Monitoring report on the security of supply on the Belgian natural gas market. Year (based on article 5 of the directive 2009/73/EC)

Transcription

1 Monitoring report on the security of supply on the Belgian natural gas market Year 2011 (based on article 5 of the directive 2009/73/EC) Member States are asked under the Directive 2009/73/EC of 13 July 2009 concerning common rules for the internal market in natural gas and repealing Directive 2003/55/EC to provide each year a monitoring report on the natural gas security of supply issues. This report should contain at least the following aspects: 1. the balance of supply and demand on the national market, 2. the level of expected future demand and available supplies, 3. envisaged additional capacity being planned or under construction, 4. the quality and level of maintenance of the networks, 5. measures to cover peak demand and to deal with shortfalls of one or more suppliers. The competent authorities shall publish, by 31 July each year, a report outlining the findings resulting from the monitoring of those issues, as well as any measures taken or envisaged to address them and shall forward that report to the Commission forthwith. The information in the report below gives an overview of the required information. The report is based on 2011 data and was established in cooperation with the Federal Planning Bureau and in consultation with the Commission for Regulation of Electricity and Gas (CREG) and Fluxys Belgium 1

2 Table of Contents 1. The balance of supply and demand on the national natural gas market Natural gas demand in Evolution of the total natural gas demand Breakdown of natural gas consumption per consumer category Relative share of the natural gas consumption per consumer category Seasonality in the natural gas demand Natural gas supply in Natural gas imports Wholesale and retail market Natural gas in storage Level of expected future demand and available supplies Evolution of the peak demand on the distribution network Evolution of the daily average peak demand on the L-gas distribution network Evolution of the daily average peak demand of the H-gas distribution network Natural gas demand projections up to Projections of the hourly daily average L-gas demand and entry capacity Projections of the daily average H-gas demand and entry capacity Projections of the H-gas demand at the hourly peak and the available entry capacity Natural gas infrastructure Existing and new infrastructure Quality and level of maintenance of the networks Measures to cover peak demand and to deal with shortfalls of one or more suppliers

3 1. The balance of supply and demand on the national natural gas market 1.1. Natural gas demand in 2011 The natural gas demand needs to be looked at on a yearly, monthly or daily basis as those three options provide a different insight in de needed volumes to cover the total demand in one year, in a specific month, but also on the flexibility needs of the network to cope with fluctuations of the natural gas demand during a day. Belgium consumes two different types of natural gas, namely H-gas (with high caloric value) and L-gas (with low caloric value). L-gas was mainly originating from the Groningen field in the Netherlands (Groningen gas) but now more and more L-gas is obtained by ballasting H-gas (adding nitrogen to obtain L-gas). Those two different types of natural gas request a separate network. A distinction will be made between the figures for L-gas and H-gas in order to give a reliable overview of the Belgian natural gas market Evolution of the total natural gas demand The natural gas demand can be expressed in the actually measured consumption on the transmission network and the consumption normalised to correct for the dependency of the consumption with the weather conditions 1. The normalised consumption sets out the amount of natural gas that would be consumed in a specific year provided standard winter condition occur. These standard conditions are calculated as an average of the temperatures over a period of 30 years. The demand figures are normalised for the entire distribution network to get a better indication of the annual growth rates of the natural gas demand in the distribution network as the consumption there is very dependent on the outside equivalent temperature and to be able to extract the demand for a winter with peak conditions. The graphs below show the evolution of the total measured Belgian natural gas consumption (L-gas, H-gas and combined) for the period (in GWh/year) (figure 1). Figure 2 gives the evolution of the total Belgian consumption after correction based on normalised temperature profile. In 2011, total measured natural gas demand of the Belgian end consumers amounted to TWh, of which TWh for H-gas (74.4% of total demand) and 47.0 TWh for L-gas (25.6% of total demand). This is a decrease of the natural gas demand of 14.8% compared to 2010 (215.3 TWh). The reason for this is that 2010 was an exceptionally cold year which explains the spike in consumption for that year (2 700 Degree-Day in 2010 compared with 1928 Degree-Day in 2011). The year 2011 on the other hand was characterised by a very mild winter which resulted in a much lower natural gas consumption. Also the natural gas consumption by the electric utilities decreased which resulted in a lower total natural gas consumption. 1 Weather normalization is the process by which the energy use from one year is adjusted to account for specific weather conditions. Through this procedure, the energy in a given year is adjusted to express the energy that would have been consumed under 30-year average weather conditions. 3

