and EE programmes enshrined in the upcoming 2030 National Climate and Energy Plans, of which progress is expected to be tracked by the Commission every 2 years. Answers from ENTSO-E:

Visibility of Decommissioning – It is currently being investigated whether to carry out MAFsensitivity analysis that takes amore ‘conservative’ viewof the installed capacity. To facilitatethis,datacollectionhasbeenbroadenedsoTSOcanprovidetheirperspectiveonwhichunitsmaybe at risk of being mothballed for economic reasons, like, for example, in the absence of acapacitymarketand/orunfavourablemarketconditions–‘missingmoney’problem.

Q3 Decommissioning/Mothballing - In this respect, how can you or which other stakeholders can help ENTSO-E (and its members) to obtain more reliable data on power plant availability/decommissioning or mothballing plants?



18

COMMENTS BY STAKEHOLDERS No Comment.


No comment.


So far, national TSOs remain responsible for assessing structural adequacy concerns under the scrutiny of national regulators/governments. Those assessments should be performed on a yearly basis looking at time horizons of 5-10 years, and modelling the impact of an integrated system at the regional level.

Q4 Economic Viability - In order to present in the MAF a view on the economic viability of the generation portfolio (at the national level), the TSOs must have a complete perspective of the economic/technical data and assumptions linked to these forecasts.

What would you recommend to us in order to improve the quality of the data and assumptions mentioned above?

Q5 Scenarios/Sensitivities - A solution may be to use two different sensitivity scenarios for each time horizon as described below - what is your view on this possible solution?

i) One scenario linked to the current regulatory framework based on the energy-only market; and

ii) A second scenario connected to the future regulatory framework (for example reflecting expectations of a national implementation of capacity mechanisms or any other market design instruments).

Q6 Other Comments – “Please share with us below other suggestions you may have.”



19

In the future, fully regionalised adequacy assessments, performed jointly by TSOs and based on a harmonised methodology to be developed and updated by ENTSO-E, should be mainstreamed. ACER should be entitled to oversee those assessments and determine their consistency with EU regulatory frameworks (state aid guidelines, grid code requirements, etc.). This would safeguard there was a rigorous, need-based and coordinated approach to SoS that takes into account the benefits of the internal energy market. For the sake of transparency, stakeholder consultations should be organised to provide opportunities to provide input on a) the harmonised methodology to be developed by ENTSO-E and b) the assessments to be conducted at the regional level.

Answer by ENTSO-E: The ENTSO-E adequacy methodology considers that the large-scale exploitation of renewable energy sources of variable generation poses challenges for electricity system operation. In addition to sufficient levels of back-up capacity, additional resources for system flexibility will be needed in the future. The current focus is on the development of a robust methodology that will allow for performing a diagnosis of the power system with respect to the main adequacy risks at the Pan-European level. One of the key objectives is to be able to inform the system with the 'need for flexibility'. Flexibility in adequacy assessments is based on weather-dependent effects related to load variation, generation patterns of wind and solar power plants with a one-hour resolution and the consideration of the resources for flexibility. The PECD will be utilised for these adequacy assessments. Although these studies will not encompass every potential issue involved in the future power system, ENTSO-E strongly believes that this methodology will permit the development of a model able to capture all the key features and risks regarding adequacy for the Pan-European power system. This is a major aim of ENTSO-E; together with other studies/activities performed by ENTSO-E, these analyses will ultimately assist improving market design and network codes. In that sense, it is worth mentioning that ENTSO-E is aware of the impact of adequacy on designing an optimal and economically feasible set of investment plans within a generation portfolio.

ENTSO-E is also cognizant of the choice of the mathematical approach affecting, in a significant manner, the indicators that can be assessed through the simulation as well as the structure and complexity of the input data. Hence, the assessment approach will be further revised in each step of the methodology improvement process. The methodological details will be properly evaluated at each step in the form of continuous consultations and expert workshops.

Although System Security Centres, as defined by the newly adopted System Operation Guidelines (e.g., Coreso, TSC) may have a part in performing seasonal outlooks, their involvement is not relevant beyond this operational time horizon as long-term adequacy concerns are more linked to investment cycles.

From a different viewpoint, it must be ensured that future adequacy assessments take into account the objectives in terms of RES deployment and EE programmes enshrined in the upcoming 2030 National Climate and Energy Plans, of which progress is expected to be tracked by the Commission every 2 years."




20

Scenarios for MAF 2020 and 2025 and TYNDP 2030 and 2040 – differences: Exploratory Scenarios: ENTSO-E has developed long-term scenarios, e.g., 2030 scenarios used within TYNDP are exploratory with respect to generation, demand and Pan-European adequacy. Each m follows different storylines indicating possible futures with certain Pan-European (top – down) political – economic – technological assumptions. Over this long-term horizon, e.g., RES deployment and EE are considered explicitly, e.g., taking into account National Renewable Energy Action Plans (NREAPs) and EC targets with intrinsic uncertainties surrounding the evolution of the energy mix for a long horizon.

As a result of this uncertainty and because of the shorter investment cycle for new generation, the suitable time frame to assess generation adequacy at national and regional resolution, anticipating possible adequacy issues, is 5-10 years (maximum) as recommended by the ECG – subgroup in adequacy. Within this time frame, trustable diagnoses of generation adequacy risks are possible by use of a sound, widely accepted and transparent methodology. Bottom-up (best estimate) scenarios: ‘… built national generation adequacy outlooks prepared by each individual TSO’ are therefore used within the MAF so the risks of adequacy are identified in due time (beyond 5 years ahead and up to 10 years). Adequacy assessments are therefore performed for ‘predictive’ mid-term scenarios rather than those that are ‘exploratory’ over the long-term.

TSOs provide data to their best knowledge on the evolution of their generation mix. Regarding RES forecasts within these bottom-up (best estimate) scenarios, TSOs take into account the current knowledge regarding supporting mechanisms for renewable energy sources in each country and the expected development of these support mechanisms if changes are under discussion. Therefore, the 2020 and 2025 scenarios provided are as realistic as possible as the forecasts for the years 2020 and 2025 from TSOs, meaning that some slight deviation – delays in development of National Renewable Energy Action Plans (NREAPs) – might be possible if deemed realistic.



21


Ø EDF believes that European, regional and national adequacy studies are complementary tools for assessing generation adequacy in Europe. In our opinion, the elaboration of European and regional adequacy assessments represents an opportunity to strengthen TSO coordination concerning the data used as input in national studies (e.g., when modelling neighbouring electricity systems) and to favour the sharing of best practices regarding the methodologies employed for such analyses (e.g., probabilistic approach, etc.).

Answers from ENTSO-E: On the articulation between national - regional - Pan-European reports -- methodology and report scope:


a) ENTSO-E describes and proposes common definitions for the multiple concepts/aspects under SoS and adequacy, providing all stakeholders with a common language;

b) ENTSO-E creates or adopts and then standardises methodologies and processes to perform adequacy analyses and market modelling based on its members’ expertise; and

c) ENTSO-E performs Pan-European adequacy analyses relying on dedicated Pan-European data collection processes. ENTSO-E’s ability to propose common definitions and methodologies is unchallenged. However, the diversity of SoS concerns across Europe entails more complex modelling and processes, ultimately requiring trade-offs, to comply with the many diverse issues, computing power and actual data availability from all 42 TSOs/35 countries, and more fundamentally with conflicting model features (e.g., hydro-modelling for Nordic and Alpine countries based on different logics, unit commitment against forced outages, thermal sensitivities to demand, etc. ).

Hence, Pan-European, regional and national studies are a must as they complement each other rather than competing - studies covering a larger perimeter (Pan-European compared to regional or regional/local) are limited by potentially contradicting modelling goals and challenging data collection, though provide a consistent appraisal, if not of the highest accuracy, of a series of commonly and consistently assessed indicators at the Pan-European level. Conversely, more local


EDF SA



22

studies need the consistent boundary conditions supplied by the larger perimeter studies to address certain specific issues more accurately.

Regional and national studies shall use the common definition and methodology standards compiled by ENTSO-E to enable comparison. For a given country or region, the prevailing conclusions and decision making regarding market design adaptations to achieve the desired SoS level shall stem from the study implementing the most advanced modelling for the investigated parameter. ENTSO-E adequacy studies will thus be needed as a pre-requisite for more local regional and national studies, but they will be the reference from which to decide upon implementing risk mitigation measures only where no more detailed study is performed. TSO/ENTSO-E processes and bottom-up common methodologies, along with the use of several TSO tools across the EU perimeter in MAF, guarantees that consistency and articulation between Pan-European (MAF), regional (e.g., PLEF) and national studies exists in this respect.

Ø EDF wishes to highlight that regional and European adequacy studies provide an overly

simplified analysis of generation adequacy at the national level, thereby making the elaboration of national assessments necessary to complete the analysis of the impact of all the relevant contingencies for SoS. For instance, ENTSO-E bases its analysis on a limited number of climatic years (14 to be extended to 35 in subsequent MAF editions) and it does not take into account specific correlations between climatic conditions and RES production (wind and PV) as is the case with certain TSOs (e.g., RTE) in their adequacy assessments. National TSOs also have greater visibility and knowledge of the management of hydro stock and availability of single power plants in their control area. This makes the methodology used in the MAF less exact when seeking to identify adequacy issues compared to national assessments. Another example of the limits of the analyses in MAF is explicitly mentioned in the “Country Comments” section on France (paragraph 6.2.11, page 76). ENTSO-E underscores that the methodology used to build load in the French Generation Adequacy Report is different than that utilised in MAF. ENTSO-E also states that this could prevent MAF from determining possible load-shedding situations caused by extreme weather events.


Methodological Improvements


The purpose of the ENTSO-E adequacy methodology and MAF reports under evolution is the development of a robust methodology which will allow conducting regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks.

Althoughthesestudieswillnotencompasseverypotentialissueoffuturepowersystem,ENTSO-Estronglybelievesthatthey:i) Will permit the development of a model able to capture all key features and risks regarding adequacy for the Pan-European power system. This is a main objective for ENTSO-E;

ii) Will act as a common basis for methodologies and definitions; and

iii) Will greatly facilitate consistent articulation between Pan-European, regional and national adequacy studies.



23


ENTSO-E is additionally cognizant of the impact that designing an optimal and economically feasible set of investments plans in generation portfolio has on adequacy levels. We refer to the answers below.

Ø Moreover, in the current version of the MAF, each country is considered one node and

network constraints are considered only at the country border level (except for countries composed by more than one bidding zone). This simplification could limit the determining of possible SoS problems because of internal network congestions. We deem this remark to be relevant as we currently observe variations of particular cross-border capacity limitations obviously originating from such congestions.





Ø Therefore, EDF welcomes the efforts made by TSOs to cooperate on the elaboration of the

MAF and supports all the methodological iterations seeking to improve the ability of this study to accurately identify load-shedding risks. The application of best practices for MAF currently adopted in national adequacy studies can be appropriate for enhancing the consistency of the results. A greater continuity between the scenarios utilised in the MAF and the assumptions made at the national level are also valuable.



24

Yet, the MAF cannot be used in any case as the unique reference to assess generation adequacy in Europe because a certain level of simplification is necessary to safeguard the feasibility of a Pan-European analysis. For this reason, national assessments will remain necessary to carry out the tailor-made analyses required to better reflect national specificities, consistently with TSOs’ responsibility in ensuring electricity SoS at the national level.


The added value of ENTSO-E MAF reports is through Pan-European assessments not being questioned by member states as they provide a consistent appraisal of a series of commonly and continuously evaluated indicators at the Pan-European level within the necessary relevant accuracy limits. It is important to note that today, such indicator review at the Pan-European level does not exist. ENTSO-E reports fill this gap. Still, local/regional studies are always needed to complete the picture painted by the Pan-European reports.

Making ENTSO-E Pan-European reports the ‘binding single’ generation adequacy assessment for member states to exclusively rely upon when arguing for CMs is, however, not a recommended solution - the diversity of SoS concerns across Europe encompasses complicated modelling and processes, necessitating overall trade-offs and, therefore, complementarity between Pan-European-, regionally and nationally focused analyses of generation and resource adequacy. Articulation of Pan-European, regional and national studies is hence a must as these complement, rather than compete, with each other. National studies must be consistent with the boundary conditions and assumptions provided by the larger perimeter Pan-European and regional studies and have to employ the common definitions and methodology standards defined and used by ENTSO-E.


Ø EDF welcomes the significant methodological improvements introduced in this new ENTSO-E adequacy report (e.g., probabilistic approach) and supports ENTSO-E’s efforts in continuously developing and improving both the modelling tools and the data assumptions for the MAF. EDF is also generally in favour of the future methodological refinements listed on page 7 of the Report.

Nevertheless, EDF wishes to suggest further enhancements:

• More consistency between the assumptions used for the evolution of the energy mix and those related to the development of interconnection capacity;

• More accurate calculation of cross-border exchange capacity;




25

• Introduction/description of the assumptions on the management of simultaneous scarcity situations; and

• DSR-explicit modelling focused on unexploited DSR potential.


We refer to the answers below. Regarding Planned Methodological Improvements

Demand: ENTSO-E agrees with stakeholders that demand forecasting (levels, but also profiles) as one of the main items affecting the future adequacy outlook. Therefore, ENTSO-E is working on improving modelling and the reliability of demand forecasting through enhancing methodologies. In order to ensure consistency and transparency, ENTSO-E has set up a process to define, in a centralized manner, demand forecast levels and profiles. TSOs are requested to provide national figures for the main parameters to model demand: temperature sensitivity of load, electric vehicle (EV) penetration, heat pumps (HPs), energy efficiency (EE), DSM/demand-side response (DSR), expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. In this way, consistent regional/Pan-European demand figures and profiles are defined that are also in line with national expectations and capture national specificities properly. Hydro: Hydro modelling details can have a significant influence on adequacy indicators, especially of regions, including countries, with large shares of hydro production. Dry and wet hydro conditions (inflow and reservoir) have been considered in MAF 2016. In addition, the ENTSO-E PECD is currently being extended to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the probability of occurrence of these dry-normal and wet conditions is being investigated systematically in order to define so-called hydrological regions for which the same probability of similar conditions can be accounted for in Monte Carlo probabilistic simulations. Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices. New Methodological Improvements:

Economic Viability: Sensitivity analysis that will take a more ‘conservative’ view of installed capacity might be needed in future MAFs. To enable this, ENTSO-E seeks to include in the data collection process for future MAF reports estimates of the volume of capacity that could be at risk of being mothballed for economic reasons, like, for example, in the absence of a capacity market and/or unfavourable market conditions – ‘missing money’ problem. These data will be analysed and sensitivities will be defined for the simulations. Visibility of Data: ENTSO-E strives toward a high level of transparency. The so-called 'MAF2016_market_modelling_data' package, including all input NGC, demand and NTC figures,



26

has been made available upon publication of the report. ENTSO-E strives to enhance transparency regarding other key parameters used for the simulations. Transparency also implies responsibility not only of the publisher, but also of the recipient. Improvement of these key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the different generation technologies available. Such exchange could also make better the available data by TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as it concurs that the results should always be understood within the context of the data and assumptions used. Transparency of Algorithms:


Considering the above, what additional methodological improvements should ENTSO-E consider for future MAFs? Please justify why through suggestions for methodological improvements.


1. Consistency between energy mix and interconnection scenarios

TSOs should pay particular attention to making certain there is consistency in the energy mix assumptions used in the expected progress/best estimates scenarios and the values used for the “Adequacy Reference Transfer Capacity”. If the anticipated evolution of the power plant park is not in line with the assumptions on the available transmission capacity, this could result in misleading results based on the adequacy forecast.

For instance, we noted that the net generating capacity values used for the adequacy forecast of GB corresponded to those employed in the Gone Green Scenario developed by National Grid while the transmission capacity values used (around 5 GW for 2020 and 10 GW for 2025) were much lower than that chosen for that scenario at the national level (respectively 7,5 GW in 2020 and 19 GW in 2025). Whereas we understand the conservative approach of ENTSO-E in providing transfer capacity values for adequacy analyses, we believe that the adequacy risk can be overestimated if the energy mix taken into account (e.g., high penetration of RES) is clearly incompatible with the transfer capacity considered, as would seem to be the case for GB.




27

NTC – Conservative. Conservative assumptions have been made regarding cross-border transmission capacity relating to uncertainty in the commissioning dates of cross-border transmission projects and adequacy risks therein. These assumptions are within the scope of the methodology and report capturing all key features and risks relevant at the Pan-European level - this is a main objective for ENTSO-E. Aconservative approach of ENTSO-E in providing transfer capacity values for adequacy analyses is sensible, even if adequacy risk might be overestimated. This will mean that the energy mix considered will not be compatible with the assumed/available transmission infrastructure. In this respect, the MAF results provide a signal in the form of ‘adequacy risk’. This could very well be higher than the desired adequacy level that a country is willing to accept or prepare for, expressed in the form of its adequacy standard. As a consequence of the fact that investments in grid infrastructure require a longer planning and decision horizon (~15 years) than the typical investment cycles for investments in new generation (3-5-7) years, the MAF exercise will detect risks and provide important feedback regarding the establishment of countermeasures by relevant stakeholders (e.g., member state authorities, policy makers, regulatory agencies, energy producers) in order to ensure the desired adequacy levels and the risks linked to their choices of generation mix and its evolution as well the impact of cross-border capacity availability. Furthermore, national or regional assessments might consider different assumptions that could be more optimistic regarding availability of cross-border transmission infrastructure and availability of imports to complete the assessment of the impact of such assumptions on adequacy levels.

2. Calculation of cross-border exchange capacity

EDF understands that for computational reasons, and because most adequacy criteria are defined at the national level, ENTSO-E has chosen to model every bidding zone as a single node. It should, however, be noted that a change in generation/demand dispatch within a country may lead to a significant variation of NTCs, even with an unchanged transmission infrastructure (e.g., NTCs on German borders are different in the case of high wind). This may justify considering even more conservative NTC values in the MAF as generation and demand will still evolve in 2020 and 2025, at least until finer models are available permitting establishing the interdependencies between relevant parameters.

Answers from ENTSO-E: For MAF 2017, ENTSO-E is considering to account for seasonality of NTC (off peak, peak, seasonality, etc.)

3. Introduction of assumptions on the management of simultaneous scarcity situations

EDF asks for more transparency in terms of the assumptions employed for the management of simultaneous scarcity situations in the framework of adequacy assessment. In our view, this is an important factor deserving attention in regional adequacy assessments and a description of the approach used would be extremely helpful for evaluating the findings of the analysis.




28

As planned for MAF 2017 and future reports, ENTSO-E is considering how to detect situations with simultaneously high ENS for neighbouring countries (i.e., simultaneous scarcity situations) by post-processing of the yielded results.

4. Modelling of the currently unexploited DSR potential

Even if DSR is not explicitly modelled, EDF posits that, when calibrating thermal sensitivity of load, ENTSO-E already considers the demand response potential currently exploited (implicitly or explicitly) as it influences a reduction of demand in peak situations (i.e., in cases of very low or very high temperatures that depends on the country). This is probably the main reason for the saturation effect revealed by the lower gradient dP/dT identified when temperatures are particularly low or high (see Figures on page 32).

Therefore, ENTSO-E should clarify that explicit modelling of DSR in the future editions of the MAF will only improve the assessment of the impact of the DSR potential that is unexplored today, yet continues to be economical. Moreover, in order to explicitly quantify the total volumes of DSR, that which is currently existing as well as in the future, ENTSO-E should distinguish between demand net forecast for demand response and demand forecasts that include demand response.


Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices.


Ø EDF believes that TSOs can construct a vision of the possible evolutions of the generation fleet in every country in a more reliable way than using the sum of the estimations elaborated by generators. TSOs have, in fact, a global visibility of the evolution of demand




29

and interconnection capacity together with the information on the decisions already made by generators to mothball or decommission specific power plants. This can be used to assess future generation adequacy through the introduction of an analysis of the economic viability of the generation portfolio.

The elaboration of national adequacy assessments allows TSOs to collect up-to-date information on the future availability of power plants (e.g., based on the declarations of generators) and on the energy policy targets defined by public authorities. Therefore, this information can be shared by TSOs with ENTSO-E to build the MAF upon trustworthy assumptions.

Yet, EDF acknowledges that adequacy assessments focused on a mid-term horizon (up to 10 years) are subject to large uncertainties, in particular with respect to the evolutions of electricity supply and, to a lesser extent, demand and interconnection capacity. In our opinion, a useful instrument to address these uncertainties would be the elaboration of scenarios anticipating sufficiently differentiated trends in the evolution of the energy mix with different consequences in terms of adequacy.

Answers from ENTSO-E: ENTSO-E is aware of the impact upon adequacy of designing an optimal and economically feasible set of investment plans within a generation portfolio. TSOs rely on decisions made by generators to mothball or decommission specific power plants.

Generally speaking, when TSOs provide their best estimate forecast regarding NGC and demand evolution, they typically utilise information provided to them by their national market parties, generators and, in some cases, data approved first by their NRAs. However, the availability of this data is not widespread across the EU and varies significantly between countries. Such heterogeneity presents a challenge and a risk of inconsistency for the scenarios that ENTSO-E is able to collect data on from TSOs.

A detailed analysis of the economic viability of the generation portfolio is therefore difficult without proper visibility into/reliable data on power plant availability/decommissioning/mothballing plans. See the Q4 answers below on ‘Economic Viability’.






30

As already mentioned, EDF believes that the information held by TSOs, as a result of their activities, on the functioning, economics and possible evolutions of the electricity system are sufficient to formulate a perspective on the economic viability of the generation portfolio considered in the MAF. Furthermore, a strengthened cooperation between TSOs in the elaboration of the next iteration of the MAF will help improve the quality and consistency of the assumptions used across Europe in order ensure the robustness of the Pan-European scenarios proposed.

EDF believes that ENTSO-E is not supposed to perform a detailed assessment of the economic viability of power plants unit by unit but it could limit its economic viability assessment through use of the available data on the variable costs of each generation technology. This approach would permit making assumptions on the actual utility of generation units to meet demand in the time horizon of the MAF. As put forth in the answer to the next question, different scenarios can be elaborated with varying assumptions on the capacity that is likely to leave the market because of economic viability problems.


ENTSO-E is improving its data collection process regarding input data which affects the ‘likelihood of units to run and stay online’ within the market modelling assessments performed in TYNDP - MAF - these input data items are crucial to carrying any sensible sensitivity determination pertaining to ‘viability’ of the (central) best estimate scenarios collected from TSOs. ENTSO-E and the TSOs know of the importance of these assumptions regarding the definition of the scenarios and ENTSO-E’s process for a common and consistent data collection has been revised to elevate the quality of the scenarios and possibly evaluate sensitivities surrounding them.

Nonetheless, it cannot be 100% guaranteed that the forecasted generation mix considered by TSOs in their simulations will be economically viable in 2020 and 2025. For 2030 and 2040, such appraisal is even more difficult because of the intrinsic uncertainties of such a long-term horizon. As part of the adequacy assessment in future MAF reports, ENTSO-E is debating whether to conduct a portion of sensitivity analysis with a more ‘conservative’ view of installed capacity. The sensitivities will attempt to assess the impact for adequacy of parts of the generation portfolio that may be at risk of being mothballed for economic reasons, like, for example, in the absence of a capacity market and/or unfavourable market conditions – ‘missing money’ problem.

Note that the aforementioned sensitivities, through consideration of different scenarios that can be elaborated by using different assumptions on the capacity, still need to fulfil the requirements of consistent Pan-European-wide scenarios relevant to the Pan-European studies of the MAF. Economic viability appraisal is therefore linked to the construction of Pan-European consistent scenarios/sensitivities, which implicitly means that not all nationally relevant sensitivities can be captured within consistent Pan-EU scenarios. Detailed national specificities might need to be considered in depth by additional national and regional studies, which in turn need to be consistent and complement the Pan-European study.



31


Ø As mentioned above, EDF is in favour of introducing sensitivity scenarios into the MAF in order to take into account the uncertainties related to the evolution of the energy mix in a 5-to-10 year horizon. Nevertheless, EDF believes that the selected scenarios should represent different levels of stress in the electricity system that can be influenced by a variety of factors (e.g., fuel prices, macroeconomic conditions, etc.) and not only by the progression of the regulatory framework.

Thus, EDF suggests introducing into the next version of the MAF two sensitivity scenarios based on differentiated assumptions leading to varied levels of system stress, such as, for example, a low-stress and high-stress scenario:

• The “low-stress” scenario could be based on the best vision that TSOs have of the available capacities for each time horizon considering the already planned decommissionings as well as new connections (for each technology and each country), regardless of demand forecasts and interconnection capacities; and

• The “high-stress” scenario could be derived from simulations integrated in the “low-stress” scenario assumptions for variable costs for each generation technology. Indeed, the simulations made using the first scenario would show an expected number of running hours for each technology and country. To build the high-stress scenario, ENTSO-E could then assume that N% of generation units running less than M hours on average could be decommissioned (N and M would have to be appropriately defined to provide realistic figures). This analysis could be later complemented by an assessment of the probability for these power plants to cover their fixed costs.

