3. Forward-looking risk assessments
3.2 Generation adequacy
Generation adequacy refers to the system's ability to procure
0
Eastern Denmark Western Denmark
2037
FIGURE 17: EXPECTED NOMINAL IMPORT CAPACITY OF DENMARK'S INTERNATIONAL CONNECTIONS.
Source: Energinet's analysis assumptions 2017.
GOALS FOR GENERATION ADEQUACY
• EUE (Expected Unserved Ener-gy) indicates unserved energy per year, including the risk of blackout, i.e. the total energy consumption need which cannot be met by production.
• Weighted minutes indicates the number of hours with a power shortage irrespective of the extent of the shortage.
• LOLE (Loss Of Load Expecta-tion) indicates the number of hours with a power shortage irrespective of the extent of the shortage.
• LOLP (Loss Of Load Probabi- lity) indicates the probability of a power shortage in a given hour (LOLE=8760 x LOLP).
risk in the Danish electricity system. The results are cal- culated separately for Eastern and Western Denmark.
3.2.1 Generation adequacy analyses
Generation adequacy analyses in Denmark towards 2025 show that Eastern Denmark faces the greatest risk of pow-er shortages. This can be ascribed, among othpow-er things, to increased domestic production capacity as well as more high-capacity international connections in the Western Denmark system than in the Eastern Denmark system.
With Energinet's analysis assumptions, all the analysis re- sults for Western Denmark show a risk of less than one output minute per year, thus complying with Energinet's generation adequacy objective. Therefore, only results for Eastern Denmark are presented in the following.
Basic results, where maintenance and outage probabilities are based on historical data, show that the risk of power shortages in Eastern Denmark generally increase in the long term. The power situation will improve in 2019 fol- lowing the commissioning of the Kriegers Flak offshore wind farm. The seven calculated minutes equates that the average consumer is expected to have 99.9987% of the demanded energy supplied, while 33 minutes correspond to 99.9937%.
The overall more strained power balance owes to changed expectations in the analysis assumptions. Firstly, less wind capacity than previously is forecast for Eastern Denmark, TABLE 5: GENERATION ADEQUACY ASSESSMENT IN EA-STERN DENMARK. BASIC RESULTS.
EASTERN DENMARK 2018-2025
2025 1,002 33 1.8 0.02
FSI MODEL
The Forsyningssikkerhedsindeks (FSI) model is Energinet's tool for analysing expected future generation adequacy situations. The model is stochastic and simulates incidents in the electri-city system which can lead to power shortages on an hourly basis.
A number of calculation runs (typically 300) are performed for each year being analysed, to re- present possible incident combinations. Thus the model results illustrate an average of all calcula- tion runs for a single year.
The FSI model is based on historical time series on an hourly basis for consumption and fluctua-ting production (wind and solar power). Production from thermal power plants and imports via inter- national connections are stochastic. The stocha-stic element is represented using probabilities of breakdowns or maintenance. Thermal generation plants and international connections will therefore be unable to supply energy to meet the consump- tion needs for a number of hours in each calculati-on run. Which and for how many hours are deci- ded at random. During the hours when large ge- neration plants and/or international connections drop out, consumption needs must be met by fluctuating production from wind and solar power or the remaining thermal plants and international connections.
The model estimates the risk of outages in the Danish system due to power shortages. Since the model operates on an hourly basis, variations within each hour of delivery are not covered. The FSI model is best suited to perform calculations on the basis of assumptions resembling the cur- rent situation, as it is based on historical time se- ries and without the derived effects of shortage situations, such as flexible consumption. The FSI model will tend to overestimate the risk of power shortages in a future electricity system which is significantly different from today's system in terms of both physics, the market and internatio-nal operatiointernatio-nal cooperation (e.g. in 2025).
Note that the expected number of hours with power shortages in the model simulations is low, indicating that power shortage situations are rare incidents, also in DK2. Even though the results are subject to uncertainty, they indicate that a few brownout situations should be expected over the next 10 years.
The risk of brownouts may be reduced by, for example, maintenance planning, market solutions (e.g. strengthened efforts to ensure flexible consumption), a new production capacity framework and the establishment of new infra-structure for new areas.
More specifically, a project has been initiated to assess whether an additi-onal electricity connection to Western Denmark is a socio-economically attra-ctive solution for strengthening the power balance in Eastern Denmark.
3.2.2 Generation adequacy sensitivity Risk assessments of the future power situation are highly impacted by input data, which is illustrated in the follow-ing sensitivity analyses. The various parameter variations are examples and do not reflect Energinet's assessment of the input parameters.
Availability of international connec-tions
As Denmark is well-connected to the electricity grids in our neighbouring countries, the availability of these con-nections is important when assessing generation adequacy. The availability of international connections directly impacts import options – no cable, no energy.
The results show that reduced availa- bility of international connections in- creases the risk of power shortages significantly. Doubling the outage time on all international connections sig- nificantly increases the risk of power as well as declining thermal capacity. Secondly, electricity
consumption is higher in this year's analysis assumptions.
As from 2020, the risk of power shortages therefore exceeds Energinet's 5-minute target.
TABLE 6: SENSITIVITY RESULTS FOR AVAILABILITY OF INTERNATIONAL CONNECTIONS IN EASTERN DENMARK.
AVAILABILITY OF INTERNATIONAL CONNECTIONS EUE
doubled 1,442 51 3.3 0.04
2020 - outage
halved 119 4 0.2 0.00
TABLE 7: SENSITIVITY RESULTS FOR DATA CENTRE CAPACITY IN EASTERN DENMARK.