This sensitivity analysis would allow market players and public authorities to appreciate the risk related to the absence of regulatory interventions (e.g., the introduction of capacity mechanisms, etc.) seeking to guarantee the evolution of the generation fleet necessary to meet the SoS targets defined at the national level.


i) One scenario linked to the current regulatory framework based on the energy-only market; and ii) A second scenario connected to the future regulatory framework (for example reflecting expectations of a national implementation of capacity mechanisms or any other market design instruments).



32


The baseline scenario and economic viability sensitivity planed for MAF 2017 and what has been mentioned are in agreement with the low- and high-stress proposals provided by the stakeholder feedback. We appreciate and thank the stakeholder for this comment.


Ø With respect to data assumptions (paragraph 4 on page 40), ENTSO-E highlights that data for the year 2020 should be in line with the TYNDP 2016 data used for the 2020 Expected Progress Scenario. Nevertheless, differences between MAF 2016 and TYNDP 2016 can occur because the data for MAF 2016 were collected at the beginning of 2016, after the data collection process for TYNDP 2016 (end of 2014). Therefore, it would be interesting for stakeholders to have more visibility into such differences and the reasons that led to the change of certain assumptions over the period concerned.

Answer by ENTSO-E:

Visibility of Data: ENTSO-E strives toward a high level of transparency. The so-called 'MAF2016_market_modelling_data' package, including all input net generating capacity (NGC), demand and NTC figures, has been made available upon publication of the report. ENTSO-E strives to enhance transparency regarding other key parameters used for the simulations. Transparency also implies responsibility not only of the publisher, but also of the recipient. Improvement of these key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the different generation technologies available. Such exchange could also make better the available data by TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as it concurs that the results should always be understood within the context of the data and assumptions used.

The differences between the MAF 2016 and TYNDP 2016 market modelling data packages indicate the difficult and dynamic nature of the forecasting exercise. It is common to encounter updates of data releases with > 1 year of difference. ENTSO-E is always looking to use a common format for both TYNDP and MAF market modelling data packages, so differences can be easily observed and identified. The rationale for the differences can be understood through the country comments provided by TSO in the appendix of the MAF reports. These typically relate to more conservative assumptions in relation to cross-border transmission capacity and the risks of unavailability of generation capacity because of e.g., mothballing risks or other intrinsic uncertainties.

Ø More transparency into the data used to build scenarios for hydro production are recommended in order to better understand the modelling results. ENTSO-E could make




33

available the installed capacity per country and the main assumptions employed to derive the three hydro regimes proposed - “dry”, “wet” and “normal”.

Answer by ENTSO-E: Hydro: Hydro modelling details can have a significant influence on adequacy indicators, especially of regions, including countries, with large shares of hydro production. Dry and wet hydro conditions (inflow and reservoir) have been considered in MAF 2016. In addition, the ENTSO-E PECD is currently being extended to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the probability of occurrence of these dry-normal and wet conditions is being investigated systematically in order to define so-called hydrological regions for which the same probability of similar conditions can be accounted for in Monte Carlo probabilistic simulations. ENTSO-E must fulfil obligations of Reg. 714/2009, and contribute to the overall European/national debates on adequacy concerns while respecting national legislation and confidentiality agreements between TSOs and national stakeholders. Confidentiality issues might require certain data to be publicly released in an aggregated manner by ENTSO-E.

§ EDF acknowledges that the EC ECG asked ENTSO-E to update their adequacy methodology and assessments to better account for the risks to SoS and the need for flexibility. Nevertheless, EDF wishes to highlight that capacity adequacy and flexibility are two different issues that are both necessary for a well-functioning electricity markets but should be addressed through different measures:

o Capacity adequacy is the need to ensure that enough capacity exists and isavailableintheelectricitysystemduringpeakdemandperiodsinordertoensureSoSata controlled levelof risk (inagreementwithSoS targets)andat the leastcost.Adequacyobjectivescanbeattainedthroughtheimplementationofcapacitymechanisms.

o Flexibility is the capability of the electricity system to cope with sudden injection variations (e.g., variable RES production). This capability should be properly valued in intra-day and balancing markets or by reserving balancing capacity.

Answers from ENTSO-E: ECG: ENTSO-E welcomes the ECG as a framework for TSOs, NRAs and member states, ACER, EC and ENTSO-E to interact at the relevant and necessary technical level regarding regional assessments of SoS and common definitions as well as in terms of discussion to ensure regional coordination regarding SoS measures by member states. ENTSO-E MAF adequacy reports provide an important input for discussions at the ECG. However, choosing to implement measures to ensure SoS at the national level are directly connected to the responsibility for SoS. Member states, as entities responsible for SoS, should make the decision to implement or not measures to guarantee SoS, and this should be carried out in coordination with neighbouring member states and TSOs as a consequence of the cross-border impact that such decisions may have. National



34

studies, if consistent methodologically with Pan-European studies, might be needed to complete the assessment of the MAF studies and facilitate the best possible informed decisions by member states.

The Total Need for Reliable/Dispatchable Capacity (MW) and the Capacity Surplus/Deficit (MW):

ENTSO-E will consider carefully the possibility to report on the figures described previously in the presentation of the results of future MAF reports. Providing views on specific generation technologies is not necessarily the role of TSOs as regulated entities. These indicators appear connected to the deterministic indicator, remaining capacity margin (RCM), used by ENTSO-E in SO&AF 2015 and previous reports. This indicator was not chosen in MAF 2016 in order to focus attention on the new probabilistic method deployed and the LOLE/ENS indicators used. Reporting the RCM back into forthcoming MAF reports will be considered.

Flexibility:Residual load analysis, as considered in SO&AF2015, will be appraised for future MAF reports in order to inform them with respect to the need for flexibility necessary in power systems, typically within intra-day and balancing markets. One of the main goals is to be able to integrate 'need for flexibility' into the system. Flexibility in adequacy assessments is based on weather-dependent effects related to load variation, generation patterns of wind and solar power plants with a one-hour resolution and the consideration of the resources for flexibility. The existing PECD will be used for the adequacy assessments.



35

COMMENTS BY STAKEHOLDERS Regulation 714/2009 defines that ENTSO-E must produce a “generation adequacy outlook” that has European dimensions and be built bottom-up from TSOs’ national assessments. Coupling these two elements leads to the following consistency issues. First, the need for coordination in terms of (a) methodology used, (b) criteria applied to construct the different demand and generation mix scenarios and (c) reliability metrics is problematic. These three elements should be defined in a clear and transparent manner so that the national assessments upon which the European assessment is constructed are comparable with respect to what they represent. This does not mean that more sophisticated national assessments might be carried out - they might be needed to reflect in detail local variations, specificities and circumstances, which is too ambitious an objective for a European assessment. Second, the requirement for coordination of cross-border contributions to reliability to avoid inconsistencies between the European and national assessments poses a challenge. It seems that a European assessment could be a valuable tool to analyse such contributions when considering the flow-based approach to determine the interconnection capacities and the rules governing market coupling, although this would somewhat compromise the bottom-up approach prescribed in the current EU regulation (i.e., tackling this would necessitate a kind of top-down iteration). Without these coordination vectors, the European assessment risks (a) turning into a patchwork of national assessments that have nothing to do with each other and, hence, cannot be combined in any sensible manner, and (b) suffer from an internal inconsistency with regards to cross-border contributions considered in the European and national assessments. Both issues might severely distort the results of the European assessment.

Answers from ENTSO-E: On articulation between national - regional - Pan-European reports -- methodology and report scope



IBERDROLA SA



36




ENTSO-E’s ability to propose common definitions and methodologies is unchallenged. However, the diversity of SoS concerns across Europe entails more complex modelling and processes, ultimately requiring trade-offs to comply with the many diverse issues, computing power and actual data availability from all 42 TSOs in 35 countries, and more fundamentally with conflicting model features (e.g., hydro-modelling for Nordic and Alpine countries based on different logics, unit commitments against forced outages, thermal sensitivities to demand, etc.).

Hence, Pan-European, regional and national studies are an absolute necessity as they complement each other rather than compete - studies covering a larger perimeter (Pan-European compared to regional or regional/local) are limited by potentially contradicting modelling goals and challenging data collection, but provide a consistent appraisal, if not of highest accuracy, of a series of commonly and continuously assessed indicators at the Pan-European level. Conversely, more local studies need the consistent boundary conditions supplied by the larger perimeter studies to address more precisely certain specific issues. Regional and national studies shall use the common definition and methodology standards compiled by ENTSO-E to enable comparison. For a given country or region, the prevailing conclusions and decision making regarding market design adaptations, to ensure the desired SoS level, shall stem from the study implementing the most advanced modelling for the investigated parameter. ENTSO-E adequacy studies will therefore be needed as a pre-requisite for more local regional and national studies, but they will be the reference to decide whether to implement risk mitigation measures only where no more detailed study is performed. TSO/ENTSO-E processes and bottom-up common methodology along with the use of several TSO tools across the EU perimeter in MAF makes certain there is consistency and articulation between Pan-European (MAF), regional (e.g., PLEF) and national studies in this respect.

COMMENT BY STAKEHOLDER

1. Use 35 climatic years instead of 14

2. Model the flow-based interconnection capacity calculation

3. Explicit modelling of demand response




37

Other basic additional enhancements should include:

4. Increase transparency

5. Improve the hydro modelling

6. Treat forced outages as non-independent events

7. The reliability metrics resulting from the assessment should be compared to the actual reliability standards of each country

8. Methodologies for building the future total and peak demand probability distribution in each country should be included and elaborated upon

9. There should be definitions of several generation mix scenarios.

As a conclusion, considering both the inconsistency issues described in the previous question along with the limited number of risk factors analysed (i.e., climatic and hydro conditions and forced outages), it must be concluded that MAF 2016 does not reflect a realistic adequacy outlook and, hence, should not be the basis for any energy policy or regulatory decisions. Significant improvements must be incorporated into future MAF versions, especially in terms of worst cases of demand and generation mixes.

Answers from ENTSO-E Methodological improvements


The purpose of the ENTSO-E adequacy methodology and MAF reports under evolution is the development of a robust methodology which will allow conducting regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Although these studies will not encompass every potential issue of future power system, ENTSO-E strongly believes that they: i) Will permit the development of a model able to capture all key features and risks regarding adequacy for the Pan-European power system. This is a main objective for ENTSO-E;



ENTSO-E is also aware that the choice of the mathematical approach could significantly affect the indicators that may be assessed through the simulation as well as the structure and complexity of the input data and modelling assumptions used. ENTSO-E is additionally cognizant of the impact that designing an optimal and economically feasible set of investments plans in generation portfolio has on adequacy levels. We refer to the answers below.



38





Visibility of Data: ENTSO-E strives toward a high level of transparency. The so-called 'MAF2016_market_modelling_data' package, including all input net generating capacity (NGC), demand and NTC figures, has been made available upon publication of the report. ENTSO-E strives to enhance transparency regarding other key parameters used for the simulations. Transparency also implies responsibility not only of the publisher, but also of the recipient. Improvement of these key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the different generation technologies available. Such exchange could also make better the available data by TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as it concurs that the results should always be understood within the context of the data and assumptions used. Transparency of Algorithms:



39

Different market modelling tools available, owned or procured by TSOs are utilised during the MAF exercise. These tools are proven to capture nationally and regionally relevant specificities. MAF work relies on this know-how from TSOs in terms of the strong bottom-up basis of all knowledge and detailed data. Tool dependency and reliability is a major focus for MAF. Use of exclusively open-source tools might be beneficial from a purely ‘academic’ point of view, though on the other hand, does not guarantee a strong link to bottom-up national details – know-how of TSOs and national studies. Full transparency is not always possible because these tools are, in certain cases, proprietary materials from different tool providers in a similar way that generation and market data is confidential to market parties as communicated via the stakeholder feedback and answers throughout this document. Role and Impact that the Results of the MAF Reports:

The MAF 2016 report features an ‘Overview Table’ on page 3 complemented by an appendix (Appendix 2) presenting the comments from each national TSO on the results in MAF 2016 in relation to each country’s own assessment of adequacy levels, its national adequacy standards and the measures taken to maintain them in the case of problems.The choice of LOLE = 1 hour as a threshold in such an overview table was chosen only because of technical reasons related to the presentation of the results. Only values of LOLE > 1 hour after averaging the results of the whole probabilistic Monte Carlo exercise were deemed significant enough to be highlighted, i.e.. the occurrence of adequacy problems across all the Monte Carlo ensemble of situations analysed was deemed significant rather than marginal. This choice of 1 hour should NOT be interpreted as a statement by ENTSO-E or its member TSOs on any proposals of harmonized SoS standards for EU of LoLE = 1 hour (!).

We know that such a choice of threshold for the previously referred to overview table could be nonetheless confusing and leads to misinterpretation. ENTSO-E will carefully consider the role and impact of the presentation of the results in future MAF reports. ENTSO-E is aware of the impact that the MAF results might have on discussions by member states on regionally coordinated definitions of those standards. Total Need for Reliable/Dispatchable Capacity (MW):


ENTSO-E will cautiously account for the possibility of reporting the aforementioned Figure in the presentation of the results in future MAF reports. Providing views on specific generation technologies is not necessarily the role of TSOs as regulated entities. This indicator seems related to the deterministic indicator, remaining capacity margin (RCM), employed by ENTSO-E in SO&AF 2015 and previous reports. This indicator was not chosen in MAF 2016 in order to focus attention to the new probabilistic method deployed and LOLE and ENS indicators used. Reporting the RCM again in forthcoming MAF reports will be considered. Transparency of Key Parameters, Assumptions and Data:

ENTSO-E must fulfil obligations of Reg. 714/2009 and contribute to the overall European/national debates on adequacy concerns. ENTSO-E is committed to transparency of data and products delivered while respecting national legislation and confidentiality agreements between TSOs and



40

national stakeholders. Furthermore, confidentiality issues might also require data to be publically released in an aggregated manner. ENTSO-E endeavours towards a high level of transparency on key parameters and assumptions. Improvement of key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the various generation technologies available. Such exchange could enhance the available data from TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as ENTSO-E agrees that the results should always be understood within the context of the data and assumptions used.


ENTSO-EmustfulfilobligationsofReg.714/2009.TheEuropeangenerationadequacyoutlookshall build upon national generation adequacy outlooks prepared by each individualtransmission system operator. The first goal of ENTSO-E is to collect and set up a contiguous(central)bestestimate‘Pan-European’scenariobasedondatacollectedfromTSOsasmeaningfulsensitivitiesaredefinedaroundthis.ENTSO-Eisthereforerevisingitsdatacollectionandscenario-building process in order to consider relevant sensitivities based on data provided by TSOs on,e.g., units at risk of being mothballed for economic reasons, information regardingdecommissioningofunitsforpurelytechnicalreasonsand/orshutdownsforlegal/policyreasonswithinthestorylineassumptionsofthescenariosconsidered.

Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios, which in turn needs to be consistently linked to the Pan-European scenario to foster comparison and completeness.

Considering the above, what additional methodological improvements should ENTSO-E consider for future MAFs? Please justify why through suggestions for methodological improvements - what will these improvements bring?


Although ENTSOE has made relevant methodological improvements to the MAF 2016, as a probabilistic assessment approach (with its associated new reliability metrics) or as solid geographical coverage, there is still a way to go in terms of the coordination needs described. Some improvements are already put forward in the MAF itself:



41

1. Use 35 climatic years instead of 14 – at present a too narrow set of scenarios that risk excluding the worst cases.


35 year > 14 year PECD for the whole Pan-European perimeter is a significant improvement that is sufficiently sound for the Pan-European scope and accuracy of the MAF report. This indeed does mean that some extreme, nationally specific and relevant situations could be missed. Pan-European studies consider that contextually significant and consistently constructed central EU scenarios will appraise adequacy properly but are by definition limited in the number of sensitivities of the scenarios that can be explored to capture all nationally specific extreme conditions for every country’s situation. Therefore, MAF studies should always be complemented by sound sensitivities originating within regional and national reports where detailed impacts of these risks, as sketched out at the Pan-European level by MAF, can be evaluated in greater depth by nationally specific sensitivity scenarios.

2. Model the flow-based interconnection capacity calculations, or at least refine the estimation of the NTC in order to reflect the full range of potential network situations based on variable renewable generation.


We thank the stakeholder for this relevant comment that is in agreement with the ENTSO-E roadmap for MAF methodological evolution. The modelling of seasonality of NTC (off peak, peak, other, etc.) will be considered for MAF 2017.

Flow-Based Methods

Market studies provide input to detailed grid studies. Some countries are more sensitive than others to commercial exchanges and their differences with so-called “physical exchanges”. ENTSO-E is aware that the current approach to carrying out market studies may yield certain unrealistic results corresponding to copper plates connected with HVDCs. Approaches, such as PTDF methods, facilitate the establishment of the link between market studies and grid studies by translating market results into approximated physical flows for each hour of a Monte Carlo year. Once this translation is complete, sanity checks can be performed to ensure that the bilateral transfer capacities (BTCs) between countries or price zones are not over- or underestimated. However, closing the loop is then a difficult matter.

The integration of physical grid concepts (capacities, impedances) directly into market models, usually referred to as “flow-based methods”, avoid the unrealistic results pitfall. ENTSO-E is currently determining how to define and put in place flow-based methods within long-term Pan-European market studies, the main purpose being to make commercial exchanges more realistic and faithful to actual physical flows versus the current approach. It should be noted that flow-based methods are currently deployed by a number of TSOs in national studies (Belgium and France) and are in the scope of regional studies, e.g., PLEF. Pan-



42

European studies, like MAF, will profit from these evolutions and will integrate them at the EU level once mature at the national and regional levels.

3. Explicit modelling of demand response.



Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices.

Other basic additional improvements should include:

4. Increased transparency. The MAF 2016 is not sufficiently transparency in terms of assumptions used to formulate the national generation mixes, demand/demand response/efficiency, outage rates, size of operational reserves, capacity factors, hydro modelling, etc. In addition, there is no description regarding how market coupling-models deal with simultaneous scarcity situations.


Simultaneous scarcity: ENTSO-E will cautiously consider the possibility of post-processing its results to detect situations of simultaneous scarcity on future reports. These results could act as important input for discussions by member states coordinated management of these simultaneous scarcity situations.

5. Improve hydro modelling. According to MAF 2016, hydro will represent 21% of the installed capacity in 2025. Therefore, a robust adequacy assessment is not possible without similarly robust hydro management modelling. In this sense, MAF 2016 needs significant improvements both with respect to data (from more countries) and avoidance of simplifications/subjective assumptions (for



43

instance, see the case of correlation between the hydro conditions in adjacent regions). In addition, it is not clear how the models employed deal with the possibility of successive dry years, which is not reflected either in the dry/normal/wet scenario approach or in the estimation of the water value of large multi-year reservoirs.


Detailed linear optimization market models incorporating strategy and optimization components are used to model hydro conditions in MAF reports. The strategy aspect integrates water value calculations for hydro reservoirs. Afterwards, the optimization component specifically optimizes generation for the whole market area. If hydro generation is not allocated maximally, the result may be power adequacy issues and finite LOLE values. These calculations are performed for scenarios that consider relevant and regionally consistent dry, normal and wet conditions. The countries for which dry, normal and wet conditions are accounted for are hydro-dominant countries where hydro conditions, rather than normal circumstances, have a significant impact on adequacy situations based on TSO judgement.

Additionally, the PECD is currently being expanded to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the probability of occurrence of these dry, normal and wet conditions is being investigated systematically in order to define so-called “hydrological regions” for which the same probability of similar conditions can be considered during the Monte Carlo probabilistic simulations. This improvement seeks to have a more systematic appraisal of correlation between the hydro conditions in adjacent regions while keeping the number of hydro regions within a reasonable number to ensure feasible Pan-European Monte Carlo simulations.

Successive iterations produce a representative reservoir trajectory for dry, normal or wet years. Currently, dry year reservoirs might not necessarily arrive at the same yearly end-point as their start, so there is a chance of refilling the reservoirs. This assumption is considered reasonable as the probability of two consecutive dry years is deemed to be low. The sensitivity of the adequacy results for consecutive dry years is considered low probability – local extreme situations, and is therefore thought to be more pertinent for complementary national studies rather than for a Pan-European study, like the MAF. Moreover, the consideration of simultaneous dry conditions in large hydro regions, as conducted in some runs of the MAF 2016, constitutes an ‘envelope worst-case situation’ that enables testing the ability of the interconnected system to cope with these tight hydrological situations.

6. Forced outages are not independent events. The MAF should consider the correlations of the forced outages with factors such as frequency of start/spot cycles or annual equivalent utilisation. For example, a dry year means a larger utilisation of the thermal plants and, hence, a greater probability of outages.


We thank the stakeholder for such a pertinent comment. Indeed, the global probability of forced outages might be different depending on climatic conditions, such as temperature and/or hydrological conditions.



44

The Monte Carlo approach of MAF is being continuously revisited and revised for each MAF publication in order to assess the feasibility of including these types of correlations both in the design of maintenance schedules as well as the modelling of forced outages for generation and cross-border transmissions while recognizing that these aspects, as relevant as they are, might merit a more detailed analysis through complementary national studies instead of within a full Pan-European study.

7. The reliability metrics resulting from the assessment should be compared to the actual reliability standards of each country. In the current MAF, results are compared to a “standardised” one-hour LoLE.

However, a more fundamental problem is that the MAF 2016 analyses exclusively the impact on reliability of climate and hydro conditions and forced outages. Both the expected demand and the generation mix in 2020 and 2025 are characterised with a single scenario, so the effect on reliability of the variables affecting them (GDP growth, EE and demand response development, electrification, etc.; RES penetration, potential regulatory developments at the national and EU level, economic viability of existing and new capacity, etc.; respectively) is not analysed. Therefore, MAF 2016 does not deliver a thorough assessment of the risks affecting reliability and, hence, its results are, by definition, incomplete and likely biased – i.e., underestimation of potential reliability concerns as only limited risk factors are taken into account.

An adequacy assessment should be a “stress test” rather than a “forecast”. It should aid in identifying the risks and uncertainties faced by the system, including those derived from the policy and regulatory framework, and quantify their impact on reliability. The most extreme conditions – although rare – must be appropriately introduced in the analysis as reliability is definitely about dealing with events that correspond to the tails of the probability distributions of the risk factors involved. In this sense, future MAF versions should include at least the following elements:


The MAF 2016 report features an ‘Overview Table’ on page 3 complemented by an appendix (Appendix 2) presenting the comments from each national TSO on the results in MAF 2016 in relation to each country’s own assessment of adequacy levels, its national adequacy standards and the measures taken to maintain them in the case of problems. The choice of LOLE = 1 hour as a threshold in such an overview table was chosen only because of technical reasons related to the presentation of the results. Only values of LOLE > 1 hour after averaging the results of the whole probabilistic Monte Carlo exercise were deemed significant enough to be highlighted, i.e.. the occurrence of adequacy problems across all the Monte Carlo ensemble of situations analysed was deemed significant rather than marginal.


We know that such a choice of threshold for the previously referred to Overview Table could be nonetheless confusing and leads to misinterpretation. ENTSO-E will carefully consider the role and impact of the presentation of the results in future MAF reports. ENTSO-E is aware of the impact that the MAF results might have on discussions by member states on regionally coordinated definitions of those standards.



45

Indeed, the Overview Table using a “standardised” one-hour LoLE for the colour-coding threshold appears to be misleading. We would like to apologize for this.

It should be noted however that ENTSO-E MAF report’s cannot be considered as the only source to judge whether standards are correct or not nor to justify the (the lack of/or) need for measures to ensure the desired level of adequacy. The diversity of SoS concerns across Europe entails complicated modelling and processes, ultimately requiring trade-offs and therefore complementarity between Pan-European-, regionally and nationally focused analyses of generation and resource assessments to answer these politically relevant questions satisfactorily.

Regarding the issue of whether adequacy assessments should serve as “stress tests” rather than “forecasts”, ENTSO-E is improving its data collection processes as the input data are crucial to perform any sensible adequacy assessments. The first goal of ENTSO-E is to collect and set up a consistent (central) best-estimate Pan-European scenarios based on data collected from TSOs. Meaningful sensitivities are indeed defined around this central best-estimate Pan-European scenario as ‘stress test’ cases in MAF. It should be noted that the sensitivities relevant for ENTSO-E will always be relevant to and of the scope of Pan-European scenarios. ENTSO-E welcomes the stakeholder feedback received - it supplies an important input for revising the elaboration of meaningful sensitivities. Articulation of Pan-European, regional and national studies is therefore a must as these complement rather than compete with each other. National studies must be consistent with boundary conditions and assumptions provided by the larger perimeter Pan-European and regional studies and have to use the common definitions and methodology standards defined and used by ENTSO-E.

The methodology should be common and shared between these three levels, taking into account national sensitivities and specificities. However, the decision to implement measures to ensure SoS at the national level is directly connected to the responsibility for SoS.

8. Methodology for building the future total and peak demand probability distribution in each country. This methodology should address the main variables affecting demand (i.e., GDP growth, electrification, EE potential, etc.).



Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value



46

for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices.

9. Definition of several generation mix scenarios. Turning the MAF into a fully-fledged capacity expansion model, in which the generation mix would be an output subject to economic viability and least-cost restrictions, seems to be technically unworkable, hard to reconcile with the existing regulatory framework (i.e., bottom-up) and plagued with subjectivity as it would be necessary to model the economic incentives provided by the existing market and regulatory arrangements in each country in a detailed manner with agents’ behaviour and risk aversion addressed in particular. The alternative is to consider several scenarios reflecting different policy and regulatory developments, technology progress, etc., most especially the potential worst cases. In any case, high-level criteria reflecting economic viability restrictions should be applied to such scenarios – an aspect not considered in the MAF 2016 as it depicts countries where there would be no scarcity expectations either for 2020 or 2025 (only possible assuming that agents behave irrationally so that they maintain their capacity available under any circumstances, even accepting endless losses). These scenarios should be constructed under the assumption that (a) there is no exit barrier (i.e., agents are totally free to decommission/mothball their capacity) and (b) mothballed plants are not reactivated to cope with an isolated scarcity event (i.e., reactivation driven by economics exclusively; imperfect forecast; reactivation is not an immediate process).

Answer by ENTSO-E:

ENTSO-E must fulfil obligations of Reg. 714/2009. The European generation adequacy outlook shall build upon national generation adequacy outlooks prepared by each individual transmission system operator. The first goal of ENTSO-E is to collect and set up a contiguous (central) best estimate ‘Pan-European’ scenario based on data collected from TSOs as meaningful sensitivities are defined around this.

Sensitivity of Scenarios: ENTSO-E must fulfil obligations of Reg. 714/2009. The European generation adequacy outlook shall build upon national generation adequacy outlooks prepared by each individual transmission system operator.The first goal of ENTSO-E is to collect and set up a contiguous (central) best estimate ‘Pan-European’ scenario based on data collected from TSOs as meaningful sensitivities are defined around this. ENTSO-E is therefore revising its data collection and scenario-building process in order to consider relevant sensitivities based on data provided by TSOs on, e.g., units at risk of being mothballed for economic reasons, information regarding decommissioning of units for purely technical reasons and/or shut downs for legal/policy reasons within the storyline assumptions of the scenarios considered.

Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national



47

scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios.

As a conclusion, considering both the inconsistency issues described previously and the limited number of risk factors analysed (i.e., climatic and hydro conditions and forced outages), it must be concluded that MAF 2016 does not reflect a realistic adequacy outlook and, hence, should not be the basis for any energy policy or regulatory decision. Significant improvements must be incorporated to future MAF versions, especially in terms of worst cases of demand and generation mixes.

Answer by ENTSO-E:

ENTSO-E MAF reports cannot be considered the only source to judge whether standards are correct or not nor to justify the (the lack of/or) need for measures to ensure the desired level of adequacy. The diversity of SoS concerns across Europe entails complicated modelling and processes, ultimately requiring trade-offs and therefore complementarity between Pan-European-, regionally and nationally focused analyses of generation and resource assessments to answer these politically relevant questions satisfactorily.

Articulation of Pan-European, regional and national studies is therefore a must as these complement rather than compete with each other. National studies must be consistent with boundary conditions and assumptions provided by the larger perimeter Pan-European and regional studies and have to use the common definitions and methodology standards defined and used by ENTSO-E. The methodology should be common and shared between these three levels, taking into account national sensitivities and specificities. However, the decision to implement measures to ensure SoS at the national level is directly connected to the responsibility for SoS.

The MAF report and its related Pan-European methodology have a goal of filling the gaps between member states’ assessments by providing a consistent appraisal of a series of commonly and consistently assessed indicators at the Pan-European level. Today, such common and continuous appraisal of indicators does not exist elsewhere, and ENTSO-E’s reports addresses this deficiency. Still, local/regional studies are always necessary to complete the picture painted by the Pan-European reports.


It is inefficient to oblige agents to submit decommissioning/mothballing plans 5-10 years in advance as these decisions are made in a context of great uncertainty (i.e., stagnant demand, evolution of EE and demand response, new technology developments, actual RES penetration,




48

changes in environmental requirements, price of commodities, results of regulatory initiatives, such as the new market design for the EU, etc.), so they would be utterly unreliable. Furthermore, such plans would risk distorting both the adequacy assessment itself and even competition.

The availability/decommissioning/mothballing issue is in fact directly related to the economic viability that any generation mix considered in the adequacy assessment should comply with. It is not about having better “data” on the expected situation in 5-10 years (i.e., suggested decommissioning/mothballing approaches), but making sound estimates, which inevitably requires incorporating the economic viability criterion.


Stakeholders stress the importance of ENTSO-E performing high-level economic sensitivities. ENTSO-E is revising its data-collection and scenario-building processes in order to consider relevant sensitivities based on data provided by TSOs on units at risk of being mothballed for economic reasons and information regarding decommissioning of units strictly for technical reasons, or/and shut downs based on legal/policy factors within the storyline assumptions of the scenarios considered.

It must be noted that the conclusions of the MAF report cannot be separated from the hypotheses used and can only be read in reference to them. The hypotheses are gathered by the TSOs according to their best knowledge at the time of the data collection and validated by ENTSO-E’s relevant committees. Therefore, MAF reports and sensitivities do not intend to contain all the information that a prospective investor or market participant may need. ENTSO-E emphasises that ENTSO-E and the TSOs involved in these studies cannot be responsible in the case hypotheses from these reports or the estimations based on these hypotheses are not realised in the future. Providing views on the economic viability for specific generation technologies linked to decisions by prospective investors or market participants should be the onus of market participants themselves and not by TSOs as regulated entities.

What would you recommend to us in order to improve the quality of the data and assumptions mentioned above? COMMENTS BY STAKEHOLDERS

As explained in the previous questions, turning the MAF into a fully-fledged capacity expansion model, in which the generation mix would be an output (rather than an input) subject to economic




49

viability and least-cost restrictions seems to be technically unworkable, hard to reconcile with the existing regulatory framework (i.e., bottom-up approach) and plagued with subjectivity. Furthermore, detailed economic/technical data of the existing plants is very sensitive commercial information, of which disclosure for the sake of a more “accurate” MAF would be disproportionate.


See answer above.

A proper alternative would be to (a) consider several generation scenarios reflecting different policy and regulatory developments, technological progress, etc., most especially the potential worst cases, and (b) apply to those generation scenarios high-level criteria reflecting economic viability restrictions (for instance, minimum expected annual utilisation rate, maximum mothballing period for existing plants, etc.). No detailed economic/technical data of the existing plants would be necessary to perform these analyses.


Economic/technical data on generation technologies is needed to perform market simulations. The hypotheses are gathered by the TSOs according to their best knowledge and the availability of data at the time of data collection. ENTSO-E and the TSOs involved in these studies can therefore not be held responsible in the case the hypotheses from these reports or the estimations based on these hypotheses are not realised in the future. ENTSO-E welcomes greater transparency and availability from market participants with respect to the economic/technical data to improve the accuracy of the forecasts on economic/technical data used by TSOs in the simulations performed.

Furthermore, ENTSO-E welcomes the comments provided as the economic viability sensitivities planned for the MAF reports are in line with the stakeholder feedback that ‘..no detailed economic/technical data of the existing plants would be necessary to perform these analyses..’

These generation scenarios should be built under the assumption that (a) there is no exit barrier (i.e., agents are totally free to decommission/mothball their capacity) and (b) mothballed plants are not reactivated to cope with isolated scarcity events (i.e., reactivation driven by economics exclusively and under imperfect forecast capabilities).

This approach should serve to avoid unrealistic results as those presented in MAF 2016, where it was estimated for some countries that there would be no scarcity expectations either in 2020 or in 2025 – i.e., possible only assuming agents behave irrationally so that they maintain their capacity available under any circumstances accepting endless losses.


It is very important to have visibility on decommissioning and mothballing figures in order to set up the relevant generation scenarios.



50

The MAF data collection process will be revised for MAF 2017 in order to improve visibility of these data provided by TSOs and processed to set up relevant sensitivities that will explore possibilities to consider different views on installed capacity, from ‘optimistic’ to ‘conservative’ situations. ENTSO-E welcomes the comments provided and strives for enacting economic viability sensitivities that are in agreement with stakeholder feedback.


The market and regulatory framework to be assumed should depend on the aim pursued by the adequacy assessment. If, according to MAF 2016, its objective is to “...assess potential load-shedding risks and send signals to both market players and decision makers for the generation fleet to evolve”, it seems that the most appropriate approach would be to assume that no capacity mechanism or equivalent measure is in place. This, once again, refers to the economic viability of the generation mix issue, which is in turn a key sign pointing to the existence of market and/or regulatory failures that justify the need for a capacity mechanism.

In addition, regulatory failures artificially extending overcapacity situations over time, such as exit barriers, should not be part of the regulatory framework employed to carry out the assessment. They are unsustainable over the mid-term (i.e., agents cannot be forced to make losses), so they disguise actual load-shedding risks and misguide potential investors.

Answer from ENTSO-E:

The scenarios analysed in MAF for 2020 and 2025 will be based on a best estimate of the evolution of the generation mix (thermal and renewable) and transmission capacity as well as demand forecast of each country. These scenarios are referred specifically as “best-estimate” scenarios. For 2025, this scenario will be common to the TYNDP 2018 2025 scenario. Within the principles set forth by ENTSO-E for common and consistent data collection, all TSOs will provide data considering to their best knowledge the evolution of their generation mix, in some cases including “economic viability” of the scenarios provided.





51

The scenario storylines for 2030 and 2040 of TYNDP 2018 have been, for the first time, constructed with ENTSO-G and are largely based on stakeholder, regulator and member states input.

Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios.


No Comment.




52

COMMENTS BY STAKEHOLDERS Ensuring a proper level of system adequacy during the (ongoing) energy transition, not only over the short-term but also the long-term, is extremely important to its success. ENGIE would like to emphasize two structuring elements regarding MAF studies. The first is related to the need to have robust forecasts for demand, consistent across all European countries (e.g., based on a coherent set of macro-economic assumptions) and also including proper sensitivities on certain key underlying parameters (e.g., economic activity, EE, level of decentralized generation, etc.). In this regard, ENTSO-E has a clear and important role to play, both in terms of coordination but also content. The second element is related to the decommissioning of existing assets, not only for technical reasons (e.g., end of lifetime, environmental constraints, etc.) but also for economic reasons (e.g., thermal assets not covering their fixed costs). Although obtaining a proper view of this is more complex (see Q 3.1 and Q 4.1), ENGIE stresses the fact that ENTSO-E should definitely provide as a key outcome of its system adequacy assessment at the level of reliable capacity needed for targeted adequacy criteria (for each country).

ENGIE believes that system adequacy assessments in highly interconnected countries make sense at only at the regional level, or even potentially, and better, the European level. Consistency between adequacy studies could be enhanced by sharing, in a transparent manner, the methodology deployed, the datasets utilised and the results of the computations performed.

Providing more transparency and full access to the datasets employed by ENTSO-E (e.g., outage rates, wind and solar production levels and profiles, cross-border interconnection capacities, macro-economic assumptions used for demand forecasts, level of operational reserves, etc.) is required to allow other market players to perform their own analyses on a similar basis, to challenge the assumptions and the outcomes of the various adequacy studies and to provide relevant feedback with the goal of improving these studies.

ENTSO-E should probably document in more detail within a technical appendix aspects of the methodology underlying their approach (as a complement to the high-level description found in Section 3 of the report). This could then permit proper replication by other stakeholders as well as a better understanding of various methodological choices. In addition, ENTSO-E should supply more granularity on the assumptions and methodology behind the demand forecasts.


ENGIE



53

The tools used by ENTSO-E for performing the computations (ANTARES, BID3, GRARE, PLEXOS) all have specific features and competitive advantages, but also respective drawbacks. The outcomes from the use of each tool, therefore, depends on not just the dataset used, but also on the particular features leveraged and the fundamental parameters/settings. The approach of ENTSO-E in the current MAF, which consists of several tools and blending of their results, allows partial alleviation of the modelling risk by synthetizing the outcomes of various models. ENGIE believes, however, that all market participants should have the opportunity to gain the same understanding of the results and their drivers. This is especially possible if all market participants strictly have access to the models and their technical features. Thus, ENGIE pleads to ENTSO-E to use publicly available tools (e.g., commercially available or open-source software) exclusively.


On articulation between national - regional - Pan-European reports -- methodology and report scope

ENTSO-E appreciates the positive feedback from stakeholders regarding the ENTSO-E methodology as a common basis. ENTSO-E views this role as follows: a) ENTSO-E describes and proposes common definitions for the multiple concepts/aspects under Security of Supply (SoS) and adequacy, providing all stakeholders with a common language; b) ENTSO-E creates or adopts and then standardises methodologies and processes to perform adequacy analyses and market modelling based on its members‘ expertise; and c) ENTSO-E performs Pan-European adequacy analyses relying on dedicated Pan-European data collection processes. ENTSO-E’s ability to propose common definitions and methodologies is unchallenged. However, the diversity of SoS concerns across Europe entails more complex modelling and processes, ultimately requiring trade-offs to comply with the many diverse issues, computing power and actual data availability from all 42 TSOs in 35 countries, and more fundamentally with conflicting model features (e.g., hydro-modelling for Nordic and Alpine countries based on different logics, unit commitments against forced outages, thermal sensitivities to demand, etc.). Hence, Pan-European, regional and national studies are an absolute necessity as they complement each other rather than compete - studies covering a larger perimeter (Pan-European compared to regional or regional/local) are limited by potentially contradicting modelling goals and challenging data collection, but provide a consistent appraisal, if not of highest accuracy, of a series of commonly and continuously assessed indicators at the Pan-European level. Conversely, more local studies need the consistent boundary conditions supplied by the larger perimeter studies to address more precisely certain specific issues.

Regional and national studies shall use the common definition and methodology standards compiled by ENTSO-E to enable comparison. For a given country or region, the prevailing conclusions and decision making regarding market design adaptations, to ensure the desired SoS level, shall stem from the study implementing the most advanced modelling for the investigated parameter. ENTSO-E adequacy studies will therefore be needed as a pre-requisite for more local regional and national studies, but they will be the reference to decide whether to implement risk mitigation measures only where no more detailed study is performed. TSO/ENTSO-E processes and bottom-up common methodology along with the use of several TSO tools across the EU



54

perimeter in MAF makes certain there is consistency and articulation between Pan-European (MAF), regional (e.g., PLEF) and national studies in this respect.


The purpose of the ENTSO-E adequacy methodology and MAF reports under evolution is the development of a robust methodology which will allow conducting regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Although these studies will not encompass every potential issue of future power system, ENTSO-E strongly believes that they: i) Will permit the development of a model able to capture all key features and risks regarding adequacy for the Pan-European power system. This is a main objective for ENTSO-E;





Improvements made/identified for MAF 2016:

First, the methodological developments already implemented by ENTSO-E to move from a deterministic to a probabilistic approach should be properly acknowledged. Although they already represent a significant improvement and a huge effort, further improvements should be carried out in the coming years. ENGIE welcomes the fact that ENTSO-E is already aware of limitations of its analyses and is considering additional improvements (number of climatic years, assumptions on cross-border capacities, demand response, etc.).

***

Adequacy metrics (§ 3.1.1):

ENGIE believes that the set of adequacy metrics considered in MAF 2016 has to be further complemented by at least the following:




55

• The total need for reliable/dispatchable capacity (MW): distribution, average, median, p95, standard deviation, etc.. This would assist the various stakeholders in the market to evaluate the need for additional investments/divestments based on their own perspectives on the development of existing assets. Moreover, it could become instrumental in setting up the capacity demand in capacity remuneration mechanisms (see also Q 3.1 and Q 4.1).

• The capacity surplus/deficit (MW): distribution, average, median, p95, standard deviation, etc.

Adequacy criteria/reporting:

As highlighted by DG Competition in its interim report of sector inquiry into capacity mechanisms (see Section 4), European countries have different views and practices regarding adequacy metrics and criteria. In particular, certain countries do not have legal reliability standards and, when available, there is not necessarily a link between these criteria, VoLL, the cost of new entry, etc. As underscored by DG Competition, this raises the question of how to ensure that reliability standards are based on comprehensive economic assessments.

Regarding this factor, the table on page 13 of the MAF report is clearly misleading, at least for the following reasons:

(i) in practice, not all countries share the same reliability standard;

(ii) where existing, availability standards are not homogeneous (metrics and thresholds) – for instance, the same value of LOLE might be considered sufficient in one country and not in another;

(iii) the use of a colour code (green/red/blue) in the table expressed an implicit statement on the part of ENTSO-E regarding the adequacy situation in the countries covered by the study, while neither the criteria (LOLE) nor the threshold value (one hour) is justified.

ENGIE insists that this table should thereby be either deleted or substantially amended with the proper information (e.g., adequacy metric(s) and threshold(s) per country).

Similarly, ENGIE believes that certain reporting published in the MAF report could be improved. As an example, adequacy assessment criteria are often related to LOLE – e.g., in Belgium, the LOLE average should be less than three hours and the LOLE p95 should be less than 20 hours. It would thus be helpful to present data like that in Figure 5 (page 17) also in terms of LOLE, not strictly ENS.

Finally, as a side yet very important comment, the colour codes used for France in this table (red cells for all three cases in 2020) do not match the data provided in Table 2 (LOLE = 0.8 and not > 1). This should be corrected in the final version.

***

Demand-side assumptions:



56

ENGIE maintains that the ongoing energy transition requires regional adequacy assessments to be conducted on a sound basis. One of the main elements in this analysis is clearly the demand forecasts (levels, but also profiles) for the future. Unfortunately, the assumptions upon which the load/demand data are not fully transparent in the MAF report open for consultation. In particular, it is not clear whether all countries are basing their forecasts on coherent underlying assumptions for key parameters (e.g., GDP growth, EE, etc.). ENGIE believes that ENTSO-E should make certain there is consistency in the forecasts considered in the analysis and, if needed, propose a path forward for greater harmonization.

***

Supply-side assumptions:

Similarly, supply-side assumptions are extremely vital when assessing system adequacy as they represent the “second leg” of the rationale. ENTSO-E should hence take particular care when devising the capacity assumed reliable in the future, and perform a sensitivity analysis where needed. Specifically, and given the energy transition, ENTSO-E should facilitate capacity considered as reliable in the future is also economically viable. Otherwise, the analysis could depend on capacity that might not be present at that moment in time and – a fortiori – that cannot contribute to SoS. This would significantly lower expected system adequacy.

For instance, ENTSO-E is accounting for, in its analysis, that mothballed assets could return in the case of capacity shortages and therefore be available by 2020 and/or 2025 (see e.g., page 76). However, mothballing an asset is the preliminary step towards its full decommissioning, and takes while a more in-depth analysis of its future expected economic viability is performed. ENGIE believes that mothballed assets should thus be excluded from the installed capacity in the “baseline case” of the adequacy assessment. Recent years have shown that the European markets, in their current design, do not reward capacity. Market design changes have been debated in some countries (see the development of capacity mechanisms) while in others, large volumes of generation are set aside in strategic reserves or contracted more or less transparently by TSOs (see 8 GW mentioned in AF German TSO discussion). In this context, the MAF assumption (all capacities have to be modelled as online for system adequacy assessment, even if currently mothballed) is far too optimistic unless the capacity is selected in a capacity market covering the area (see Irish TSO remarks in the MAF about closures of capacities not successful in the CM auction). In practice, ENGIE recommends that market-based capacity mechanisms are put in place so that the assets needed for SoS cover their (fixed) costs. This would incentivize the currently mothballed assets either to return to active duty (safeguarding SoS) or are closed and completely leave the market.

Decentralized generation:

Given the ongoing energy transition, decentralized generation is expected to play a more and more critical part in the future electricity supply-demand balance. It is hence extremely necessary that ENTSO-E is able to rely on accurate figures for existing capacity (and related generation) of decentralized generation as well as for the expected development (coherent with other assumptions, like RES deployment targets). This also requires other actors in the electricity systems (like DSOs) to practice more visibility and transparency with regards to the potential



57

impact of their activities on system adequacy (e.g., aggregated information on “prosumers” connected to the grid in terms of type, capacity, generation, etc.).

Operational reserves modelling:

First and most importantly, ENTSO-E factors in the sensitivity analyses (Sensitivity Cases I & II) the assumption that operational reserves could contribute to system adequacy in 2020 and 2025. However, the operational reserves are necessary in the European electricity system to ensure management and stability in real-time and they should therefore be available at all times (at least to the extent possible unless they are activated and not yet recovered). Otherwise, the risk of a blackout (complete failure of the system, as opposed to a brownout where consumption is partially curtailed) is underestimated. ENGIE contends that the correct assumption is that operational reserves do not contribute to system adequacy, which is the assumption retained in the “baseline case”.

Second, ENTSO-E should offer a view on the assumptions for operational reserves (no detailed figures were found in the report nor in the data package).

Flexibility:

Assessing whether adequate flexibility exists in the system to cope with large and fast load variations, or excess renewable energy in the summer, will probably become crucial as LOLE estimates in the near future (including those for the summer with inflexible generation and renewable sometimes exceeding demand). The Belgian TSO has looked at the issue in its latest adequacy report, but the flexibility requirements discussed are mostly those of the grid and not of providers - the harmonisation of flexibility analysis methods is sorely needed.

As acknowledged by ENTSO-E, further work should be performed regarding the inclusion of demand response in the next MAF exercises. ENGIE believes that the potential of this source of flexibility should be properly estimated provided its contribution to SoS in a market with more and more intermittent RES generation.

Similarly, taking into account the dispatch flexibility of hydro assets (e.g., pumped storages) being critical for reliability appraisal, especially with greater renewable resources, such as solar and wind generation. The hydro resources in Europe (e.g., located in the Alps, Norway, Spain, Italy, etc.) could economically adapt their dispatch throughout the year(s) according to the expected risk of unsatisfied demand (scarcity). ENGIE would like ENTSO-E to venture beyond the probabilistic approach currently taken in the MAF report, which is based on scenarios of hydro generation and likely to underestimate the contribution to flexibility and SoS of hydro. For instance, an stochastic dynamic dual programming (SDDP)-based tool could probably be potentially evaluated as a tool for ENTSO-E.

Weather scenarios:

The weather scenarios made use of by ENTSO-E are too narrow. We had several years of extreme weather in the recent past and climate change makes older events (e.g., winter 1954) less relevant. According to Meteo France, the “ [climate change] trend calculated for the 1979-2005 period is



58

0.55 ° C/decade [in France]”. In that context, even the larger scenario set used by the MAF is very restricted. It would be useful to adopt for the MAF a larger set of scenarios at “current climate”, for instance the 200 scenarios developed recently for RTE that seem to cover all of Europe (see https://cpr.concerte.fr/sites/default/files/20160229_Sc%C3%A9narios%20climatiques_Meteo%20France.pdf).

In addition, it is not clear whether the weather scenarios are correlated with other elements in the analysis (e.g., outage rates). As a case in point, extreme cold or warm events can lead to loss of generation capacity far in excess of that predicted by usual models (reduced cooling source availability, frozen captors, etc.). While modelling these events is quite difficult and data intensive, they should be accounted for in the long-term for inclusion in the MAF methodology.

Interconnections:

In a highly interconnected electricity system, the availability of interconnections can influence the system adequacy of a country or a region significantly. In the current exercise, interconnections are considered according to an ATC approach. ENGIE believes that a better representation of the flow-based market coupling, including a link between meteorological conditions (wind and solar production) and network availability, would be beneficial.

Stopping criteria:

In section3.1.2of theMAFreportopen forconsultation, it isnot fullyclearwhichstoppingcriteriaareutilisedbyENTSO-Eandwhethertheyareappropriateforsatisfactoryinsightsintotheadequacysituationofacountry/region. Answers from ENTSO-E

Methodological improvements

We thank the stakeholder for their constructive comments that are in line with the scope of the methodological evolution foreseen for forthcoming MAFs. The purpose of the ENTSO-E adequacy methodology and MAF reports under evolution is the development of a robust methodology which will allow conducting regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks.