DATA CENTRE CAPACITY
data center capacity in both DK1 and DK2
542 18 1.1 0.01
2025 – 100 MW extra data center capacity in both DK1 and DK2
1,725 53 3.1 0.04
2025 – Data center capacity doubled in DK1 relative to ana-lysis assumptions, no data centers in DK2.
1,051 34 1.9 0.02
energy consumers with expected high continuous power usage. As a result, the establishment of additional data centres, in addition to already known or similar energy consumers, will pos- sibly impact general generation ade-quacy in the Danish electricity system to a great extent.
Results show that data centres located in Eastern Denmark will increase the risk of power shortages in the Eastern Denmark system. The location of se-veral data centres in Western Denmark alone will only have a marginal impact on the power situation in Eastern Denmark via the Great Belt Power Link.
A doubling of the data centre capacity in Western Denmark in 2025, for exam- ple, will not significantly increase the risk of power shortages in Eastern Denmark. As the FSI model does not take local conditions into account, there may still be local challenges in conne-ction with the commissioning of large energy consumers.
Availability of central power plants The availability of thermal production capacity is material to risk assess-ments of the power situation.
Additional power plant capacity of 200 MW in Eastern Denmark will probably reduce the risks in 2018 and 2020 to a level within Energinet's objective as well as more than halving the risk in 2025. In contrast, reduced availability of power plants, for example via a shortages in Eastern Denmark. This means that the risk
of outages increases significantly in years with extra-long outage times due to maintenance work or breakdowns.
Conversely, the risk also decreases in years without long-term maintenance work or breakdowns.
Increased data centre expansion
Interest in locating large data centres in Denmark has been high in recent years. Data centres are significant TABLE 8: SENSITIVITY RESULTS FOR POWER PLANT CA-PACITY/AVAILABILITY IN EASTERN DENMARK.
POWER PLANT CAPACITY/AVAILABILITY EUE
(MWh/
year)
Weighted minutes (min./year)
LOLE (hours/
year) LOLP
(%) 2018 – 200 MW extra
power plant capacity 46 2 0.1 0.00
2019 – 200 MW extra
power plant capacity 47 2 0.1 0.00
2020 – 200 MW extra
power plant capacity 116 4 0.2 0.00
2025 – 200 MW extra
power plant capacity 452 15 0.8 0.01 2020 – Outage
proba-bility for central power plants doubled
613 22 1.4 0.02
Outage probabilities for HVAC connec- tions will be included in the 2017 edition. The work on MAF 2017 is still ongoing, and the first draft is expected to be published in the summer of 2017.
The results of Energinet's market si- mulations for MAF 2016 included a base case and one sensitivity analysis.
In the base case, reserves and outage probabilities for international connecti-ons are not included in the calculaticonnecti-ons.
Reserves and outage probabilities for HVDC connections have been added in the sensitivity analysis. The sensitivity analysis focuses on 2020.
The results from MAF 2016 do not paint quite the same picture of a strained po- wer situation in the Danish system as the FSI model. This may partly be at- tributable to differences in methodology and input parameters (e.g. outage times on international connections), but may also indicate that the FSI assumptions about international data and internati- onal availability are too pessimistic.
In the long term, Energinet expects to replace the FSI model with the Europe-an model.
doubling of the outage probability at central power plants, will somewhat double the risk of power shortages in 2020.
3.2.3 Common Nordic and European generation adequacy assessments
Energinet cooperates with other European and Nordic TSOs on common generation adequacy assessments. The method used is based on probabilistic principles.
The method is suitable for evaluating how international con- nections contribute to generation adequacy and add other types of risk.
In 2016, ENTSO-E published its Mid-Term Adequacy Forecast (MAF) for the first time. The purpose of the MAF is to assess the risk of power shortages in the European price areas in the medium term (up to 10 years ahead). The assessment is made on the basis of a common European analysis, where four different market models are simulated based on the same data set.
In the first edition of the MAF, the method did not yet include outage probabilities for HVAC connections (such as the Øresund cables from Zealand to Sweden), and the results in Table 10 therefore underestimate the generation adequacy risk for Denmark.
TABLE 9 : RESULTS OF THE GENERATION ADEQUACY AS-SESSMENT IN THE MAF FOR EASTERN DENMARK 2020.
EUE
(MWh/year) LOLE
(hours/year) LOLP (%) Eastern Denmark 2020 & 2025 (base case)
2020 0 0 0.00
2025 8 0.37 0.00
Eastern Denmark 2020 (sensitivity)
2020 0 0 0.00
TABLE 10: SHARE OF HOURS WITH CONSUMPTION EXCEEDING PRODUC-TION CAPACITY IN DENMARK.
SHARE OF HOURS IN ONE YEAR (%)
2018 7.7
2019 9.7
2020 18.3
2025 30.2
Since then, the European Commission has carried out a sector inquiry into capacity mechanisms. Based on this, Energinet still finds that it is not pos- sible to make a call for tenders for a strategic reserve in the 2016-2018 period. Energinet will reduce the risk of disconnecting consumers in the event of shortages in Eastern Denmark through improved coordination of maintenance work on international connections, power plants and own installations as well as by rewarding shorter outage times in connection with construction projects and using shorter start-up warning times at po- wer plants. In addition, Energinet has ensured a shorter start-up warning time at a single power plant.