Although these studies will not encompass every potential issue of future power system, ENTSO-E strongly believes that they: i) Will permit the development of a model able to capture all key features and risks regarding adequacy for the Pan-European power system. This is a main objective for ENTSO-E;






59






Visibility of Data: ENTSO-E strives toward a high level of transparency. The so-called 'MAF2016_market_modelling_data' package, including all input net generating capacity (NGC), demand and NTC figures, has been made available upon publication of the report. ENTSO-E strives to enhance transparency regarding other key parameters used for the simulations. Transparency also implies responsibility not only of the publisher, but also of the recipient. Improvement of these key parameters would be possible by discussion with relevant market



60

parties/actors that can provide detailed technical data and feedback regarding the different generation technologies available. Such exchange could also make better the available data by TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as it concurs that the results should always be understood within the context of the data and assumptions used. Transparency of Algorithms:

Different market modelling tools available, owned or procured by TSOs are utilised during the MAF exercise. These tools are proven to capture nationally and regionally relevant specificities. MAF work relies on this know-how from TSOs in terms of the strong bottom-up basis of all knowledge and detailed data. Tool dependency and reliability is a major focus for MAF. Use of exclusively open-source tools might be beneficial from a purely ‘academic’ point of view, though on the other hand, does not guarantee a strong link to bottom-up national details – know-how of TSOs and national studies. Full transparency is not always possible because these tools are, in certain cases, proprietary materials from different tool providers in a similar way that generation and market data is confidential to market parties as communicated via the stakeholder feedback and answers throughout this document. Role and Impact that the Results of the MAF Reports:

The MAF 2016 report features an ‘Overview Table’ on page 3 complemented by an appendix (Appendix 2) presenting the comments from each national TSO on the results in MAF 2016 in relation to each country’s own assessment of adequacy levels, its national adequacy standards and the measures taken to maintain them in the case of problems. The choice of LOLE = 1 hour as a threshold in such an overview table was chosen only because of technical reasons related to the presentation of the results. Only values of LOLE > 1 hour after averaging the results of the whole probabilistic Monte Carlo exercise were deemed significant enough to be highlighted, i.e.. the occurrence of adequacy problems across all the Monte Carlo ensemble of situations analysed was deemed significant rather than marginal. This choice of 1 hour should NOT be interpreted as a statement by ENTSO-E or its member TSOs on any proposals of harmonized SoS standards for EU of LoLE = 1 hour (!). We know that such a choice of threshold for the previously referred to Overview Table could be nonetheless confusing and leads to misinterpretation. ENTSO-E will carefully consider the role and impact of the presentation of the results in future MAF reports. ENTSO-E is aware of the impact that the MAF results might have on discussions by member states on regionally coordinated definitions of those standards.

Total Need for Reliable/Dispatchable Capacity (MW):


ENTSO-E will cautiously account for the possibility of reporting the aforementioned Figure in the presentation of the results in future MAF reports. Providing views on specific generation technologies is not necessarily the role of TSOs as regulated entities. This indicator seems related to the deterministic indicator, remaining capacity margin (RCM), employed by ENTSO-E in SO&AF 2015 and previous reports. This indicator was not chosen in MAF 2016 in order to focus



61

attention to the new probabilistic method deployed and LOLE and ENS indicators used. Reporting the RCM again in forthcoming MAF reports will be considered. Transparency of Key Parameters, Assumptions and Data:



ENTSO-E must fulfil obligations of Reg. 714/2009. The European generation adequacy outlook shall build upon national generation adequacy outlooks prepared by each individual transmission system operator.The first goal of ENTSO-E is to collect and set up a contiguous (central) best estimate ‘Pan-European’ scenario based on data collected from TSOs as meaningful sensitivities are defined around this. ENTSO-E is therefore revising its data collection and scenario-building process in order to consider relevant sensitivities based on data provided by TSOs on, e.g., units at risk of being mothballed for economic reasons, information regarding decommissioning of units for purely technical reasons and/or shut downs for legal/policy reasons within the storyline assumptions of the scenarios considered. Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios. Flow-Based Methods

Market studies provide input to detailed grid studies. Some countries are more sensitive than others to commercial exchanges and their differences with so-called “physical exchanges”. ENTSO-E is aware that the current approach to carrying out market studies may yield certain unrealistic results corresponding to copper plates connected with HVDCs. Approaches, such as PTDF methods, facilitate the establishment of the link between market studies and grid studies by translating market results into approximated physical flows for each hour of a Monte Carlo year.



62

Once this translation is complete, sanity checks can be performed to ensure that the bilateral transfer capacities (BTCs) between countries or price zones are not over- or underestimated. However, closing the loop is then a difficult matter. The integration of physical grid concepts (capacities, impedances) directly into market models, usually referred to as “flow-based methods”, avoid the unrealistic results pitfall. ENTSO-E is currently determining how to define and put in place flow-based methods within long-term Pan-European market studies, the main purpose being to make commercial exchanges more realistic and faithful to actual physical flows versus the current approach.


NTC – Conservative. Conservative assumptions have been made regarding cross-border transmission capacity relating to uncertainty in the commissioning dates of cross-border transmission projects and adequacy risks therein. These assumptions are within the scope of the methodology and report capturing all key features and risks relevant at the Pan-European level - this is a main objective for ENTSO-E. A conservative approach of ENTSO-E in providing transfer capacity values for adequacy analyses is sensible, even if adequacy risk might be overestimated. This will mean that the energy mix considered will not be compatible with the assumed/available transmission infrastructure. In this respect, the MAF results provide a signal in the form of ‘adequacy risk’. This could very well be higher than the desired adequacy level that a country is willing to accept or prepare for, expressed in the form of its adequacy standard. As a consequence of the fact that investments in grid infrastructure require a longer planning and decision horizon (~15 years) than the typical investment cycles for investments in new generation (3-5-7) years, the MAF exercise will detect risks and provide important feedback regarding the establishment of countermeasures by relevant stakeholders (e.g., member state authorities, policy makers, regulatory agencies, energy producers) in order to ensure the desired adequacy levels and the risks linked to their choices of generation mix and its evolution as well the impact of cross-border capacity availability. Furthermore, national or regional assessments might consider different assumptions that could be more optimistic regarding availability of cross-border transmission infrastructure and availability of imports to complete the assessment of the impact of such assumptions on adequacy levels.

Frequency Containment Reserves (FCR): Operating reserves necessary for constant containment of frequency deviations (fluctuations) from nominal value in order to constantly maintain the power balance in the whole synchronously interconnected system. Activation of these reserves results in a restored power balance at a frequency deviating from the nominal value. This category typically includes operating reserves with an activation time of up to 30 seconds. Operating reserves of this category are usually activated automatically and locally. Frequency Restauration Reserves (FRR): Operating reserves are there to restore the imbalance of a control area after an outage or forecast error for that area. They should be available within 7.5 minutes (automatic) or 15 minutes (manual). These can be contracted or procured by TSOs through the market but the TSO must guarantee their availability in accordance with the reserve requirements of each control area. FRR should be always available.



63

Modelling of Operational Reserves: With MAF, the main imbalances addressed in the simulations relate to planned outages (maintenance schedules) and forced (unplanned) outages. Forced outages are modelled in MAF as occurring randomly at a given hour, though can have different durations in reality.

If a forced outage happens and lasts for > 24 hours, it is visible in the day-ahead (D-1) market of the next x day(s). This unit is then not available neither for the market nor for the balancing reserves (FCR, FRR). So, in the end, capacity that has to be provided for the D-1 is the sum of what is required to cover the load + FCR + FRR requirements. This is what is modelled in MAF when modelling FCR + FRR as an extra load. If a forced outage is of a short duration, lasting a few hours (a “failed to start”, for example, which is very common), then the balancing responsible party (BRP) might not have time to recover this capacity in the market (intra-day market gate closing being longer than the outage). In this case, TSO will need to restore such an imbalance with the ‘remaining’ available balancing reserves. Under these circumstances, the modelling FCR + FRR as an extra load might represent a double-counting during the assessment as reserves might be used to cover the FOR fault. Even if those short outages are able to have a hourly instantaneous high impact to the adequacy level, their weight in the total forced outage rate calculated by TSOs and used in MAF is low (longer outages have much more weight), so the forced outage rates deployed in MAF account mostly for long-lasting outages impacting the adequacy situation significantly in a way that it is visible to the day-ahead market for subsequent days.

Other arguments favour the modelling of FCR + FRR as an extra load - BRPs are/will be more and more incentivized to restore their balances in real-time, so if an outage takes place within their portfolio, they need to recover it as soon as possible without the intervention of TSOs employing FCR and FRR. These reserves are not designed to cover shortfalls in capacity or energy in the market but to ensure frequency stability. Furthermore, not only imbalances because of forced outages are covered by balancing reserves. Forecasting errors for load, wind and solar production are also addressed. With the increase of RES capacity, it is occasionally difficult to forecast hour by hour (even if improvements are made year after year), so this is more and more important. ENTSO-E aims to achieve a high level of transparency for key parameters and assumptions. ENTSO-E will consider making available the values of FCR + FRR assigned as an extra load in the simulations.

15-Minute Flexibility: The purpose of the ENTSO-E target adequacy methodology being developed is the generation of a sound methodology that will foster performing regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Considering any modelling implies simplifications, so ENTSO-E adequacy target modelling focuses on hourly power balance modelling. Although these studies do not encompass every potential issue involved in the state of the future power system, ENTSO-E strongly believes that this methodology will permit the development of a model able to capture all the key features and risks regarding adequacy for the Pan-European power system - this is a main objective for ENTSO-E.

The effect of imbalances based on forced outages occurring close to real-time and the effect of forecasting errors of load, wind and solar production is accounted for in MAF by consideration of operational reserve requirements as an extra constraint while modelling. The detailed modelling



64

of these aspects is outside of the scope of MAF as these imbalances are typically covered by the intra-day and primarily balancing markets and are not observed systematically in the day-ahead market during several days of a week. Flexibility: Residual load analysis, as considered in SO&AF2015, will be appraised for future MAF reports in order to inform them with respect to the need for flexibility necessary in power systems, typically within intra-day and balancing markets. One of the main goals is to be able to integrate 'need for flexibility' into the system. Flexibility in adequacy assessments is based on weather-dependent effects related to load variation, generation patterns of wind and solar power plants with a one-hour resolution and the consideration of the resources for flexibility. The existing PECD will be used for the adequacy assessments.

Considering the above, what additional methodological improvements shall ENTSO-E consider for future MAFs? Please justify why through suggestions for methodological improvements and what these improvements will offer.


The improvements suggested in Q 2.1 should enhance the quality of the approach followed by ENTSO-E, the confidence with respect to the results yielded as well as their usefulness for all market players, especially regarding the capacity needed to ensure reliable supply in the coming 5 to 10 years. Additional justifications are provided next to the suggestions in the previous question.


Seeanswersaboveandbelow.

COMMENTS BY STAKEHOLDERS The fact that ENTSO-E does not have full visibility into availability/decommissioning/mothballing of power plants for the next 5 to 10 years is unfortunate, but obviously normal in the European market context. To start, it goes beyond the information available on transparency platforms and the requirements for transparency set by market players. Next, the market players themselves do not have a precise view of their own assets over this time horizon, excepting some specific assets (e.g., nuclear) or given specific constraints (e.g. technical lifetime, Industrial Emission Directive compliance). Regarding mothballing, ENGIE suggests that market-based capacity mechanisms are introduced so that the assets needed for SoS cover their (fixed) costs. This would incentivize the




65

currently mothballed assets either to return to active duty (and ensure SoS) or closed and completely leave the market. Given the range of possible decisions regarding investment/decommissioning/mothballing and the impact they could have system adequacy, it would be wise for ENTSO-E to carry out a proper sensitivity analysis on the installed/reliable generation capacity or provide the relationship between the level of reliable capacity and adequacy. A first approach to determine proper sensitivities on the reliable capacity available would be to perform an economic assessment based on several classes of assets (e.g., old or new coal units, CCGT units, etc.) and on certain assumptions on the fixed costs (i.e., costs to be covered at least sufficiently to maintain the corresponding units in the system). However, this approach might be very time-consuming with a Monte Carlo simulation approach (the plants’ profitability needs to be verified, the capacity level has to be adapted and the simulation has to be rerun, etc.). Also, much discussion could be expected on assumed price levels and on underlying parameters that are only indirectly related to adequacy issues. A second (simpler) approach would focus on quantities, e.g., adjusting the level of reliable capacity as long as the LOLE targets are not met. Reliable capacity could then be added (in case LOLE > target) or subtracted (LOLE < target) by ENTSO-E to reach the adequacy target in each country without considering the plants’ profitability. Overall, it amounts to providing adequacy levels as a function of reliable capacity available. The capacity/generation mix would probably be incorrect, but at least this approach would supply a reasonable perspective on the total reliable capacity required in the system. Each player would then be able to adapt the view on expected capacity margin based on the baseline reliable capacity assumed by ENTSO-E in the analysis and equilibrium levels. A side question is, however, in which country to adapt the reliable capacity above the baseline as there may exist multiple solutions to reach the same adequacy targets (import/export). Answers from ENTSO-E:


Nonetheless, it cannot be 100% guaranteed that the forecasted generation mix considered by TSOs in their simulations will be economically viable in 2020 and 2025. For 2030 and 2040, such appraisal is even more difficult because of the intrinsic uncertainties of such a long-term horizon. As part of the adequacy assessment in future MAF reports, ENTSO-E is debating whether to conduct a portion of sensitivity analysis with a more ‘conservative’ view of installed capacity. The sensitivities will attempt to assess the impact for adequacy of parts of the generation portfolio that



66

may be at risk of being mothballed for economic reasons, like, for example, in the absence of a capacity market and/or unfavourable market conditions – ‘missing money’ problem.

Note that the aforementioned sensitivities, through consideration of different scenarios that can be elaborated by using different assumptions on the capacity, still need to fulfil the requirements of consistent Pan-European-wide scenarios relevant to the Pan-European studies of the MAF. Economic viability appraisal is therefore linked to the construction of Pan-European consistent scenarios/sensitivities, which implicitly means that detailed national specificities could need to be considered by in-depth national and regional studies.

The approach put forth whereby there is the adjustment of the level of reliable capacity as long as the LOLE targets are not met (to be added if LOLE > target or subtracted if LOLE < target), should be considered with caution by ENTSO-E as providing perspectives on the economic viability for specific generation technologies should be conducted by market participants themselves and not by TSOs as regulated entities. Furthermore, such analyses implicitly assumes that the adequacy problems require a solution in terms of capacity, although there might exist multiple solutions to reach the same adequacy targets (import/export/capacity/generation and or demand flexibility).


The need for an evaluation of the economic viability of the assets originates from the fact that the installed capacity assumed in the system adequacy analysis might ultimately not be available as expected for economic reasons. Indeed, over a time horizon of 5 to 10 years, a lack of proper remuneration for any asset will trigger decisions (mothballing or decommissioning) by the market players so as to end their losses.

Performing such an analysis is evidently challenging given the multitude of assets and market players, and this should be properly acknowledged by all stakeholders. Actually, ENGIE maintains that a very precise analysis might not necessarily be required and that alternative options for deriving sensitivities should be investigated. In addition, recent major markets shifts (like extreme renewable expansion leading to prices collapsing well below long-term marginal costs in CWE, the gas price decrease putting pressure on lignite operators, or the ever decreasing price of carbon) are hard to anticipate. Therefore, there is no point in asking utilities to provide data on (firm) availabilities years in advance as changing economic conditions may make such promises unaffordable. Government interventions, in the form of TSO contracts or various taxes, compound the problem.




67

On the one hand, the analysis of the economic viability of the assets should help ENTSO-E in designing a credible sensitivity scenario on the installed capacity. A precise and detailed analysis might be too complex to perform when compared to the added value it would offer. As such, this analysis might be simplified and based on generic categories of assets and underlying costs (see Q 3.1). The assumptions of this analysis could, however, be subject to many discussions (e.g., all market players would have different viewpoints on expected revenues and costs for the [classes of] assets).

Yet, especially with the expected development of capacity mechanisms across Europe, an alternative option would be that ENTSO-E provides best estimates of the need for reliable/dispatchable capacity (MW). In other words, ENTSO-E could provide its view on the capacity necessary to service the demand minus the contribution from RES generation/capacity while respecting the adequacy targets. This is exactly the second approach sketched above in Q 3.1. If available, the market participants could then value the capacity needed via market-based capacity mechanisms and this approach would then provide an appropriate market signal for the investments or the decommissioning of assets (as exemplified by the capacity mechanism in UK).

Finally, ENGIE would like to emphasize that coherent demand forecasts are at least as important as relevant supply forecasts in a system adequacy exercise. ENTSO-E should make sure that the various demand forecasts employed in this exercise are based on similar assumptions. This is not obvious at all - “central demand” scenarios of various TSOs could be based on completely different electricity system outlooks. As a matter of fact, adequate sensitivities should be performed on the demand side. Similarly, it would be best if ENTSO-E tackled the issue along with uncertainty surrounding decentralized generation." Answers from ENTSO-E:

See answer above.


Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands



68

for DSM activation should be based on a forecasted expectations of future prices rather than current prices.


Today, adequacy levels are still a question with respect to national preferences - it depends on what a country considers a reasonable adequacy target. A higher adequacy can only be achieved with more installed reliable capacity, which comes at a certain cost. Overall, the system costs and benefits need to be brought in balance.

In an energy-only market design, ENGIE believes that adequacy cannot be guaranteed in practice. Part of the trade-off is theoretically through setting a price cap in the energy markets - the higher the combination of the price caps and the hours of capacity shortage (i.e., system inadequacy), the higher will be the remuneration above marginal costs for capacity and therefore the incentive for adequacy. However, peaking/marginal units should cover their fixed costs (and investment costs if a new unit should enter the market), mainly within scarcity periods (when demand is curtailed and prices reach the cap). Actually, the uncertainties surrounding market revenues and the ability to cover costs tend to discourage investors from keeping their merchant power plants online or – a fortiori - to invest in new capacity. This creates a risk for the security of electricity supply.

In a market-based capacity mechanism, the adequacy level is controlled by the authorities via the level(s) of capacity to be procured. However, accomplishing 100% reliability is likely to involve extremely high costs and be technically impossible, as highlighted by DG Competition.

***

Inany case, the linkbetweenadequacy levelsandeconomicviability isnotnecessarily straightforward.The first question iswhether a systemwith LOLE of less than one hour is adequate or not is open. AsrecalledfromQ2.1,somecountrieshavenoexplicittargetatallwhilelocalauthoritiesinothercountries(e.g., France) consider a LOLE less than three hours adequate, etc. Second, if the system is overlyadequate, a portion of the units should be loss-making and therefore exit the system (without anyconstraint).Thisprocesswouldthusreducethelevelofsystemadequacytoalevelwhereanequilibrium(includingnew investments) is reached. This dynamicprocess explains theneed for a proper economicassessmentand/orforaperspectiveonthereliablecapacitynecessaryinthesystemtoachieveacertain





69

levelofadequacy.Thoughthe installedcapacitycould indeeddependonthecoremarketdesign,otherriskfactors(e.g.,demandlevels)alsodrivethelevelofadequacyandapropersensitivityanalysisshouldbeperformed,aswell. Answers from ENTSO-E:

Role and Impact that the Results of the MAF Reports:

The MAF 2016 report features an ‘Overview Table’ on page 3 complemented by an appendix (Appendix 2) presenting the comments from each national TSO on the results in MAF 2016 in relation to each country’s own assessment of adequacy levels, its national adequacy standards and the measures taken to maintain them in the case of problems.The choice of LOLE = 1 hour as a threshold in such an overview table was chosen only because of technical reasons related to the presentation of the results. Only values of LOLE > 1 hour after averaging the results of the whole probabilistic Monte Carlo exercise were deemed significant enough to be highlighted, i.e.. the occurrence of adequacy problems across all the Monte Carlo ensemble of situations analysed was deemed significant rather than marginal.


WeknowthatsuchachoiceofthresholdforthepreviouslyreferredtoOverviewTablecouldbenonethelessconfusingandleadstomisinterpretation.ENTSO-EwillcarefullyconsidertheroleandimpactofthepresentationoftheresultsinfutureMAFreports.ENTSO-Eisawareoftheimpactthat the MAF results might have on discussions by member states on regionally coordinateddefinitionsofthosestandards. Total Need for Reliable/Dispatchable Capacity (MW):


ENTSO-E will cautiously account for the possibility of reporting the aforementioned Figure in the presentation of the results in future MAF reports. Providing views on specific generation technologies is not necessarily the role of TSOs as regulated entities. This indicator seems related to the deterministic indicator, remaining capacity margin (RCM), employed by ENTSO-E in SO&AF 2015 and previous reports. This indicator was not chosen in MAF 2016 in order to focus attention to the new probabilistic method deployed and LOLE and ENS indicators used. Reporting the RCM again in forthcoming MAF reports will be considered.


Irrespective of the market design, one should keep in mind that adequacy can only be ensured to a certain extent, even in the best case that includes a market-based capacity mechanism (see e.g., the




70

Interim Report - Sector Inquiry on Capacity Mechanisms by DG Competition, § 4.2.4). Within energy-only markets, price spikes (reflecting a.o. a shortage of supply compared to demand) are expected to cover the fixed (and investment) costs of the marginal units (like CCGT in the current market context), but it might not be sufficient to keep online existing capacity (which should then trigger decommissioning) or to attract new investments (which should then lower adequacy levels). With capacity mechanisms, the level of capacity necessary (and, therefore, remunerated) should ensure SoS only up to a given level of adequacy. In both cases, it is however expected that LOLE should be raised above the threshold retained by ENTSO-E in the current report (one hour).

***

From a regulatory perspective the current adequacy problems might be seen as the consequences of the (patchwork of) market designs and conditions, but also of the system interface between generation and network. The questionnaire gives the impression that the TSOs are responsible for managing this interface. Going forward, and to avoid any conflict of interest between market players and regulated entities, the setup of a European Independent System Operator (ISO) could be considered for managing such interfaces.

***

Modelling requires good data - we are not sure that the current implementations of regulation 543/2013 with regards to hydro reservoir contents reporting assist in this context as the numbers published do not reflect energy that can be produced any longer (reservoirs energy is now reported based on local potential production when going through the local turbine as opposed to the old UCTE assessment of energy available through all dams down to the sea; this means that the water equivalent to 1 GWh in one dam could be materialized into 1.5 GWh to 7 GWh of hydro production when released depending on location).

Answer by ENTSO-E:

Regarding the system interface between generation and network, ENTSO-E and the TSOs agree they are not solely responsible for managing this interface – indeed, TSOs are not responsible for safeguarding the economic viability of specific generation technologies. This clearly should be performed by market participants themselves and not by TSOs as regulated entities. Yet, ENTSO-E is mandated to provide a view on the adequacy risks and to manage the security of the grid and electricity supply. These two elements are very much interlinked in the power system and should be considered together. Still, TSOs do not assume and should therefore not be assumed i) it is their sole responsibility to solve these problems by managing the generation and network interface and ii) that adequacy problems require a solution in terms of capacity exclusively as there may exist multiple solutions to reach the same adequacy targets (import/export/capacity generation and or demand flexibility).

Hydro: Hydro modelling details can have a significant influence on adequacy indicators, especially of regions, including countries, with large shares of hydro production. Dry and wet hydro conditions (inflow and reservoir) have been considered in MAF 2016. In addition, the ENTSO-E PECD is currently being extended to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the probability of occurrence of these dry-normal and wet conditions is being investigated



71

systematically in order to define so-called hydrological regions for which the same probability of similar conditions can be accounted for in Monte Carlo probabilistic simulations.

National specificities and geographical/location sensitivities should be well captured by the bottom-up data collection process in which TSOs provide data to ENTSO-E.



72

COMMENTS BY STAKEHOLDERS Core contributions of having a common adequacy analysis at the EU level includes:

1) Highlighting interdependencies of the European electricity markets; 2) Taking into account the regional influences in SoS estimations; 3) Making cross-border comparisons; and 4) Allowing for learning across borders.

To underpin these objectives, it is important to ensure harmonisation of inputs across member states and market areas. To this end, the format of inputs should be as standardised as possible. There should also be a standardisation of how data is treated by the model, as well as standardisation of assumptions. Deviations from the standardised approach may be relevant to take into account location conditions, but it should be well-analysed and well-argued in the report.

We appreciate the annex where each TSO could comment on the MAF as it increases transparency surrounding how results are modelled and perceived differently across the different TSOs and where TSOs agree to compromises in order to complete the European modelling.








NORDENERGI



73


Hence, Pan-European, regional and national studies are an absolute necessity as they complement each other rather than compete - studies covering a larger perimeter (Pan-European compared to regional or regional/local) are limited by potentially contradicting modelling goals and challenging data collection, but provide a consistent appraisal, if not of highest accuracy, of a series of commonly and continuously assessed indicators at the Pan-European level. Conversely, more local studies need the consistent boundary conditions supplied by the larger perimeter studies to address more precisely certain specific issues. Regional and national studies shall use the common definition and methodology standards compiled by ENTSO-E to enable comparison. For a given country or region, the prevailing conclusions and decision making regarding market design adaptations, to ensure the desired SoS level, shall stem from the study implementing the most advanced modelling for the investigated parameter. ENTSO-E adequacy studies will therefore be needed as a pre-requisite for more local regional and national studies, but they will be the reference to decide whether to implement risk mitigation measures only where no more detailed study is performed. TSO/ENTSO-E processes and bottom-up common methodology along with the use of several TSO tools across the EU perimeter in MAF makes certain there is consistency and articulation between Pan-European (MAF), regional (e.g., PLEF) and national studies in this respect.


Nordenergi supports that ENTSO-E, for the first time in this version of the MAF, has now parted with traditional capacity balances. There are good reasons for this, as ENTSO-E justifies in the MAF. Irrespective of the approach, a robust analysis of SoS should be as transparent as possible. Nordenergi has found that the MAF 2016 is a start, but it does leave quite a bit to be desired in terms of transparency on a number of accounts:

• The presentation of the results appears to be lacking. In MAF 2016, the findings for the capacity adequacy situation are compared to a LOLE of one hour/year, like a binary indicator. It would be highly appreciated if concrete numbers for LOLE and ENS are reported for each area.

• It is difficult to understand the inner workings of the algorithm. Nordenergi observed that a greater level of transparency is required on this matter. This is a clear loss of transparency over the capacity balance. To improve, there must be transparency for all data (insofar that it does not violate the privacy of market participants), assumptions and methods.




74

• Nordenergi would also highlight the need for publication of the algorithm so it can be used by market participants to formulate their private estimations. This would permit better understanding of the SoS situation and serve as a third-party evaluation of the model utilised.

• No forecast is ever going to be 100% correct. For this reason, sensitivity analysis is a key component of any SoS analysis. Nordenergi found that the sensitivities used in MAF 2016 are quite difficult to understand. It is not entirely apparent which elements were investigated and how the different sensitivities relate to one another. Are the four different analytical tools to model regional specificities also meant to reflect sensitivities? Looking ahead to the next MAF, we would highly encourage ENTSO-E to improve on this point.

• Attempts should be made to include demand response in future MAF exercises. It is already visible in the Nordic countries of which demand reacts to high day-ahead prices and to implied scarcity. If data is difficult to obtain, it could be included in the sensitivities or in the cases of countries where better data is available as a regional specificity.

Nordenergi believes that sensitivity analysis should (at the very least) contain the consequences of alternative assumptions on:

• speed of closure of power stations; • roll-out speed of DSR and storage; • availability of interconnectors; • timing of newly built transmission capacity (in case of delays); • weather (dry years in the Nordics, heat waves); • variation in consumption; • cross-border effects of having a capacity market in one region on its neighbouring regions.

We understand that ENTSO-E is aware of this issue already; and • due considerations to simultaneous scarcity situations across borders.

Answers from ENTSO-E



Although these studies will not encompass every potential issue of future power system, ENTSO-E strongly believes that they:





75


ENTSO-E is also aware that the choice of the mathematical approach could significantly affect the indicators that may be assessed through the simulation as well as the structure and complexity of the input data and modelling assumptions used. ENTSO-E is additionally cognizant of the impact that designing an optimal and economically feasible set of investments plans in generation portfolio has on adequacy levels. We refer to the answers below. Regarding Planned Methodological Improvements

Demand: ENTSO-E agrees with stakeholders that demand forecasting (levels, but also profiles) as one of the main items affecting the future adequacy outlook. Therefore, ENTSO-E is working on improving modelling and the reliability of demand forecasting through enhancing methodologies. In order to ensure consistency and transparency, ENTSO-E has set up a process to define, in a centralized manner, demand forecast levels and profiles. TSOs are requested to provide national figures for the main parameters to model demand: temperature sensitivity of load, electric vehicle (EV) penetration, heat pumps (HPs), energy efficiency (EE), DSM/demand-side response (DSR), expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. In this way, consistent regional/Pan-European demand figures and profiles are defined that are also in line with national expectations and capture national specificities properly. Hydro: Hydro modelling details can have a significant influence on adequacy indicators, especially of regions, including countries, with large shares of hydro production. Dry and wet hydro conditions (inflow and reservoir) have been considered in MAF 2016. In addition, the ENTSO-E PECD is currently being extended to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the probability of occurrence of these dry-normal and wet conditions is being investigated systematically in order to define so-called hydrological regions for which the same probability of similar conditions can be accounted for in Monte Carlo probabilistic simulations. Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices. New Methodological Improvements:




76

Visibility of Data: ENTSO-E strives toward a high level of transparency. The so-called 'MAF2016_market_modelling_data' package, including all input net generating capacity (NGC), demand and NTC figures, has been made available upon publication of the report. ENTSO-E strives to enhance transparency regarding other key parameters used for the simulations. Transparency also implies responsibility not only of the publisher, but also of the recipient. Improvement of these key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the different generation technologies available. Such exchange could also make better the available data by TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as it concurs that the results should always be understood within the context of the data and assumptions used. Transparency of Algorithms:



The MAF 2016 report features an ‘Overview Table’ on page 3 complemented by an appendix (Appendix 2) presenting the comments from each national TSO on the results in MAF 2016 in relation to each country’s own assessment of adequacy levels, its national adequacy standards and the measures taken to maintain them in the case of problems.The choice of LOLE = 1 hour as a threshold in such an overview table was chosen only because of technical reasons related to the presentation of the results. Only values of LOLE > 1 hour after averaging the results of the whole probabilistic Monte Carlo exercise were deemed significant enough to be highlighted, i.e.. the occurrence of adequacy problems across all the Monte Carlo ensemble of situations analysed was deemed significant rather than marginal.


We know that such a choice of threshold for the previously referred to Overview Table could be nonetheless confusing and leads to misinterpretation. ENTSO-E will carefully consider the role and impact of the presentation of the results in future MAF reports. ENTSO-E is aware of the impact that the MAF results might have on discussions by member states on regionally coordinated definitions of those standards. Total Need for Reliable/Dispatchable Capacity (MW):

ENTSO-E will carefully consider the possibility of reporting the aforementioned Figure in the presentation of the results in future MAF reports. However, providing a view on specific generation technologies is not necessarily the role of TSOs as regulated entities.



77

Capacity Surplus/Deficit (MW):



Thermal Outage Correlation with Climatic Conditions: Extensions of the methodology might beconsidered to capture correlations of thermal outages, thermal production and climateconditions,notablytemperatureconditions.Itisworthnotingthatthiseffectisverypronouncedin certain countries, e.g., PL, but might not require a Pan-European deployment, and insteadrequirelocalsensitivitieswithrespecttotheMAFresults. Sensitivity of Scenarios:

ENTSO-E must fulfil obligations of Reg. 714/2009. The European generation adequacy outlook shall build upon national generation adequacy outlooks prepared by each individual transmission system operator.The first goal of ENTSO-E is to collect and set up a contiguous (central) best estimate ‘Pan-European’ scenario based on data collected from TSOs as meaningful sensitivities are defined around this. ENTSO-E is therefore revising its data collection and scenario-building process in order to consider relevant sensitivities based on data provided by TSOs on, e.g., units at risk of being mothballed for economic reasons, information regarding decommissioning of units for purely technical reasons and/or shut downs for legal/policy reasons within the storyline assumptions of the scenarios considered. Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios. Flow-Based Methods



78




NTC – Conservative. Conservative assumptions have been made regarding cross-border transmission capacity relating to uncertainty in the commissioning dates of cross-border transmission projects and adequacy risks therein. These assumptions are within the scope of the methodology and report capturing all key features and risks relevant at the Pan-European level - this is a main objective for ENTSO-E. A conservative approach of ENTSO-E in providing transfer capacity values for adequacy analyses is sensible, even if adequacy risk might be overestimated. This will mean that the energy mix considered will not be compatible with the assumed/available transmission infrastructure. In this respect, the MAF results provide a signal in the form of ‘adequacy risk’. This could very well be higher than the desired adequacy level that a country is willing to accept or prepare for, expressed in the form of its adequacy standard. As a consequence of the fact that investments in grid infrastructure require a longer planning and decision horizon (~15 years) than the typical investment cycles for investments in new generation (3-5-7) years, the MAF exercise will detect risks and provide important feedback regarding the establishment of countermeasures by relevant stakeholders (e.g., member state authorities, policy makers, regulatory agencies, energy producers) in order to ensure the desired adequacy levels and the risks linked to their choices of generation mix and its evolution as well the impact of cross-border capacity availability. Furthermore, national or regional assessments might consider different assumptions that could be more optimistic regarding availability of cross-border transmission infrastructure and availability of imports to complete the assessment of the impact of such assumptions on adequacy levels. 15-Minute Flexibility: The purpose of the ENTSO-E target adequacy methodology being developed is the generation of a sound methodology that will foster performing regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Considering any modelling implies simplifications, so ENTSO-E adequacy target modelling focuses on hourly power balance modelling. Although these studies do not encompass every potential issue involved in the state of the future power system, ENTSO-E strongly believes that



79

this methodology will permit the development of a model able to capture all the key features and risks regarding adequacy for the Pan-European power system - this is a main objective for ENTSO-E. The effect of imbalances based on forced outages occurring close to real-time and the effect of forecasting errors of load, wind and solar production is accounted for in MAF by consideration of operational reserve requirements as an extra constraint while modelling. The detailed modelling of these aspects is outside of the scope of MAF as these imbalances are typically covered by the intra-day and primarily balancing markets and are not observed systematically in the day-ahead market during several days of a week. Flexibility: Residual load analysis, as considered in SO&AF2015, will be appraised for future MAF reports in order to inform them with respect to the need for flexibility necessary in power systems, typically within intra-day and balancing markets. One of the main goals is to be able to integrate 'need for flexibility' into the system. Flexibility in adequacy assessments is based on weather-dependent effects related to load variation, generation patterns of wind and solar power plants with a one-hour resolution and the consideration of the resources for flexibility. The existing PECD will be used for the adequacy assessments. Sensitivity of Scenarios:

ENTSO-E must fulfil obligations of Reg. 714/2009. The European generation adequacy outlook shall build upon national generation adequacy outlooks prepared by each individual transmission system operator.The first goal of ENTSO-E is to collect and set up a contiguous (central) best estimate ‘Pan-European’ scenario based on data collected from TSOs as meaningful sensitivities are defined around this.

Data Collection & Scenarios: ENTSO-E is therefore revising its data collection and scenario-building process in order to consider relevant sensitivities based on data provided by TSOs on, e.g., units at risk of being mothballed for economic reasons, information regarding decommissioning of units for purely technical reasons and/or shut downs for legal/policy reasons within the storyline assumptions of the scenarios considered. Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios. ENTSO-E is improving its data collection process regarding input data which affects the ‘likelihood of units to run and stay online’ within the market modelling assessments performed in TYNDP - MAF - these input data items are crucial to carrying any sensible sensitivity determination pertaining to ‘viability’ of the (central) best estimate scenarios collected from TSOs.

ENTSO-E and the TSOs know of the importance of these assumptions regarding the definition of the scenarios and ENTSO-E’s process for a common and consistent data collection has been revised to elevate the quality of the scenarios and possibly evaluate sensitivities surrounding them.



80


Note that the aforementioned sensitivities, through consideration of different scenarios that can be elaborated by using different assumptions on the capacity, still need to fulfil the requirements of consistent Pan-European-wide scenarios relevant to the Pan-European studies of the MAF. Economic viability appraisal is therefore linked to the construction of Pan-European consistent scenarios/sensitivities, which implicitly means that detailed national specificities could need to be considered by in-depth national and regional studies.

The approach put forth whereby there is the adjustment of the level of reliable capacity as long as the LOLE targets are not met (to be added if LOLE > target or subtracted if LOLE < target), should be considered with caution by ENTSO-E as providing perspectives on the economic viability for specific generation technologies should be conducted by market participants themselves and not by TSOs as regulated entities. Furthermore, such analyses implicitly assumes that the adequacy problems require a solution in terms of capacity, although there might exist multiple solutions to reach the same adequacy targets (import/export/capacity/generation and or demand flexibility). Role and Impact that the Results of the MAF Reports:



We know that such a choice of threshold for the previously referred to overview table could be nonetheless confusing and leads to misinterpretation. ENTSO-E will carefully consider the role and impact of the presentation of the results in future MAF reports. ENTSO-E is aware of the impact that the MAF results might have on discussions by member states on regionally coordinated definitions of those standards. Simultaneous scarcity: ENTSO-E will cautiously consider the possibility of post-processing its results to detect situations of simultaneous scarcity on future reports. These results could act as important input for discussions by member states coordinated management of these simultaneous scarcity situations.



81

Considering the above, what additional methodological improvements should ENTSO-E consider for future MAFs? Please justify why through suggestions and what these improvements will offer. COMMENTS BY STAKEHOLDERS

See justification above.


Nordenergi recognizes that it is nearly impossible to feature a full overview of future availability and the decommissioning/mothballing of plants for 5 to 10 years into the future - even the industry cannot predict this. We understand that this needs to be based on close dialogue between the TSOs and their NRAs and stakeholders so that a "best guess" can be implemented into the model. This method necessitates transparency in terms of the assumptions and underscores the importance of thorough sensitivity analysis to reflect uncertainty. Answers from ENTSO-E:

Regarding the issue of whether adequacy assessments should serve as “stress tests” rather than “forecasts”, ENTSO-E is improving its data collection processes as the input data are crucial to perform any sensible adequacy assessments. The first goal of ENTSO-E is to collect and set up a consistent (central) best-estimate Pan-European scenarios based on data collected from TSOs. Meaningful sensitivities are indeed defined around this central best-estimate Pan-European scenario as ‘stress test’ cases in MAF. It should be noted that the sensitivities relevant for ENTSO-E will always be relevant to and of the scope of Pan-European scenarios. ENTSO-E welcomes the stakeholder feedback received - it supplies an important input for revising the elaboration of meaningful sensitivities. Articulation of Pan-European, regional and national studies is hence a must as these complement, rather than compete, with each other. National studies must be consistent with the boundary conditions and assumptions provided by the larger perimeter Pan-European and regional studies and have to employ the common definitions and methodology standards defined and used by ENTSO-E.




82


Nordenergi suggests a slightly different technique for the proposed economic assessment. The adequacy forecast should concentrate on defining the level of capacity to ensure SoS at maximum load while taking into account available capacity and expected possibilities for demand response in the mid- to long-term. Given a predefined reliability standard, the TSOs should conduct an assessment of how much capacity should be present in the system (including cross-border capacity availability), considering the differences between short-term (hours) and longer term (days to weeks) deficit situations to safeguard SoS in the mid- to long-term. Answers from ENTSO-E: Total Need for Reliable/Dispatchable Capacity (MW):


ENTSO-E will cautiously account for the possibility of reporting the aforementioned figure in the presentation of the results in future MAF reports. Providing views on specific generation technologies is not necessarily the role of TSOs as regulated entities. This indicator seems related to the deterministic indicator, remaining capacity margin (RCM), employed by ENTSO-E in SO&AF 2015 and previous reports. This indicator was not chosen in MAF 2016 in order to focus attention to the new probabilistic method deployed and LOLE and ENS indicators used.




i) One scenario linked to the current regulatory framework based on the energy-only market; and

ii) A second scenario connected to the future regulatory framework (for example reflecting expectations of a national implementation of capacity mechanisms or any other market design instruments).



83


Nordenergi understands that there is a need to evaluate the cross-border effects of capacity markets in one region and its neighbouring regions, and the proposed solution could potentially satisfy this requirement. The importance lies in clearly stating, in detail, which assumptions are made in both scenarios. While it is relatively easy to model an energy-only market, assumptions on capacity mechanisms are more difficult as there are different views on future capacity adequacy targets and level of support, while existing capacity mechanisms change and may evolve. Answers from ENTSO-E:

ENTSO-E is improving its data collection process regarding input data which affects the ‘likelihood of units to run and stay online’ within the market modelling assessments performed in TYNDP - MAF - these input data items are crucial to carrying any sensible sensitivity determination pertaining to ‘viability’ of the (central) best estimate scenarios collected from TSOs.

ENTSO-E and the TSOs know of the importance of these assumptions regarding the definition of the scenarios and ENTSO-E’s process for a common and consistent data collection has been revised to elevate the quality of the scenarios and possibly evaluate sensitivities surrounding them. Nonetheless, it cannot be 100% guaranteed that the forecasted generation mix considered by TSOs in their simulations will be economically viable in 2020 and 2025. For 2030 and 2040, such appraisal is even more difficult because of the intrinsic uncertainties of such a long-term horizon. As part of the adequacy assessment in future MAF reports, ENTSO-E is debating whether to conduct a portion of sensitivity analysis with a more ‘conservative’ view of installed capacity. The sensitivities will attempt to assess the impact for adequacy of parts of the generation portfolio that may be at risk of being mothballed for economic reasons, like, for example, in the absence of a capacity market and/or unfavourable market conditions – ‘missing money’ problem.

Note that the aforementioned sensitivities, through consideration of different scenarios that can be elaborated by using different assumptions on the capacity, still need to fulfil the requirements of consistent Pan-European-wide scenarios relevant to the Pan-European studies of the MAF. Economic viability appraisal is therefore linked to the construction of Pan-European consistent scenarios/sensitivities, which implicitly means that not all nationally relevant sensitivities can be captured within consistent Pan-EU scenarios. Detailed national specificities might need to be considered in depth by additional national and regional studies. The approach put forth whereby there is the adjustment of the level of reliable capacity as long as the LOLE targets are not met (to be added if LOLE > target or subtracted if LOLE < target), should be considered with caution by ENTSO-E as providing perspectives on the economic viability for specific generation technologies should be conducted by market participants themselves and not by TSOs as regulated entities. Furthermore, such analyses implicitly assumes that the adequacy problems require a solution in terms of capacity, although there might exist multiple solutions to reach the same adequacy targets (import/export/capacity/generation and or demand flexibility).



84


The theoretical calculation of SoS should be accompanied by a set of ‘resiliency indicators’, whereby system operators’ use of remedial actions (a number of which are defined in the Guideline on System Operation) are tracked. The purpose of this is to aid in comprehending the essential level of stress in the electrical system that results in TSOs having to take non-market based actions to ensure SoS. Possible indicators include:

• Rejection of market participants’ applications to close down or mothball power stations; • Brown outs; • Announcements of brown outs; • Number and duration of controlled voltage reduction; • Use of energy reserves to cover foreseen generation deficits (not what reserves are for); • Occurrences where countries change cross-border capacity to cover own stress situations in the grid; • Occurrences of defined stress situations in capacity markets; • Activation of strategic reserves; • Instances where market participants are forced by system operators to take certain actions outside markets for SoS reasons, such as forced operation; • Number of hours without clearing price in the spot markets; • Voltage quality; • Changes to market participant outage planning; and • Changes in availability of interconnectors because of SoS concerns.

Denmark seems to be treated as a single area, but is in fact two separate areas. Between them, there are significant differences in Energinet.dk’s estimated SoS. Given this difference, we would recommend that these areas are treated as two separate areas in the MAF.

Russian imports and exports require more clarity. How are they considered and taken into account in the analysis? It appears that flows from third countries are based on historical values. For the Russian imports, this, for example, may produce biased estimates, and does not tell what the imports could be in a scarcity situation with high prices in the Nordic region. Nordenergi proposes that certain price dynamics be applied.

Demand responses seem to be rather neglected in the analysis. However, it is evident that when power prices are high, the demand does react, as, for example, was experienced in 2010 when Finnish strategic reserve last were activated. Concerning demand flexibility, we also want to point out Statnett's comment regarding the MAF report. They do not necessarily believe that a customer not receiving electricity automatically indicates a system failure. If a customer is not willing to pay a high price, choosing to reduce consumption instead, it is a sign of contradictory system success. Demand flexibility should be considered in the next MAF, even if it is just a matter of sensitivity.




85

Nordenergi also calls for updating the information on existing capacities. It appears that, for example, for Finland, capacities already de-commissioned are deemed mothballed in the report.


MAF methodology might indeed not capture certain extreme, nationally specific and relevant situations. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios.

MAF methodology does not aim to supply full insights into possible overloads and voltage problems that could take place as a consequence of tight adequacy situations inside a country. These situations have typically a rather local nature and are outside the scope of a Pan-European appraisal of adequacy within the current Pan-European MAF report.

TSOs supported by the ENTSO-E framework and regional security coordinators (RSC) are committed to facilitation of the deployment of potential gradual measures to maintain supply in all cases of local/national, regional or Pan-European impact. Examples of measures that TSOs utilise are:

i) Internal and cross-border structural grid reinforcement; ii) Local optimisation of the grid;

iii) Extra generation reserves; and iv) Demand Response. Simultaneous scarcity: ENTSO-E will cautiously consider the possibility of post-processing its results to detect situations of simultaneous scarcity on future reports.These results could have a significant function regarding the establishment of a framework to express national SoS targets on the basis of commonly agreed adequacy standards and discussions by member states on regionally coordinated definitions of those standards as well as on the management of these simultaneous scarcity situations.



86

COMMENT BY STAKEHOLDER To ensure comparability, baseline case assumptions of demand and installed capacity should be the same on a national, regional and European level. Further, a harmonised methodology should be applied to ensure consistency. An adequacy assessment encompassing a larger region is more complete because it takes interconnection and interdependency between countries into account. In our view, this is the most profound benefit of the MAF reports. Adequacy studies at the regional and national levels, on the other hand, still have an important role to play. In those studies, it is possible to conduct sensitivity analyses on the inputs as the scope is more limited and there is greater amounts of detailed information available. From our perspective, the MAF should be restrained from sensitivities as long as the input data are not fully harmonized. As the data from different countries is gathered and processed in varied ways, creating scenarios would not be a good idea at this moment seeing that it would reduce objectivity and hence credibility. Robust generation of adequacy assessments is a prerequisite for any public intervention in this area. Defining national SoS levels, as desired by politicians and the public, is a key element of the assessment of generation adequacy because it allows gauging actual reliability levels. The MAF 2016 report created a skewed image of the adequacy situation as a LOLE of one was chosen as a threshold. Better choices for a threshold to flag potential scarcity would be either the nationally determined LOLE target, a European average or, at the very least, a frequently used number. Member states should be encouraged to set these in terms of loss of load. Furthermore, another colour should be introduced to highlight ‘oversecurity’, which signals a power system that is potentially too expensive.



ENTSO-E appreciates the positive feedback from stakeholders regarding the ENTSO-E methodology as a common basis. ENTSO-E views this role as follows: a) ENTSO-E describes and proposes common definitions for the multiple concepts/aspects under Security of Supply (SoS) and adequacy, providing all stakeholders with a common language;


VATTENFALL



87




Hence, Pan-European, regional and national studies are an absolute necessity as they complement each other rather than compete - studies covering a larger perimeter (Pan-European compared to regional or regional/local) are limited by potentially contradicting modelling goals and challenging data collection, but provide a consistent appraisal, if not of highest accuracy, of a series of commonly and continuously assessed indicators at the Pan-European level. Conversely, more local studies need the consistent boundary conditions supplied by the larger perimeter studies to address more precisely certain specific issues. Regional and national studies shall use the common definition and methodology standards compiled by ENTSO-E to enable comparison. For a given country or region, the prevailing conclusions and decision making regarding market design adaptations, to ensure the desired SoS level, shall stem from the study implementing the most advanced modelling for the investigated parameter. ENTSO-E adequacy studies will therefore be needed as a pre-requisite for more local regional and national studies, but they will be the reference to decide whether to implement risk mitigation measures only where no more detailed study is performed. TSO/ENTSO-E processes and bottom-up common methodology along with the use of several TSO tools across the EU perimeter in MAF makes certain there is consistency and articulation between Pan-European (MAF), regional (e.g., PLEF) and national studies in this respect. ENTSO-E is also aware that the choice of mathematical approach could significantly affect the indicators that can be assessed through the simulation as well as the structure and complexity of the input data and modelling assumptions used. Therefore, the MAF report’s results have been benchmarked with four different calculation software tools.


Publication of the MAFs should include the actual national set of reliability standards (where available) to permit comparison of the MAF results with these. In the current version, LOLE values are provided out of context and in colour schemes that risk the reader drawing the wrong conclusions.




88

Regarding the simulations, Vattenfall would welcome more explanations of the methodology. From the illustration on page 9, it seems that load, RES, hydro, thermal and cross-border capacity are all considered to be independent factors, while, for instance, high temperature in the summer could lead to high demand in the summer as well as to low availability of thermal plants (following a lack of cooling water).




The purpose of the ENTSO-E adequacy methodology and MAF reports under evolution is the development of a robust methodology which will allow conducting regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Although these studies will not encompass every potential issue of future power system, ENTSO-E strongly believes that they: i) Will permit the development of a model able to capture all key features and risks regarding adequacy for the Pan-European power system. This is a main objective for ENTSO-E; ii) Will act as a common basis for methodologies and definitions; and



Demand: ENTSO-E agrees with stakeholders that demand forecasting (levels, but also profiles) as one of the main items affecting the future adequacy outlook. Therefore, ENTSO-E is working on improving modelling and the reliability of demand forecasting through enhancing methodologies. In order to ensure consistency and transparency, ENTSO-E has set up a process to define, in a centralized manner, demand forecast levels and profiles. TSOs are requested to provide national figures for the main parameters to model demand: temperature sensitivity of load, electric vehicle (EV) penetration, heat pumps (HPs), energy efficiency (EE), DSM/demand-side response (DSR), expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. In this way, consistent regional/Pan-European demand figures and profiles are defined that are also in line with national expectations and capture national specificities properly. Hydro: Hydro modelling details can have a significant influence on adequacy indicators, especially of regions, including countries, with large shares of hydro production. Dry and wet hydro conditions (inflow and reservoir) have been considered in MAF 2016. In addition, the ENTSO-E PECD is currently being extended to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the



89

probability of occurrence of these dry-normal and wet conditions is being investigated systematically in order to define so-called hydrological regions for which the same probability of similar conditions can be accounted for in Monte Carlo probabilistic simulations. Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices. New Methodological Improvements:

Economic Viability: Sensitivity analysis that will take a more ‘conservative’ view of installed capacity might be needed in future MAFs. To enable this, ENTSO-E seeks to include in the data collection process for future MAF reports estimates of the volume of capacity that could be at risk of being mothballed for economic reasons, like, for example, in the absence of a capacity market and/or unfavourable market conditions – ‘missing money’ problem. These data will be analysed and sensitivities will be defined for the simulations. Visibility of Data: ENTSO-E strives toward a high level of transparency. The so-called 'MAF2016_market_modelling_data' package, including all input NGC, demand and NTC figures, has been made available upon publication of the report. ENTSO-E strives to enhance transparency regarding other key parameters used for the simulations. Transparency also implies responsibility not only of the publisher, but also of the recipient. Improvement of these key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the different generation technologies available. Such exchange could also make better the available data by TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as it concurs that the results should always be understood within the context of the data and assumptions used. Transparency of Algorithms:

Different market modelling tools available, owned or procured by TSOs are utilised during the MAF exercise. These tools are proven to capture nationally and regionally relevant specificities. MAF work relies on this know-how from TSOs in terms of the strong bottom-up basis of all knowledge and detailed data. Tool dependency and reliability is a major focus for MAF. Use of exclusively open-source tools might be beneficial from a purely ‘academic’ point of view, though on the other hand, does not guarantee a strong link to bottom-up national details – know-how of TSOs and national studies. Full transparency is not always possible because these tools are, in certain cases, proprietary materials from different tool providers in a similar way that generation and market data is confidential to market parties as communicated via the stakeholder feedback and answers throughout this document. Total Need for Reliable/Dispatchable Capacity (MW):

ENTSO-E will carefully consider the possibility of reporting the aforementioned Figure in the presentation of the results in future MAF reports. However, providing a view on specific generation technologies is not necessarily the role of TSOs as regulated entities.



90

Capacity Surplus/Deficit (MW):




ENTSO-E must fulfil obligations of Reg. 714/2009. The European generation adequacy outlook shall build upon national generation adequacy outlooks prepared by each individual transmission system operator.The first goal of ENTSO-E is to collect and set up a contiguous (central) best estimate ‘Pan-European’ scenario based on data collected from TSOs as meaningful sensitivities are defined around this. ENTSO-E is therefore revising its data collection and scenario-building process in order to consider relevant sensitivities based on data provided by TSOs on, e.g., units at risk of being mothballed for economic reasons, information regarding decommissioning of units for purely technical reasons and/or shut downs for legal/policy reasons within the storyline assumptions of the scenarios considered.

Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios. Flow-Based Methods



91




NTC – Conservative. Conservative assumptions have been made regarding cross-border transmission capacity relating to uncertainty in the commissioning dates of cross-border transmission projects and adequacy risks therein. These assumptions are within the scope of the methodology and report capturing all key features and risks relevant at the Pan-European level - this is a main objective for ENTSO-E.

Aconservative approach of ENTSO-E in providing transfer capacity values for adequacy analyses is sensible, even if adequacy risk might be overestimated. This will mean that the energy mix considered will not be compatible with the assumed/available transmission infrastructure. In this respect, the MAF results provide a signal in the form of ‘adequacy risk’. This could very well be higher than the desired adequacy level that a country is willing to accept or prepare for, expressed in the form of its adequacy standard. As a consequence of the fact that investments in grid infrastructure require a longer planning and decision horizon (~15 years) than the typical investment cycles for investments in new generation (3-5-7) years, the MAF exercise will detect risks and provide important feedback regarding the establishment of countermeasures by relevant stakeholders (e.g., member state authorities, policy makers, regulatory agencies, energy producers) in order to ensure the desired adequacy levels and the risks linked to their choices of generation mix and its evolution as well the impact of cross-border capacity availability. Furthermore, national or regional assessments might consider different assumptions that could be more optimistic regarding availability of cross-border transmission infrastructure and availability of imports to complete the assessment of the impact of such assumptions on adequacy levels. 15-Minute Flexibility: The purpose of the ENTSO-E target adequacy methodology being developed is the generation of a sound methodology that will foster performing regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Considering any modelling implies simplifications, so ENTSO-E adequacy target modelling focuses on hourly power balance modelling. Although these studies do not encompass every potential issue involved in the state of the future power system, ENTSO-E strongly believes that



92

this methodology will permit the development of a model able to capture all the key features and risks regarding adequacy for the Pan-European power system - this is a main objective for ENTSO-E. The effect of imbalances based on forced outages occurring close to real-time and the effect of forecasting errors of load, wind and solar production is accounted for in MAF by consideration of operational reserve requirements as an extra constraint while modelling. The detailed modelling of these aspects is outside of the scope of MAF as these imbalances are typically covered by the intra-day and primarily balancing markets and are not observed systematically in the day-ahead market during several days of a week. Flexibility: Residual load analysis, as considered in SO&AF2015, will be appraised for future MAF reports in order to inform them with respect to the need for flexibility necessary in power systems, typically within intra-day and balancing markets. One of the main goals is to be able to integrate 'need for flexibility' into the system. Flexibility in adequacy assessments is based on weather-dependent effects related to load variation, generation patterns of wind and solar power plants with a one-hour resolution and the consideration of the resources for flexibility. The existing PECD will be used for the adequacy assessments.





We know that such a choice of threshold for the previously referred to Overview Table could be nonetheless confusing and leads to misinterpretation. ENTSO-E will carefully consider the role and impact of the presentation of the results in future MAF reports. ENTSO-E is aware of the impact that the MAF results might have on discussions by member states on regionally coordinated definitions of those standards.

ENTSO-E MAF adequacy provides an important discussion point regarding SoS. However, the decision to implement measures to ensure SoS at the national level is directly connected to the responsibility for SoS. Member states, as entities responsible for SoS, should make the decision to implement or not measures to ensure SoS while ensuring this is performed in accordance with neighbouring member states and TSOs based on the cross-border impact that such decisions might have.



93



No Comment.

COMMENTS BY STAKEHOLDERS In general, this is a commercially sensitive situation that is therefore not easily shared by market participants. Determining whether to decommission/mothball a production unit involves quantitative analysis (i.e., at the portfolio level) combined with strategic considerations, hence it is impossible for TSOs to have full visibility into future installed capacity. In addition, the situation can change rapidly, so we simply need to accept that all input is subject to change. This is inherent to forecasting and therefore not a problematic issue as long as this is clearly explained. Generators already disclose information regarding their production units as inputs for national adequacy assessments. Thus, we are hesitant to support “further rules at the EU level”, as ENTSO-E can just make use of the information from TSOs. When the EU can initiate harmonisation of the reporting obligation, this will be fine. Close cooperation with national supervisory authorities is advised, and standardised questionnaires could be considered. Answers from ENTSO-E:

Economic/technical data on generation technologies is needed to perform market simulations.

The hypotheses are gathered by the TSOs according to their best knowledge and the availability of data at the time of data collection. ENTSO-E and the TSOs involved in these studies can therefore not be held responsible in the case the hypotheses from these reports or the estimations based on these hypotheses are not realised in the future.

ENTSO-E welcomes greater transparency and availability from market participants with respect to the economic/technical data to improve the accuracy of the forecasts on economic/technical data used by TSOs in the simulations performed. It is very important to have superior visibility into decommissioning and mothballing figures in order to set up the relevant generation scenarios. The MAF data collection process will be revised for MAF 2017 in order to improve visibility of these data provided by TSOs and processed to set up relevant sensitivities that will explore possibilities to consider different views on installed capacity, from ‘optimistic’ to ‘conservative’ situations. The sensitivities planned are in line with the feedback provided by the stakeholder.




94


It will be very hard to determine the economic viability of a certain generation portfolio within this study because it requires a totally different approach, and the result would be very dependent on a wide variety of assumptions and the models used.

We would argue that a TSO should restrain itself from assessing economic viability as this will damage the objectivity and credibility of the study. To warn the readers that outcomes are very sensitive to (small) changes of input parameters, ENTSO-E could decide to qualitatively describe this. Conducting sensitivities, like a higher/lower demand rather than the best estimate, are more suitable for a national- or regional adequacy assessment - its inherent subjectivity will always trigger debate.


Regarding the issue of whether adequacy assessments should serve as “stress tests” rather than “forecasts”, ENTSO-E is improving its data collection processes as the input data are crucial to perform any sensible adequacy assessments. The first goal of ENTSO-E is to collect and set up a consistent (central) best-estimate Pan-European scenarios based on data collected from TSOs. Meaningful sensitivities are indeed defined around this central best-estimate Pan-European scenario as ‘stress test’ cases in MAF. It should be noted that the sensitivities relevant for ENTSO-E will always be relevant to and of the scope of Pan-European scenarios. ENTSO-E welcomes the stakeholder feedback received - it supplies an important input for revising the elaboration of meaningful sensitivities. Articulation of Pan-European, regional and national studies is hence a must as these complement, rather than compete, with each other. National studies must be consistent with the boundary conditions and assumptions provided by the larger perimeter Pan-European and regional studies and have to employ the common definitions and methodology standards defined and used by ENTSO-E.

Sensitivity analysis that will take a more ‘conservative’ view of installed capacity might be needed in future MAFs. To enable this, ENTSO-E seeks to include in the data collection process for future MAF reports estimates the volume of capacity that could be at risk of being mothballed for economic reasons, like, for example, in the absence of a capacity market and/or unfavourable market conditions – ‘missing money’ problem. These data will be analysed and sensitivities will be defined for the simulations.





95


Speculation on future regulatory developments can be easily misrepresentative and discredit the meaningfulness of the MAF. The criteria for including newly built assets should be fact-based. We do not see a reason why there should be distinguishing between countries with an energy-only market or a capacity market. In both variants, new builds are possible and either triggered by attractive forward prices or capacity tenders/obligations. We contend that a newly built unit should only be taken into account when the final investment decision is made to ensure a ‘clean’ result. As mentioned earlier, sensitivities are more suitable for a national- or regional adequacy assessment. Answers from ENTSO-E:

See answers above.

ENTSO-E and TSOs fully agree they are not the responsible for managing this issue. TSOs are not responsible for ensuring the economic viability for specific generation technologies. This clearly should be on the part of market participants themselves and not by TSOs as regulated entities. However, ENTSO-E is mandated to provide a view on the adequacy risks and to manage the security of the grid and electricity supply. These two elements are very much interlinked in the power system and should be considered together. Still, TSOs do not assume and should therefore not be assumed i) it is their sole responsibility to solve these problems by managing the generation and network interface and ii) that adequacy problems require a solution in terms of capacity exclusively as there may exist multiple solutions to reach the same adequacy targets (import/export/capacity generation and or demand flexibility). Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national





96

scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios.

Adequacy assessments should serve as “stress tests” rather than “forecasts”; ENTSO-E is improving its data collection processes as the input data are crucial to perform any sensible adequacy assessments. The first goal of ENTSO-E is to collect and set up consistent (central) best-estimate Pan-European scenarios based on data collected from TSOs. Meaningful sensitivities are indeed defined around this central best-estimate Pan-European scenario as ‘stress test’ cases in MAF. It should be noted that the sensitivities relevant for ENTSO-E will always be relevant to and of the scope of Pan-European scenarios. ENTSO-E welcomes the stakeholder feedback received - it supplies an important input for revising the elaboration of meaningful sensitivities.


• Indication of mothballing assumptions would lead to further transparency. • More charts to obtain a better understanding of the outcomes.

AnswersfromENTSO-E:

The ENTSO-E adequacy methodology considers that the large-scale exploitation of renewable energy sources of variable generation poses challenges for electricity system operation. In addition to sufficient levels of back-up capacity, additional resources for system flexibility will be needed in the future. The current focus is on the development of a robust methodology that will allow for performing a diagnosis of the power system with respect to adequacy risks. TSOs provide data considering their best knowledge on the evolution of their generation mix for these predictive forecasts. Regarding the forecasting of RES within these bottom-up (best estimate) scenarios, ENTSO-E is improving its data collection process regarding input data which affects the ‘likelihood of units to run and stay online’ within the market modelling assessments performed in TYNDP - MAF - these input data items are crucial to carrying any sensible sensitivity determination pertaining to ‘viability’ of the (central) best estimate scenarios collected from TSOs.

ENTSO-E and the TSOs know of the importance of these assumptions regarding the definition of the scenarios and ENTSO-E’s process for a common and consistent data collection has been revised to elevate the quality of the scenarios and possibly evaluate sensitivities surrounding them.

ENTSO-E strives towards a high level of transparency. ENTSO-E is committed to transparency of data and the products delivered, including the presentation of results and assumption therein. ENTSO-E also respects national legislation and confidentiality agreements between TSOs and national stakeholders. Furthermore, confidentiality issues might also require data to be publicly released in an aggregated manner.




97

COMMENTS BY STAKEHOLDERS The consistency between European, regional and national adequacy studies can be ensured through the definition of harmonized standards to be applied at all levels. These standards should include a set of common, clearly defined and transparent: (i) inputs (scenarios, assumptions, data sources); (ii) methodologies and models; and (iii) metrics for conclusions, evaluations and comparisons. In particular, the input assumptions should be consistent across all adequacy studies. It is understandable that each national TSO provides input data to ENTSO-E based on their best knowledge of the national electricity markets. Nevertheless, additional effort should be made to use harmonised hypotheses (optimistic vs. conservative) for the main parameters employed in the simulation. Answers from ENTSO-E:








ENEL SpA



98



Compared to previous European adequacy assessments, the new report represents a stepping stone in the right direction. In fact, a probabilistic approach has been introduced and the study has been extended to Turkey.

As a general recommendation, it is important that more data become available for disclosure and that transparency is enhanced with respect to the assumptions made, input data considered and analytical models utilised.

At the same time, the bigger concern is that the current methodology does not include any sensitivity linked to the economic viability of existing assets and future “newly built” ones. With this in mind, the report should also clarify the assumptions used on already installed or newly built conventional and renewable capacity, taking into account the establishment of capacity mechanisms and/or long-term price signals in a certain number of countries (i.e., the assumptions made on market design evolution over the next few years should be clarified).




99

The MAF should initially determine the amount of capacity needed in the system and, in the next step, evaluate its adequacy (i.e., compare the capacity required with the expected capacity present in the system all while considering economic viability).

With the current electricity market design, there are many uncertainties regarding future availability of existing conventional capacity and new investments in renewable and conventional capacity.

Electricity DAM prices and reduced utilisation rates are insufficient for existing thermal power plants to recover fixed and capital costs (many are still not fully financially depreciated). There is a high risk of early decommissioning/mothballing these plants if no adequate CRMs are introduced.

Besides, investments in new generation, especially in renewables, will be at risk if long-term price signals are not be provided. Renewables are characterized by large investments costs and small marginal costs. Nowadays, they do not rely upon market prices and the huge investments over the last number years have been encouraged by national support mechanisms. In the absence of such mechanisms, and with no long-term price signals, renewables will be not sufficiently remunerated while also accounting for the “cannibalization effect” in short-term markets and the fact that they do not usually benefit from peak prices during scarcity periods.

In addition, the following improvements to the proposed methodology should be made:

1. Data on demand and mothballed capacity: similar assumptions for all member states should be based on the evolution of demand (i.e., growth rates taking into consideration EE measures and EV penetration). The assumption of availability of mothballed capacity is too optimistic;

2. Results of the capacity adequacy situation are presented compared to a LOLE of one hour/year; at the same time, member states may have a different reliability standard, thus these differences should be noted as disclaimers in the MAF report;

3. The model should consider that during dry years, outages rise because of low river levels, hence the possibility of cooling down conventional power plants, as well. In addition, better specifications are needed for S2 and S6;

4. The report should specify in more detail how simultaneous scarcity situations are modelled and their effects in terms of NTCs and operational reserves; and

5. Given that one of the sensitivities is based on operational reserves, it is important that the data provided indicate the size of reserves for each country.

Finally, the following methodological improvements, already suggested by ENTSO-E, would be most welcome:

1. Use the data for 35 climatic years instead of 14.



100

2. Improve assumptions on NTC: assumptions should be sufficiently conservative, even in the baseline case scenarios, in order to maintain reliability of interconnections, maintenance outages and unexpected unavailability of system elements.

3. Model demand response: demand response and distributed storage capacities should be fully included in future MAFs, at least in sensitivity scenarios.




The purpose of the ENTSO-E adequacy methodology and MAF reports under evolution is the development of a robust methodology which will allow conducting regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Although these studies will not encompass every potential issue of future power system, ENTSO-E strongly believes that they: i) Will permit the development of a model able to capture all key features and risks regarding adequacy for the Pan-European power system. This is a main objective for ENTSO-E; ii) Will act as a common basis for methodologies and definitions; and



Demand: ENTSO-E agrees with stakeholders that demand forecasting (levels, but also profiles) as one of the main items affecting the future adequacy outlook. Therefore, ENTSO-E is working on improving modelling and the reliability of demand forecasting through enhancing methodologies. In order to ensure consistency and transparency, ENTSO-E has set up a process to define, in a centralized manner, demand forecast levels and profiles. TSOs are requested to provide national figures for the main parameters to model demand: temperature sensitivity of load, electric vehicle (EV) penetration, heat pumps (HPs), energy efficiency (EE), DSM/demand-side response (DSR), expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. In this way, consistent regional/Pan-European demand figures and profiles are defined that are also in line with national expectations and capture national specificities properly. Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value



101

for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices.

Hydro: Hydro modelling details can have a significant influence on adequacy indicators, especially of regions, including countries, with large shares of hydro production. Dry and wet hydro conditions (inflow and reservoir) have been considered in MAF 2016. In addition, the ENTSO-E PECD is currently being extended to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the probability of occurrence of these dry-normal and wet conditions is being investigated systematically in order to define so-called hydrological regions for which the same probability of similar conditions can be accounted for in Monte Carlo probabilistic simulations. New Methodological Improvements:





102



ENTSO-E must fulfil obligations of Reg. 714/2009 and contribute to the overall European/national debates on adequacy concerns. ENTSO-E is committed to transparency of data and products delivered while respecting national legislation and confidentiality agreements between TSOs and national stakeholders. Furthermore, confidentiality issues might also require data to be publically released in an aggregated manner. ENTSO-E endeavours towards a high level of transparency on key parameters and assumptions. Improvement of key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the various generation technologies available. Such exchange could enhance the available data from TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as ENTSO-E agrees that the results should always be understood within the context of the data and assumptions used. Thermal Outage Correlation with Climatic Conditions: Extensions of the methodology might be considered to capture correlations of thermal outages, thermal production and climate conditions, notably temperature conditions. It is worth noting that this effect is very pronounced in certain countries, e.g., PL, but might not require a Pan-European deployment, and instead require local sensitivities with respect to the MAF results. Sensitivity of Scenarios:

ENTSO-E must fulfil obligations of Reg. 714/2009. The European generation adequacy outlook shall build upon national generation adequacy outlooks prepared by each individual transmission system operator.The first goal of ENTSO-E is to collect and set up a contiguous (central) best estimate ‘Pan-European’ scenario based on data collected from TSOs as meaningful sensitivities are defined around this. ENTSO-E is therefore revising its data collection and scenario-building process in order to consider relevant sensitivities based on data provided by TSOs on, e.g., units at risk of being mothballed for economic reasons, information regarding decommissioning of units for purely technical reasons and/or shut downs for legal/policy reasons within the storyline assumptions of the scenarios considered. Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national



103

scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios. Flow-Based Methods


The integration of physical grid concepts (capacities, impedances) directly into market models, usually referred to as “flow-based methods”, avoid the unrealistic results pitfall. ENTSO-E is currently determining how to define and put in place flow-based methods within long-term Pan-European market studies, the main purpose being to make commercial exchanges more realistic and faithful to actual physical flows versus the current approach. It should be noted that flow-based methods are currently deployed by a number of TSOs in national studies (Belgium and France) and are in the scope of regional studies, e.g., PLEF. Pan-European studies, like MAF, will profit from these evolutions and will integrate them at the EU level once mature at the national and regional levels. NTC – Conservative: Conservative assumptions have been made regarding cross-border transmission capacity relating to uncertainty in the commissioning dates of cross-border transmission projects and adequacy risks therein. These assumptions are within the scope of the methodology and report capturing all key features and risks relevant at the Pan-European level - this is a main objective for ENTSO-E.

Aconservative approach of ENTSO-E in providing transfer capacity values for adequacy analyses is sensible, even if adequacy risk might be overestimated. This will mean that the energy mix considered will not be compatible with the assumed/available transmission infrastructure. In this respect, the MAF results provide a signal in the form of ‘adequacy risk’. This could very well be higher than the desired adequacy level that a country is willing to accept or prepare for, expressed in the form of its adequacy standard. As a consequence of the fact that investments in grid infrastructure require a longer planning and decision horizon (~15 years) than the typical investment cycles for investments in new generation (3-5-7) years, the MAF exercise will detect risks and provide important feedback regarding the establishment of countermeasures by relevant stakeholders (e.g., member state authorities, policy makers, regulatory agencies, energy producers) in order to ensure the desired adequacy levels and the risks linked to their choices of generation mix and its evolution as well the impact of cross-border capacity availability.

Furthermore, national or regional assessments might consider different assumptions that could be more optimistic regarding availability of cross-border transmission infrastructure and availability of imports to complete the assessment of the impact of such assumptions on adequacy levels. 15-Minute Flexibility: The purpose of the ENTSO-E target adequacy methodology being developed is the generation of a sound methodology that will foster performing regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks.



104

Considering any modelling implies simplifications, so ENTSO-E adequacy target modelling focuses on hourly power balance modelling. Although these studies do not encompass every potential issue involved in the state of the future power system, ENTSO-E strongly believes that this methodology will permit the development of a model able to capture all the key features and risks regarding adequacy for the Pan-European power system - this is a main objective for ENTSO-E.



Considering the above, what additional methodological improvements should ENTSO-E consider for future MAFs? Please justify why through suggestions and what these improvements will offer. COMMENTS BY STAKEHOLDERS

No Comment.

COMMENTS BY STAKEHOLDERS These data should not be provided by generators themselves as a consequence of commercial confidentiality issues and uncertainties on the market’s evolution. Instead, a sensitivity analysis on the availability/decommissioning/mothballing of plants should be conducted based on assumptions proposed by ENTSO-E in consultation with stakeholders regarding: (i) the evolution of the electricity market design (capacity market vs. energy-only market, long-term price signals); (ii) the evolution of variable/fuel costs from international sources of data (e.g., IEA, investment banking reports, others); iii) CO2 price; iv) merit order effect; v) reliability margins; vi) policies and regulation; and vii) political decisions.




105

Rather than concentrating on particular power plants, sensitivity analyses should be performed at the level of economic viability of the different available generation technologies (top-down approach). In any case, the study should not be the central form of evidence for the introduction of a capacity remuneration mechanism in a member state. Answers from ENTSO-E:

A framework to express national targets for SoS on the basis of commonly agreed adequacy standards at the regional or Pan-European level is necessary. ENTSO-E methodology provides the basis for such a framework. The adequacy levels should be nationally set but understood within a coordinated regional approach and based on the principles of subsidiarity regarding the definition of measures necessary to fulfil these standards by member states. Risks associated with these standards must be commonly understood across Europe through common definitions. Indeed, decisions to implement measures to ensure SoS at the national level is directly connected to the responsibility for SoS. Only the entity responsible for SoS can make the decision to implement or not measures to guarantee SoS. This however should be carried out in coordination with neighbouring member states and TSOs as a result of the cross-border impact that these decision might have on neighbouring areas.

The first goal of ENTSO-E is to collect and set up a consistent (central) best-estimate Pan-European scenarios based on data collected from TSOs. Meaningful sensitivities are indeed defined around this central best-estimate Pan-European scenario as ‘stress test’ cases in MAF. It should be noted that the sensitivities relevant for ENTSO-E will always be relevant to and of the scope of Pan-European scenarios. ENTSO-E welcomes the stakeholder feedback received - it supplies an important input for revising the elaboration of meaningful sensitivities.


As already described in the previous response, economic viability of generation assets should be established for classes of generation technologies and utilised in sensitivity analysis scenarios.

Assumptions and data would be best based on internationally reliable sources (international agencies or private institutions) rather than on data provided by the individual producers.





106

The objective of the MAF should be determining the necessary adequate capacity in the integrated market considering, in particular, sensitivity scenarios for economic viability. Only the final market design, including possible schemes of capacity remuneration, will inform the viability of the different technologies. For this reason, the study should not be the central evidence for the introduction of a capacity remuneration mechanism in a member state.


Economic/technical data on generation technologies is needed to perform market simulations. The hypotheses are gathered by the TSOs according to their best knowledge and the availability of data at the time of data collection. ENTSO-E and the TSOs involved in these studies can therefore not be held responsible in the case the hypotheses from these reports or the estimations based on these hypotheses are not realised in the future. ENTSO-E welcomes greater transparency and availability from market participants with respect to the economic/technical data to improve the accuracy of the forecasts on economic/technical data used by TSOs in the simulations performed. It is very important to have superior visibility into decommissioning and mothballing figures in order to set up the relevant generation scenarios. The MAF data collection process will be revised for MAF 2017 in order to improve visibility of these data provided by TSOs and processed to set up relevant sensitivities that will explore possibilities to consider different views on installed capacity, from ‘optimistic’ to ‘conservative’ situations. The sensitivities planned are in line with the feedback provided by the stakeholder. Indeed, decisions to implement measures to ensure SoS at the national level is directly connected to the responsibility for SoS. Only the entity responsible for SoS can make the decision to implement or not measures to guarantee SoS. This however should be carried out in coordination with neighbouring member states and TSOs as a result of the cross-border impact that these decision might have on neighbouring areas.

For this, articulation of Pan-European, regional and national studies is hence a must as these complement rather than compete with each other. The Pan-European study should not be used as single reference and must be complemented by regional and national studies, which in turn must also be consistent with boundary conditions and assumptions provided by the larger perimeter Pan-European and regional studies.





107


We strongly support the implementation of the two aforementioned scenarios in the next releases of the MAF.

Particular attention should be paid to the assumptions on the capacity market scenarios, specifically taking into consideration (i) the next developments regarding the market design package and (ii) the different stages of discussion/implementation in the member states.


The baseline and economic viability sensitivities planed for MAF 2017 are in line with the “low-“and “high-stress” proposals provided via stakeholder feedback. We appreciate and thank the stakeholders for this comment.


Focusing on the assumptions made for Spain:

A) Spain generation capacity assumptions:

The hypothesis on future generation capacity underestimates the current status of the electricity sector.

European electricity markets do not supply adequate price signals to guarantee the permanency of existing generation plants nor in terms of acquiring new investments in renewable capacity to fulfil the decarbonisation objectives. Such issues are not short-term problems, but rather a structural failure (electricity markets have been designed to reflect and optimize systems where conventional technologies are dominant and not for systems mainly driven by technologies with very low/close to zero marginal costs).

1. Electricity market prices are insufficient for existing thermal plants to recover their fixed and capital costs, even taking into account the revenues from capacity payments. Indeed, generation plants are not fully financially depreciated, but rather the opposite. There is a high risk of early decommissioning/mothballing of these generation plants. Both the Spanish government and NRA are aware of this situation. The Spanish Master Plan was used as a reference input source for MAF (we assume MAF refers to the document “Planificación Energética, plan de desarrollo de la red de transporte de energía eléctrica 2015-2020” from the Spanish Ministry of Energy, MINETUR) and outlines the expectation the of decommissioning/mothballing 6,000 MW of CCGTs in the near term. Furthermore, the NRA (CNMC) considers, in its report on the Master Plan, an additional decommissioning of 2,000-3,000 MW CCGTs in the horizon of 2015-2017. Apparently, these have not been taken into account in the MAF assumptions for 2020.




108

2. Electricity prices do not serve as an adequate long-term signal for investing in new generation capacity, renewable sources in particular.

RES are characterized by a cost structure with high investment costs and low/close to zero marginal costs. Presently, support mechanisms guarantee a reasonable rate of return on investments so that RES do not rely on market prices. RES are highly affected by the so-called “cannibalization effect” and do not usually benefit from peak prices. In the absence of support mechanisms and with no long-term price signals, all this would lead to a higher cost of capital with resultant investment risk. The steady growth of renewable capacity predicted in the MAF seems unlikely in the absence of support mechanisms and/or long-term price signals.

3. The permanence of 7,390 MW of CHP in 2020 appears to be overestimated. Seeing that the Spanish reform of the electricity sector and, specifically, the adjustment of the remuneration framework for RES and co-generation, many CHPs have shut down. The latest report from REE noted 6.684 MW of installed capacity of CHP in 2015. In addition, co-generation will be forced to comply with facility renovation programs under Law 24/2013, which should be implemented during the first quarter of the establishment of the new government. Such obligations will require additional investments that could result in further decommissioning.

4. It is more reasonable to assume that in 2020, the installed capacity of coal will be 4.88 GW, not 9,40 GW. The MAF considers the permanence in 2020 of 9.40 GW of coal power, which is reduced to 4.88 GW in 2025 based on the decommissioning of coal plants unable to realize the necessary investments to produce this with emission caps set by the DEI. It should be noted that plants not engaging in DEI investments and operating under the National Transitional Plan, which foresees a decline in cap emissions, will have to shut down no later than June 30, 2020, and probably before as plants operators will seek to optimize margins and reduce unrecoverable fixed costs.

Therefore, we believe there is a room for improvement in MAF assumptions on future generation installed capacity. To this end, and as previously indicated in the comments to questions 1, 2 and 3 of the consultation, the MAF should include sensitivity analysis of the viability of power plants.

B) MAF assumption of Spain peak demand:

Peak demand in 2025 is probably underestimated. The Spanish Master Plan predicts a steady annual demand growth of 2% from 2015 to 2020, reaching a peak demand of 47,300 MW in 2020. Indeed, MAF considers a demand peak of 47,000 MW in 2020.

However, from 2020 to 2025, MAF predicts an annual demand growth of 0.4% and peak demand reaching 48,000 MW. That means 4,000 MW less than continuing with the Master Plan´s demand growth of 2% and 3,000 MW less than peak demand estimated in the SO&AF 2015 for 2025.

Such demand slowdown from 2020 to 2025 is not fully justified, especially if MAF does not include demand response. Additionally, the energy sector will have to work with a quite challenging decarbonisation goal, in which electrification of demand would play an essential part.



109


The ENTSO-E adequacy methodology considers that the large-scale exploitation of renewable energy sources of variable generation poses challenges for electricity system operation. In addition to sufficient levels of back-up capacity, additional resources for system flexibility will be needed in the future. The current focus is on the development of a robust methodology that will allow for performing a diagnosis of the power system with respect to adequacy risks.


i) Will permit the development of a model able to capture all key features and risks regarding adequacy for the Pan-European power system. This is a main objective for ENTSO-E; and

ii) Together with other studies/activities performed by ENTSO-E, these analyses will ultimately assist market design and network codes.

With that said, it is worth mentioning that ENTSO-E is aware of the impact on adequacy by setting forth an optimal and economically feasible set of investment plans in generation portfolio.




110

COMMENTS BY STAKEHOLDERS ENTSO-E should take the lead by identifying best practices and data definitions that TSOs can adhere to. ENTSO-E should also promote synchronized TSO timelines for national resource adequacy assessments The resource adequacy landscape across Europe is uneven and a regional approach could potentially even out capacity deficits and surpluses. ENTSO-E might also lead the establishment of the most appropriate boundaries for regional resource adequacy assessments and coordinate TSO efforts to produce those assessments using the MAF methodology template. Voluntaryassessments, such as that produced by PLEF, are ofvaluebutmaynotnecessarilyutilisethemost appropriate national groupings from a resource adequacy perspective. Answers from ENTSO-E:








REGULATORY ASSISTANCE

PROJECT



111



It seems highly likely that demand response will significantly contribute to resource adequacy over the MAF planning horizon. Therefore, a harmonised approach to identifying that contribution should be established as soon as possible. Although demand response’s role will vary from country to country, a harmonised approach to analysing demand response potential is desirable and is an endeavour of which ENTSO-E should spearhead.


Regarding Planned Methodological Improvements

Demand: ENTSO-E agrees with stakeholders that demand forecasting (levels, but also profiles) as one of the main items affecting the future adequacy outlook. Therefore, ENTSO-E is working on improving modelling and the reliability of demand forecasting through enhancing methodologies. In order to ensure consistency and transparency, ENTSO-E has set up a process to define, in a centralized manner, demand forecast levels and profiles. TSOs are requested to provide national figures for the main parameters to model demand: temperature sensitivity of load, electric vehicle (EV) penetration, heat pumps (HPs), energy efficiency (EE), DSM/demand-side response (DSR),




112

expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. In this way, consistent regional/Pan-European demand figures and profiles are defined that are also in line with national expectations and capture national specificities properly. Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices.

Hydro: Hydro modelling details can have a significant influence on adequacy indicators, especially of regions, including countries, with large shares of hydro production. Dry and wet hydro conditions (inflow and reservoir) have been considered in MAF 2016. In addition, the ENTSO-E PECD is currently being extended to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the probability of occurrence of these dry-normal and wet conditions is being investigated systematically in order to define so-called hydrological regions for which the same probability of similar conditions can be accounted for in Monte Carlo probabilistic simulations. New Methodological Improvements:


Different market modelling tools available, owned or procured by TSOs are utilised during the MAF exercise. These tools are proven to capture nationally and regionally relevant specificities. MAF work relies on this know-how from TSOs in terms of the strong bottom-up basis of all knowledge and detailed data. Tool dependency and reliability is a major focus for MAF. Use of exclusively open-source tools might be beneficial from a purely ‘academic’ point of view, though on the other hand, does not guarantee a strong link to bottom-up national details – know-how of



113

TSOs and national studies. Full transparency is not always possible because these tools are, in certain cases, proprietary materials from different tool providers in a similar way that generation and market data is confidential to market parties as communicated via the stakeholder feedback and answers throughout this document. Total Need for Reliable/Dispatchable Capacity (MW):


ENTSO-E will cautiously account for the possibility of reporting the aforementioned figure in the presentation of the results in future MAF reports. Providing views on specific generation technologies is not necessarily the role of TSOs as regulated entities. This indicator seems related to the deterministic indicator, remaining capacity margin (RCM), employed by ENTSO-E in SO&AF 2015 and previous reports. This indicator was not chosen in MAF 2016 in order to focus attention to the new probabilistic method deployed and LOLE and ENS indicators used. Transparency of Key Parameters, Assumptions and Data:

ENTSO-E must fulfil obligations of Reg. 714/2009 and contribute to the overall European/national debates on adequacy concerns. ENTSO-E is committed to transparency of data and products delivered while respecting national legislation and confidentiality agreements between TSOs and national stakeholders. Furthermore, confidentiality issues might also require data to be publically released in an aggregated manner.


ENTSO-E must fulfil obligations of Reg. 714/2009. The European generation adequacy outlook shall build upon national generation adequacy outlooks prepared by each individual transmission system operator.The first goal of ENTSO-E is to collect and set up a contiguous (central) best estimate ‘Pan-European’ scenario based on data collected from TSOs as meaningful sensitivities are defined around this. ENTSO-E is therefore revising its data collection and scenario-building process in order to consider relevant sensitivities based on data provided by TSOs on, e.g., units at risk of being mothballed for economic reasons, information regarding decommissioning of units for purely technical reasons and/or shut downs for legal/policy reasons within the storyline assumptions of the scenarios considered. Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national



114

scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios. Flow-Based Methods


The integration of physical grid concepts (capacities, impedances) directly into market models, usually referred to as “flow-based methods”, avoid the unrealistic results pitfall. ENTSO-E is currently determining how to define and put in place flow-based methods within long-term Pan-European market studies, the main purpose being to make commercial exchanges more realistic and faithful to actual physical flows versus the current approach. It should be noted that flow-based methods are currently deployed by a number of TSOs in national studies (Belgium and France) and are in the scope of regional studies, e.g., PLEF. Pan-European studies, like MAF, will profit from these evolutions and will integrate them at the EU level once mature at the national and regional levels. NTC – Conservative: Conservative assumptions have been made regarding cross-border transmission capacity relating to uncertainty in the commissioning dates of cross-border transmission projects and adequacy risks therein. These assumptions are within the scope of the methodology and report capturing all key features and risks relevant at the Pan-European level - this is a main objective for ENTSO-E.

A conservative approach of ENTSO-E in providing transfer capacity values for adequacy analyses is sensible, even if adequacy risk might be overestimated. This will mean that the energy mix considered will not be compatible with the assumed/available transmission infrastructure. In this respect, the MAF results provide a signal in the form of ‘adequacy risk’. This could very well be higher than the desired adequacy level that a country is willing to accept or prepare for, expressed in the form of its adequacy standard. As a consequence of the fact that investments in grid infrastructure require a longer planning and decision horizon (~15 years) than the typical investment cycles for investments in new generation (3-5-7) years, the MAF exercise will detect risks and provide important feedback regarding the establishment of countermeasures by relevant stakeholders (e.g., member state authorities, policy makers, regulatory agencies, energy producers) in order to ensure the desired adequacy levels and the risks linked to their choices of generation mix and its evolution as well the impact of cross-border capacity availability.

Furthermore, national or regional assessments might consider different assumptions that could be more optimistic regarding availability of cross-border transmission infrastructure and availability of imports to complete the assessment of the impact of such assumptions on adequacy levels. 15-Minute Flexibility: The purpose of the ENTSO-E target adequacy methodology being developed is the generation of a sound methodology that will foster performing regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks.



115

Considering any modelling implies simplifications, so ENTSO-E adequacy target modelling focuses on hourly power balance modelling. Although these studies do not encompass every potential issue involved in the state of the future power system, ENTSO-E strongly believes that this methodology will permit the development of a model able to capture all the key features and risks regarding adequacy for the Pan-European power system - this is a main objective for ENTSO-E.





A harmonised approach to assessing demand response potential and resource adequacy potential will improve consistency.


See above Demand and Demand-Side Assumptions.


It should be possible for TSOs to make informed judgements on likely plant closures or mothballing by using the market and other intelligence they have access to. It may also be useful to establish European-level rules with respect to supplying notice of intended decommissioning or




116

plant closures. However, these rules should not place unreasonable constraints on the closure or mothballing of a plant or force plants that are no longer economically viable to remain operational. Answers from ENTSO-E:




It would be best if TSOs had access to sufficient market and member state policy information to make informed decisions concerning the economic viability of individual generation plants so that the "expected progress/best-estimate" approach would seem reasonable. Answer by ENTSO-E:

Economic/technical data on generation technologies is needed to perform market simulations. The hypotheses are gathered by the TSOs according to their best knowledge and the availability of data at the time of data collection. ENTSO-E and the TSOs involved in these studies can therefore





117

not be held responsible in the case the hypotheses from these reports or the estimations based on these hypotheses are not realised in the future.

ENTSO-E welcomes greater transparency and availability from market participants with respect to the economic/technical data to improve the accuracy of the forecasts on economic/technical data used by TSOs in the simulations performed. It is very important to have superior visibility into decommissioning and mothballing figures in order to set up the relevant generation scenarios.

The MAF data collection process will be revised for MAF 2017 in order to improve visibility of these data provided by TSOs and processed to set up relevant sensitivities that will explore possibilities to consider different views on installed capacity, from ‘optimistic’ to ‘conservative’ situations. The sensitivities planned are in line with the feedback provided by the stakeholder.


It seems most sensible to base the scenarios on the foreseen regulatory framework, including any proposals for capacity support or reliability standards, for each member state. If in future years, the foreseen framework changes, then that can be reflected in subsequent MAFs.

However, the use of a single scenario for each of the two planning years seems questionable. There are a range of possible outcomes in terms of economic progress and decarbonisation that could affect future capacity requirements and resource adequacy. An attempt to assess resource adequacy over a range of possible economic and decarbonisation outcomes could be advantageous, although this would definitely add to the computational burden. Answers from ENTSO-E:

Scenarios for MAF 2020 and 2025 and TYNDP 2030 and 2040 – differences:

Exploratory Scenarios: ENTSO-E has developed long-term scenarios, e.g., 2030 scenarios used within TYNDP are exploratory with respect to generation, demand and Pan-European adequacy. Each follows different storylines indicating possible futures with certain Pan-European (top – down) political – economic – technological assumptions. Over this long-term horizon, e.g., RES deployment and EE are considered explicitly, e.g., taking into account National Renewable Energy Action Plans (NREAPs) and EC targets with intrinsic uncertainties surrounding the evolution of the energy mix for a long horizon.





118

As a result of this uncertainty and because of the shorter investment cycle for new generation, the suitable time frame to assess generation adequacy at national and regional resolution, anticipating possible adequacy issues, is 5-10 years (maximum) as recommended by the ECG – subgroup in adequacy. Within this time frame, trustable diagnoses of generation adequacy risks are possible by use of a sound, widely accepted and transparent methodology. Bottom-up (best estimate) scenarios: ‘… built national generation adequacy outlooks prepared by each individual TSO’ are therefore used within the MAF so the risks of adequacy are identified in due time (beyond 5 years ahead and up to 10 years). Adequacy assessments are therefore performed for ‘predictive’ mid-term scenarios rather than those that are ‘exploratory’ over the long-term. Sensitivity of Scenarios:




The MAF acknowledges that there is more work in terms of including demand response potential in resource adequacy assessments. However, given this potential and the focus on customer participation in the Commission's vision for the future electricity market, the cautious approach to demand response adopted in the MAF is disappointing. A more sophisticated and consistent approach to modelling demand response and EE is absolutely necessary to account for end-users in power systems. As with the introduction of a stochastic approach for more accurate modelling of the effects of increased intermittent capacity, there is a need to model demand-side resources more accurately taking into consideration their importance in helping reduce system peaks and integrating intermittent renewable generation.




119


The ENTSO-E adequacy methodology considers that the large-scale exploitation of renewable energy sources of variable generation poses challenges for electricity system operation. In addition to sufficient levels of back-up capacity, additional resources for system flexibility will be needed in the future. The current focus is on the development of a robust methodology that will allow for performing a diagnosis of the power system with respect to the main adequacy risks at the Pan-European level. One of the key objectives is to be able to inform the system with the 'need for flexibility'. Flexibility in adequacy assessments is based on weather-dependent effects related to load variation, generation patterns of wind and solar power plants with a one-hour resolution and the consideration of the resources for flexibility. The PECD will be utilised for these adequacy assessments. Although these studies will not encompass every potential issue involved in the future power system, ENTSO-E strongly believes that this methodology will permit the development of a model able to capture all the key features and risks regarding adequacy for the Pan-European power system. This is a major aim of ENTSO-E; together with other studies/activities performed by ENTSO-E, these analyses will ultimately assist improving market design and network codes. In that sense, it is worth mentioning that ENTSO-E is aware of the impact of adequacy on designing an optimal and economically feasible set of investment plans within a generation portfolio. See Methodological Improvements above - Demand and Demand-Side Assumptions.



120

COMMENT BY STAKEHOLDER Consistency between European, regional and national adequacy studies should be ensured through aligned (1) input structure (including scenarios) and assumptions; (2) methodologies and parameters; and (3) metrics and granularity for conclusion/evaluation. Without alignment on these three elements (input – processing – output), it is not possible to meaningfully compare/combine adequacy studies performed at different levels. The objective is not necessarily to use the same numbers everywhere –local variations, specificities and circumstances should be correctly reflected – but they should follow a similar structure so that variations/changes/evolutions can be easily detected, queried/challenged and changed/substantiated. The use of publicly available software (commercial/open source) would assist making certain there is buy-in from all stakeholders regarding the processing aspect. Furthermore, it is crucial to take into account import/export and cross-border capacity availability in a contiguous manner, in particular for national adequacy studies. An adequacy assessment covering a larger geographical scope is more comprehensive in this manner as it considers interdependencies between neighbouring countries. EURELECTRIC therefore sees strong added value in developing European/regional adequacy studies to complete assessments made at the national level. In addition, we view an increasing role for RSC to also address SoS and capacity adequacy issues to enhance consistency of adequacy assessments. Last but not least, the time horizon and years of investigation should also be aligned to enhance consistency and comparability among adequacy studies. Answers from ENTSO-E:

On articulation between national - regional - Pan-European reports -- methodology and report scope ENTSO-E appreciates the positive feedback from stakeholders regarding the ENTSO-E methodology as a common basis. ENTSO-E views this role as follows: a) ENTSO-E describes and proposes common definitions for the multiple concepts/aspects under Security of Supply (SoS) and adequacy, providing all stakeholders with a common language;


EURELECTRIC



121






"EURELECTRIC welcomes the methodological improvements brought forth by ENTSO-E in the MAF 2016, in particular the introduction of a probabilistic assessment of adequacy in Europe and the Pan-European coverage of the study, including Turkey.




122

The following improvements, as already suggested by ENTSO-E to enhance the robustness of the next version of the MAF, would be most welcome:

• Use the data for 35 climatic years instead of 14; • Improve assumptions on NTC: Regarding the current assumptions on NTC, we understand

that for computational reasons, and because most adequacy criteria are defined at the national level, ENTSO-E chose to model every bidding zone as a single node. It should however be noted that, even with an unchanged infrastructure, the NTC values are often dependent on the actual demand/generation. For instance, NTC in Germany is reliant upon the amount of wind generated in Northern Germany. This may justify considering more conservative NTC values in the MAF until finer and more granular flow-based models are available.

• Model demand response: Market-based demand response can provide considerable support for matching the supply and demand in peak situations. Besides industrial and commercial users, domestic consumers are expected to react pronouncedly to market prices thanks to smart metering and spot-based pricing. Batteries and electric cars could also act as additional resources in the future. In the current MAF, demand response and distributed storage capacities are just implicitly modelled. When calibrating thermal sensitivity of load, ENTSO-E indeed already accounts for the demand response potential as valued (implicitly or explicitly) today, leading to a reduction of demand when prices are high (i.e., very low or very high temperatures depending on the country). A more explicit modelling of demand response would be welcome in future versions of the MAF report to explore the economic potential that is currently unexplored.

• Use flow-based models.

On top of those improvements already suggested by ENTSO-E, EURELECTRIC proposes the following:

• More transparency: In general, we believe that the MAF 2016 report does not provide sufficient transparency on the parameter values for obtaining the presented results. It is necessary that ENTSO-E provide full information on the assumptions regarding generation portfolios, demand features and cross-border exchange capabilities for each scenario. As an example, the current dataset supplies installed capacity by plant type but does not provide information about planned and forced outage rates, size of operational reserves, capacity factors and total generation (which makes verifying the results against the demand or other sources difficult). Similarly, a key sensitivity of the MAF is based around operational reserves contributing to adequacy or not. Yet, the published dataset does not offer information regarding the size of reserves used for each country.

• Avoid treating mothballed capacity as being available in the study years and accounting for it consistently for all countries: The MAF provided a guideline to TSOs whereby all power plants that are currently mothballed were to be tagged as available for the MAF. However, this guideline does not seem to have been followed by all TSOs. Based on the information in Sections 6.2.11 and 6.2.19, France and Italy treat mothballed capacity as available capacity, while other countries, such as the Netherlands, do not. This distorts the perspective on future adequacy. We contend that mothballed capacity should not be treated



123

as being available. Recent years have certainly demonstrated that the current design of European energy markets does not properly valuate reliable and firm capacity. Several member states have thus introduced capacity mechanisms as an additional means to reduce the risk faced by capacity providers and reach their sought-after level of adequacy. In this context, the MAF assumption (all capacities have to be modelled as online for system adequacy assessment, even if currently mothballed) is far too optimistic unless the capacity is selected in a capacity market covering the area (see Irish TSO remarks in the MAF on closures of capacities not successful in the capacity market auction). Hence, EURELECTRIC puts forth running sensitivity analyses that takes explicitly into account the possible absence of the equivalent amount of reliable capacity.

• Treat demand consistently: The demand data from individual TSOs is inconsistent. While much of the demand data corresponds to baseline case growth (e.g., Greece), other countries use a more conservative (from an adequacy perspective, i.e., higher) growth rate (e.g., Italy) and others simply employ a forecast created for one of their internal studies (e.g., UK). There is also a very different approach to EE (e.g., Bulgaria: “Little to no EE measures were considered for the resulting forecasts“ vs. Croatia: “Investments in EE are expected and that will slightly slow the growth of electricity consumption“) or EV penetration. EURELECTRIC argues that ENTSO-E should define a consistent methodology for demand forecasting and assumptions of national TSOs. This would include, among others, alignment on the macro-economic assumptions (e.g., coherent GDP or demographic growth rates), on the EE gains, on prosumer development, etc. across countries.

• Addressing all adequacy issues: Assessing whether sufficient flexibility exists in the system to cope with large and fast load variations or excess renewable energy in the summer will probably become as important as LOLE estimates in the near future (including summers with inflexible generation and renewable sometimes exceeding demand). For example, the Belgian TSO looked at the issue in its latest adequacy report, but the flexibility requirements discussed were mostly those of the grid and not those of providers - harmonising flexibility analysis methods is hugely necessary. The study should therefore also estimate the system’s adequacy of flexibility requirements (i.e., ramp-up and ramp-down requirements).

• Consider the development of decentralised generation: Properly modelling such developments in the future requires DSOs to have more transparency and visibility into the (existing/expected) decentralized capacity connected to their grid as well as associated generation.

Modelling improvements:

a) MAF 2016 uses four different simulation tools and the average resultant value as a basis index for generation adequacy. Providing a view on the distribution (e.g., values of standard deviation, p50, p95, etc.) will complete the understanding of the adequacy level. The use of proprietary simulation tools does not allow stakeholders to understand in much detail how the modelling was performed (algorithms, parameters and their tuning, limitations, etc.).



124

b) Hydro management should be better integrated into the analysis - proper simulations of hydro stock management over the year are a must, just as pointed out by the Norwegian TSO in the Norway MAF comments. The models utilised (e.g., based on a SDDP approach) should be able to calculate water values over a hydrological season, at least for hydro-intensive countries (Nordics, France, etc.). In addition, sensitivities, S2 and S6, in terms of hydro optimisation, should be better explained.

c) The weather scenarios employed by ENTSO-E are too narrow: Several years of extreme weather in the recent past and climate change make older events (e.g., winter 1954) less relevant: according to Météo France, “The [climate change] trend calculated for the 1979-2005 period was 0.55 ° C/decade [in France]”. In that context, even the larger scenario set used by the MAF is too restricted. It would be valuable for the MAF to adopt a larger set of weather scenarios defined according to “current climate”, like, for instance, the 200 scenarios developed recently for RTE that seemed to cover all of Europe (see https://cpr.concerte.fr/sites/default/files/20160229_Sc%C3%A9narios%20climatiques_Meteo%20France.pdf)

d) The report does not contain any information on how simultaneous scarcity situations are modelled: The management of such events should be taken into account given its potential impact on the adequacy assessments of adjacent countries/zones.

e) Exchanges with non-ENTSO-E countries should be ideally modelled as market-based deliveries or as a minimum being fully or partially able to cover deficit situations, and not as pre-defined data series, in all cases where these deliveries can be expected to fully or partially cover deficit situations depending on market prices.

f) The model should also consider an additional connection/link between variables: Indeed, during dry years, forced outages are likely to rise as it is not possible to cool down conventional power plants when river levels are low.

g) Results of the capacity adequacy situation are compared to a generic reliability standard (LOLE of one hour/year). We welcome this approach as it permits comparing forecasts of SoS in all member states. However, many countries have a different national reliability standard, often less strict than one hour/year. We would welcome any illustration of why such strict reliability standards were opted for.

Most importantly, our biggest concern was that the current MAF does not include any parameters able to express the sensitivity linked to the demand (e.g., GDP, demographic growth rates, EE gains, prosumer development) nor to the supply side (e.g., economic viability of existing assets or development of RES capacity, including decentralised generation). It is therefore crucial to develop, in future iterations of the MAF report, realistic high-level sensitivities for the aforementioned elements. As for the decommissioning of existing power plants because of economic reasons (and not only because it is the end of their technical lifetime), a dispatch model including fuel prices and carbon was used, so a sanity check of the results must be carried out. Comparing the gross margin of various units to representative annual fixed costs (e.g., as provided in other adequacy studies) could allow validating (or not) the strong assumption that all mothballed units are able to remain economically available for SoS. We provide more input on this



125

point in the section of the consultation dedicated to “Economic viability of mid-term adequacy forecasts”.






ii) Will act as a common basis for methodologies and definitions; and iii) Will greatly facilitate consistent articulation between Pan-European, regional and national adequacy studies. ENTSO-E is also aware that the choice of the mathematical approach could significantly affect the indicators that may be assessed through the simulation as well as the structure and complexity of the input data and modelling assumptions used.



Demand-Side Assumptions: The capacity of demand reduction available in the market will be incorporated into the data collection process for future MAF reports. This demand reduction should be price responsive and will be applied in market models when prices rise above the value for defined price bands. An effort to consolidate DSM/DSR, expected EUR/MWh, MWh, duration and weekly constraints for activation, etc. is underway to incorporate these parameters into the ENTSO-E process. It should be noted, however, that currently, such figures are subject to



126

significant uncertainty and are not consolidated, which is agreed at the EU level, e.g., price bands for DSM activation should be based on a forecasted expectations of future prices rather than current prices. Hydro: Hydro modelling details can have a significant influence on adequacy indicators, especially of regions, including countries, with large shares of hydro production. Dry and wet hydro conditions (inflow and reservoir) have been considered in MAF 2016. In addition, the ENTSO-E PECD is currently being extended to consider geographical correlations of hydro production with hydrological conditions - e.g., rainfalls and snow melts. Furthermore, the probability of occurrence of these dry-normal and wet conditions is being investigated systematically in order to define so-called hydrological regions for which the same probability of similar conditions can be accounted for in Monte Carlo probabilistic simulations. New Methodological Improvements:


Visibility of Data: ENTSO-E strives toward a high level of transparency. The so-called 'MAF2016_market_modelling_data' package, including all input NGC, demand and NTC figures, has been made available upon publication of the report. ENTSO-E strives to enhance transparency regarding other key parameters used for the simulations. Transparency also implies responsibility not only of the publisher, but also of the recipient. Improvement of these key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the different generation technologies available. Such exchange could also make better the available data by TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as it concurs that the results should always be understood within the context of the data and assumptions used. Transparency of Algorithms:





127


ENTSO-E must fulfil obligations of Reg. 714/2009 and contribute to the overall European/national debates on adequacy concerns. ENTSO-E is committed to transparency of data and products delivered while respecting national legislation and confidentiality agreements between TSOs and national stakeholders. Furthermore, confidentiality issues might also require data to be publically released in an aggregated manner. ENTSO-E endeavours towards a high level of transparency on key parameters and assumptions. Improvement of key parameters would be possible by discussion with relevant market parties/actors that can provide detailed technical data and feedback regarding the various generation technologies available. Such exchange could enhance the available data from TSOs for MAF. ENTSO-E welcomes such interactions with relevant stakeholders as ENTSO-E agrees that the results should always be understood within the context of the data and assumptions used. Thermal Outage Correlation with Climatic Conditions: Extensions of the methodology might be considered to capture correlations of thermal outages, thermal production and climate conditions, notably temperature conditions. It is worth noting that this effect is very pronounced in certain countries, e.g., PL, but might not require a Pan-European deployment, and instead require local sensitivities with respect to the MAF results. Sensitivity of Scenarios:


Pan-European accuracy and scope is the focus of MAF. This indeed does mean that some extreme, nationally specific and relevant cases could be missed. Pan-European studies consider relevant and consistently constructed central EU scenarios which appraise adequacy properly but are limited by definition in the number of sensitivities of the scenarios used that can be explored to capture all nationally specific extreme conditions for every country’s situation. As such, MAF studies should always be complemented by regionally sound sensitivities by regional and national scenarios/studies where detailed impacts of these risks sketched at the Pan-European level can be evaluated in greater depth based on nationally specific scenarios. Flow-Based Methods



128




NTC – Conservative: Conservative assumptions have been made regarding cross-border transmission capacity relating to uncertainty in the commissioning dates of cross-border transmission projects and adequacy risks therein. These assumptions are within the scope of the methodology and report capturing all key features and risks relevant at the Pan-European level - this is a main objective for ENTSO-E. A conservative approach of ENTSO-E in providing transfer capacity values for adequacy analyses is sensible, even if adequacy risk might be overestimated. This will mean that the energy mix considered will not be compatible with the assumed/available transmission infrastructure. In this respect, the MAF results provide a signal in the form of ‘adequacy risk’. This could very well be higher than the desired adequacy level that a country is willing to accept or prepare for, expressed in the form of its adequacy standard. As a consequence of the fact that investments in grid infrastructure require a longer planning and decision horizon (~15 years) than the typical investment cycles for investments in new generation (3-5-7) years, the MAF exercise will detect risks and provide important feedback regarding the establishment of countermeasures by relevant stakeholders (e.g., member state authorities, policy makers, regulatory agencies, energy producers) in order to ensure the desired adequacy levels and the risks linked to their choices of generation mix and its evolution as well the impact of cross-border capacity availability. Furthermore, national or regional assessments might consider different assumptions that could be more optimistic regarding availability of cross-border transmission infrastructure and availability of imports to complete the assessment of the impact of such assumptions on adequacy levels. 15-Minute Flexibility: The purpose of the ENTSO-E target adequacy methodology being developed is the generation of a sound methodology that will foster performing regular Pan-European and regional diagnoses of the evolving European power system with respect to adequacy risks. Considering any modelling implies simplifications, so ENTSO-E adequacy target modelling focuses on hourly power balance modelling. Although these studies do not encompass every potential issue involved in the state of the future power system, ENTSO-E strongly believes that



129

this methodology will permit the development of a model able to capture all the key features and risks regarding adequacy for the Pan-European power system - this is a main objective for ENTSO-E. The effect of imbalances based on forced outages occurring close to real-time and the effect of forecasting errors of load, wind and solar production is accounted for in MAF by consideration of operational reserve requirements as an extra constraint while modelling. The detailed modelling of these aspects is outside of the scope of MAF as these imbalances are typically covered by the intra-day and primarily balancing markets and are not observed systematically in the day-ahead market during several days of a week. Flexibility: Residual load analysis, as considered in SO&AF2015, will be appraised for future MAF reports in order to inform them with respect to the need for flexibility necessary in power systems, typically within intra-day and balancing markets. One of the main goals is to be able to integrate 'need for flexibility' into the system. Flexibility in adequacy assessments is based on weather-dependent effects related to load variation, generation patterns of wind and solar power plants with a one-hour resolution and the consideration of the resources for flexibility. The existing PECD will be used for the adequacy assessments.

A framework to express national targets for SoS on the basis of commonly agreed adequacy standards at the regional or Pan-European level is necessary. ENTSO-E methodology provides the basis for such a framework. The adequacy levels should be nationally set but understood within a coordinated regional approach and based on the principles of subsidiarity regarding the definition of measures necessary to fulfil these standards by member states. Risks associated with these standards must be commonly understood across Europe through common definitions. Indeed, decisions to implement measures to ensure SoS at the national level is directly connected to the responsibility for SoS. Only the entity responsible for SoS can make the decision to implement or not measures to guarantee SoS. This however should be carried out in coordination with neighbouring member states and TSOs as a result of the cross-border impact that these decision might have on neighbouring areas. Role and Impact that the Results of the MAF Reports:

The MAF 2016 report features an ‘Overview Table’ on page 3 complemented by an appendix (Appendix 2) presenting the comments from each national TSO on the results in MAF 2016 in relation to each country’s own assessment of adequacy levels, its national adequacy standards and the measures taken to maintain them in the case of problems. The choice of LOLE = 1 hour as a threshold in such an overview table was chosen only because of technical reasons related to the presentation of the results. Only values of LOLE > 1 hour after averaging the results of the whole probabilistic Monte Carlo exercise were deemed significant enough to be highlighted, i.e.. the occurrence of adequacy problems across all the Monte Carlo ensemble of situations analysed was deemed significant rather than marginal. This choice of 1 hour should NOT be interpreted as a statement by ENTSO-E or its member TSOs on any proposals of harmonized SoS standards for EU of LoLE = 1 hour (!).

We know that such a choice of threshold for the previously referred to Overview Table could be nonetheless confusing and leads to misinterpretation. ENTSO-E will carefully consider the role and impact of the presentation of the results in future MAF reports. ENTSO-E is aware of the



130

impact that the MAF results might have on discussions by member states on regionally coordinated definitions of those standards.

ECG:ENTSO-E welcomes the ECG as a framework for TSOs, NRAs and member states, ACER, EC and ENTSO-E to interact at the relevant and necessary technical level regarding regional assessments of SoS and common definitions as well as in terms of discussion to ensure regional coordination regarding SoS measures by member states. ENTSO-E MAF adequacy reports provide an important input for discussions at the ECG. However, choosing to implement measures to ensure SoS at the national level are directly connected to the responsibility for SoS. Member states, as entities responsible for SoS, should make the decision to implement or not measures to guarantee SoS, and this should be carried out in coordination with neighbouring member states and TSOs as a consequence of the cross-border impact that such decisions may have. Considering the above, what additional methodological improvements should ENTSO-E consider for future MAFs? Please justify why through suggestions for methodological improvements COMMENTS BY STAKEHOLDERS

No Comment.


It is understandable that ENTSO-E does not have full visibility into availability/decommissioning/mothballing of power plants for the next 5 to 10 years; neither do the generators have such a view. In our opinion, requesting only owners of large (>100 MW) power plants or “system-relevant” generators to report their decommissioning/mothballing plans would be of limited value for the following reasons:

• Data collected would be unreliable: Recent major market shifts are hard to anticipate properly, as would be seen by referring to the renewables expansion leading to wholesale prices collapsing below marginal cost, the gas price decrease putting pressure on lignite operators, or the depressed price of carbon based on the excess of EU ETS allowances. Providing decommissioning/mothballing plans years ahead would therefore not make sense considering that changing economic conditions and regulatory interventions make such data very likely open to variation and therefore unreliable.

• Data collected would be incomplete: The trend towards more and more decentralised generation (wind, solar, small scale CHP, etc.) in the future should be duly noted as these assets may increasingly contribute to SoS. In general, the future developments for these technologies and the amount of installed capacity are also uncertain. ENTSO-E should




131

therefore obtain from DSOs all the relevant information on existing/expected decentralized capacity/generation connected to their distribution grids.

• Data collected are commercially sensitive information: In any case, such information is commercially sensitive and would therefore not easily be shared by market participants without strong guarantees on confidentiality rules being applied.

Instead, we would encourage TSOs and national regulatory authorities to closely follow future changes in generation capacity to have aligned national views on capacity development. It is also our understanding that estimations of decommissioning plants will be included in the national plans that member states submit to the EC as part of the Energy Union’s governance framework. However, we are convinced of the necessity to properly include elements of capacity uncertainty on the supply side in adequacy studies. In particular, we believe that ENTSO-E performs high-level economic sensitivity analysis to evaluate whether this expected capacity is likely to stay online during the MAF timeframe. To perform such sensitivities, we believe that the economic assessment should rather be based on a top-down approach looking at the economic position of ‘groups’ of power plants (e.g., technology, age, etc.) and based on assumptions linking the level of mothballing to economic viability rather than through bottom-up approach of looking at individual power plants. Such decisions may also be interdependent on other plants, so we contend it is not advisable to review each plant individually. It is also sensible to double check that each scenario is economically consistent with the “economic presence/survival” of the plants needed to ensure adequacy. Answer by ENTSO-E:

ENTSO-E is improving its data collection process regarding input data which affects the ‘likelihood of units to run and stay online’ within the market modelling assessments performed in TYNDP - MAF - these input data items are crucial to carrying any sensible sensitivity determination pertaining to ‘viability’ of the (central) best estimate scenarios collected from TSOs.

ENTSO-E and the TSOs know of the importance of these assumptions regarding the definition of the scenarios and ENTSO-E’s process for a common and consistent data collection has been revised to elevate the quality of the scenarios and possibly evaluate sensitivities surrounding them. Nonetheless, it cannot be 100% guaranteed that the forecasted generation mix considered by TSOs in their simulations will be economically viable in 2020 and 2025. For 2030 and 2040, such appraisal is even more difficult because of the intrinsic uncertainties of such a long-term horizon.

As part of the adequacy assessment in future MAF reports, ENTSO-E is debating whether to conduct a portion of sensitivity analysis with a more ‘conservative’ view of installed capacity. The sensitivities will attempt to assess the impact for adequacy of parts of the generation portfolio that may be at risk of being mothballed for economic reasons, like, for example, in the absence of a capacity market and/or unfavourable market conditions – ‘missing money’ problem. The first goal of ENTSO-E is to collect and set up a consistent (central) best-estimate Pan-European scenarios based on data collected from TSOs. Meaningful sensitivities are indeed defined around this central best-estimate Pan-European scenario as ‘stress test’ cases in MAF. It should be noted that the sensitivities relevant for ENTSO-E will always be relevant to and of the



132

scope of Pan-European scenarios. ENTSO-E welcomes the stakeholder feedback received - it supplies an important input for revising the elaboration of meaningful sensitivities.

Articulation of Pan-European, regional and national studies is hence a must as these complement, rather than compete, with each other. National studies must be consistent with the boundary conditions and assumptions provided by the larger perimeter Pan-European and regional studies.


In our view, adequacy assessment should primarily focus on defining the level of reliable/firm capacity that is necessary for ensuring SoS in the mid- to long-term. Given a predefined reliability standard and demand forecasts, the TSOs should conduct an assessment of how much reliable/firm capacity is present in the system (including cross-border capacity availability) to satisfy a certain level of SoS in the mid- to long-term versus the expected capacity. Making this « intermediate step » explicit would improve the quality and usefulness of the MAF.

As an initial step, it is crucial to define a consistent methodology for demand forecasts in order to obtain a more accurate/rigorous/coherent view on the level of demand that will need to be met (e.g., see our earlier suggestions for methodological improvements). If possible, it would be quite valuable to include sensitivities based on different reliability standards. This would foster a greater comprehension of how the desired level of adequacy influences the level of firm/reliable capacity needed.

This consistent methodology for the MAF report should also provide a common framework for national and regional adequacy studies (e.g., a consistent methodology for the demand forecasts, assumptions for constructing the scenarios to be analysed, etc.). As the European adequacy study shall build upon the national assessments according to Regulation 714/2009, there will be a definite consistency problem if such a common framework is not in place.

After this assessment, in order to evaluate whether this firm/reliable capacity will be available in the system and avoid relying on “ghost” capacity, ENTSO-E would be best to ultimately include in the next editions of the MAF reports high-level economic sensitivities based on different scenarios to evaluate whether this expected capacity is likely to stay online during the MAF timeframe.

This economic assessment should be based on the economic position of ‘groups’ of power plants (based on e.g., technology, age) and making assumptions on the level of mothballing/decommissioning linked to economic viability, i.e., a top-down approach examining aggregate availability/mothballing/decommissioning instead of a bottom-up approach reviewing individual power plants. As a matter of fact, providing a view on the economic viability of a





133

specific generation portfolio should be conducted by market participants themselves and not be the task of TSOs as they are regulated entities.

Decommissioning is not only a consequence of power plants reaching the end of their technical lifetimes. It can also happen for economic reasons because of the lack of sufficient income from the market, negating the ability to recover fixed costs on top of variable costs (i.e., low market prices or high costs). To define assumptions at the level of decommissioning, ENTSO-E could proceed as follows: based on assumptions for variable costs for each type of generation, the simulation conducted by ENTSO-E would show an expected number of running hours for each technology and country. Following an assessment of the possibility for these power plants to cover their fixed costs, ENTSO-E could then assume that N% of generation units running less than M hours (N and M would have to be appropriately defined to provide a realistic figure)on average could lead to decommission.

Special attention is worth being paid to the definition and inclusion of worst-case generation portfolio scenarios in generation adequacy assessments as well as to the assessment of their impact on SoS. In this sense, reintroducing into the next MAF versions a more “conservative scenario” (enhanced with an economic viability analysis) like in the last SO&AF would be most welcome.

In addition, it is important to keep in mind that decommissioning can also be brought about by an increase of environmental requirements that imply additional investment (i.e., Industrial Emission Directive), technical issues (i.e., excessive maintenance costs with old plants) or even industrial or political decisions (i.e., nuclear public acceptance).

To obtain a complementary view of the economic viability of groups of power plants, TSOs could take into consideration reports from consultants or investment banks and market reports on clear sparks or dark spreads. These elements could provide useful estimates of power plants’ economic viability in each market and help ENTSO-E develop different scenarios for power plant economic viability, and hence infer possible early decommissioning within MAF’s time horizon.




ENTSO-E welcomes greater transparency and availability from market participants with respect to the economic/technical data to improve the accuracy of the forecasts on economic/technical data used by TSOs in the simulations performed. It is very important to have superior visibility into decommissioning and mothballing figures in order to set up the relevant generation scenarios. Stakeholders stress the importance of ENTSO-E performing high-leveleconomicsensitivities.The MAF data collection process will be revised for MAF 2017 in order to improve visibility of these data provided by TSOs and processed to set up relevant sensitivities that will explore possibilities to consider different views on installed capacity, from ‘optimistic’ to ‘conservative’ situations.



134

ENTSO-E is revising its data-collection and scenario-building processes in order to consider relevant sensitivities based on data provided by TSOs on units at risk of being mothballed for economic reasons and information regarding decommissioning of units strictly for technical reasons, or/and shut downs based on legal/policy factors within the storyline assumptions of the scenarios considered. It must be noted that the conclusions of the MAF report cannot be separated from the hypotheses used and can only be read in reference to them. The hypotheses are gathered by the TSOs according to their best knowledge at the time of the data collection and validated by ENTSO-E’s relevant committees. Therefore, MAF reports and sensitivities do not intend to contain all the information that a prospective investor or market participant may need. ENTSO-E emphasises that ENTSO-E and the TSOs involved in these studies cannot be responsible in the case hypotheses from these reports or the estimations based on these hypotheses are not realised in the future. Providing views on the economic viability for specific generation technologies linked to decisions by prospective investors or market participants should be the onus of market participants themselves and not by TSOs as regulated entities.


Objectivity and fact-based analyses are significant when assessing the future capacity adequacy situation, and the MAF currently fulfils these high-level criteria, though further improvements are needed. The current results show the capacity adequacy situation without taking into account the implementation of future capacity mechanisms by member states. On the one hand, this approach enables the MAF to serve as “justification” for certain member states to deploy a capacity mechanism because it demonstrates the capacity adequacy risks they would otherwise experience (e.g., in the UK and France) in the future. On the other hand, this approach does not accurately reflect the adequacy situation for 2025. The capacity mechanisms in place for that time horizon will trigger, if needed, certain investments (new, re-powering, maintenance) to ensure the targeted level of system adequacy and to solve the capacity adequacy issue.

Adding a scenario for the ‘planned policy’ situation (e.g., decided and planned implementation of capacity mechanisms in certain member states) could help identify the effects of the capacity mechanisms in the member states implementing them. The planned policy scenario should not





135

assume capacity mechanisms in member states that do not plan to deploy these options or where the planning is still under discussion. However, it is worth mentioning that such an approach would only shed additional light via the analysis if the economic viability was adequately considered in both scenarios, otherwise the existence or not of capacity mechanisms does not make a difference. The scenario is also highly dependent on (economic) assumptions surrounding capacity mechanisms, both in cases of the centralised or decentralized model.

To assess the economic viability of the MAF, as mentioned in our answer to the previous question, EURELECTRIC would rather advise developing sensitivity scenarios to take into account the uncertainties on the supply side within the MAF timeframe. Such scenarios should represent different levels of stress in electricity systems that can be influenced by many factors (e.g., fuel prices, macroeconomic conditions, etc.) and not only by the evolution of the regulatory framework. Such sensitivity analyses would be more appropriate for permitting market players and public authorities to appreciate the risk related to the absence of regulatory interventions (e.g., the introduction of capacity mechanisms, etc.) aimed at ensuring the evolution of the generation fleet necessary to meet the SoS targets defined at the national level.




ENTSO-E welcomes greater transparency and availability from market participants with respect to the economic/technical data to improve the accuracy of the forecasts on economic/technical data used by TSOs in the simulations performed. It is very important to have superior visibility into decommissioning and mothballing figures in order to set up the relevant generation scenarios. The MAF data collection process will be revised for MAF 2017 in order to improve visibility of these data provided by TSOs and processed to set up relevant sensitivities that will explore possibilities to consider different views on installed capacity, from ‘optimistic’ to ‘conservative’ situations. The sensitivities planned are in line with the feedback provided by the stakeholder. Note that the aforementioned sensitivities, through consideration of different scenarios that can be elaborated by using different assumptions on the capacity, still need to fulfil the requirements of consistent Pan-European-wide scenarios relevant to the Pan-European studies of the MAF. Economic viability appraisal is therefore linked to the construction of Pan-European consistent scenarios/sensitivities, which implicitly means that not all nationally relevant sensitivities can be captured within consistent Pan-EU scenarios. Detailed national specificities might need to be considered in depth by additional national and regional studies,whichinturnmustbeconsistentandcomplementthePan-Europeanscenarios.



136


No Comment.


Documents

Consultation Feedback Report - Microsoft · • VATTENFALL • ENEL S.p.A. • REGULATORY ASSISTANCE PROJECT • EURELECTRIC The final MAF 2016 report, after taking into account stakeholder