4 FIGURE 1: EVOLUTION OF THE TOTAL MEASURED NATURAL GAS CONSUMPTION (IN GWH) Figure 2 shows the total natural gas consumption corrected with a normalised temperature profile. Under normalised average winter conditions, the demand for L-gas would have been fairly stable in the last 10 years (54 TWh in 2011 compared with 52 TWh in 2001). For H-gas, an increase in the natural gas demand is noticeable (144 TWh in 2011 compared with 120 TWh in 2001). This is mainly due to the increasing number of connections to the natural gas grid and the higher consumption by electricity plants. FIGURE 2: EVOLUTION OF THE TOTAL NORMALISED NATURAL GAS CONSUMPTION (IN GWH) 4

5 Breakdown of natural gas consumption per consumer category Total natural gas consumption can mainly be attributed to three categories of consumers : the consumers connected to the distribution network which consist mainly of households and small to medium enterprises and industrial consumers (TD), the large industrial consumers connected to the gas transport network (TI) and the electric utilities (TE). The chapters below provide a more detailed overview of the evolution per consumer category where we focus on the evolution of the natural gas demand over the last ten years and on the shares of the three consumer categories in the total natural gas consumption Evolution of the yearly natural gas consumption on the distribution network A large part of the natural gas consumption on the distribution network (TD) stems from households that use natural gas mainly for heating. Each winter period provides an updated correlation between the equivalent temperature and the measured natural gas consumption. Below 16.5 C, the natural gas consumption is mainly influenced by the outside equivalent temperature, which can be converted into a certain number of Degree Days (DD) 2. For temperatures above 16.5 C, we assume that the consumption is independent from the outside temperature. This is referred to as the base consumption that corresponds to the natural gas needs for hot water boilers and cooking, and the natural gas demand stemming from small and medium enterprises connected to the distribution grid. The figure below shows the correlation between the natural gas consumption and the DD. 2 The equivalent Degree Days (DDeq) gives an overview of the average profile of the need for heating of a dwelling in Belgium. For a given day, the DD are calculated through the difference between 16,5 C and the equivalent temperature that day. The equivalent temperature is defined as : Teq(n)= 60%*Taverg(n)+30%*Taverg(n- 1)+10%*Taverag(n-2) where T(n),T(n-1) and T(n-2) are the average day temperatures measured by the Royal Meteorological Institute in Uccle the day n, n-1 and n-2. If the average daily temperature is higher than 16,5 C during 3 consecutive days, the DD is equal to 0. 5

6 FIGURE 3: CORRELATION BETWEEN THE NATURAL GAS CONSUMPTION AND THE DEGREE DAYS (DD) Correlation between Degree-Day and Household Gas Consumption (Total) Household gas consumption (GWh/d) Degree-Day aug/11 sep/11 okt/11 nov/11 dec/11 jan/12 feb/12 mrt/12 apr/12 mei/12 jun/12 jul/12 Gas Consumpt. (GWh/d) Degree-Day In order to understand the evolution of the natural gas consumption on the distribution network, the evolution of the equivalent Degree Days (DDeq) 3 in the period needs to be analysed. From the data, it is clear that the years 2007 (1 963 DDeq) and 2011 (1 927 DDeq) noted a lower number of DD, which means that those were relatively warmer years. The year 2010 on the other hand counted a very high number of DD (2 700 DDeq), meaning it was an exceptionally cold year. This will be the cause of a very high natural gas consumption on the distribution network in that particular year. 3 To take into account the thermic inertia of buildings in order to better estimate the real heating needs, the equivalent DDeq are calculated by taking into account the DD of the 2 previous days according to the following formula: DDeq = 0.6 x DD of the day x DD of D x DD of D-2 6

7 FIGURE 4: EVOLUTION OF THE EQUIVALENT DEGREE DAYS (DDEQ) The measured consumption on the distribution network is fairly volatile and in close correlation with the number of DDeq. The normalised consumption profile is more stable and gives a better indication of the market penetration rate of the natural gas in the last ten years. We obtain the normalised consumption by calculating the yearly natural gas consumption that correlates to a normalised temperature profile that is identical for each year. The variations in the yearly natural gas consumption that we record then are no longer influenced by the meteorological conditions in those years, but by other variables that influence the natural gas consumption, like the growth rate of the natural gas consumption by the consumer categories on the distribution network. We also use the notion of equivalent temperature in order to estimate the capacity needs of the distribution network to ensure the security of supply in winter peak conditions. The infrastructure has to be able to transport all the natural gas that could be consumed at a winter peak day that occurs with a statistical probability of once in 20 years. In Belgium the reference value for the winter peak occurring once in 20 years was set at -11 Ceq or 27.5 equivalent Degree Days in the risk assessment 4. Based on the daily measured consumptions of a given winter period, we can deduct the existing linear relation between the equivalent temperature and the daily consumption. Based on this correlation, we can extrapolate the consumption to - 11 Ceq in order to deduce the amount of natural gas that the distribution network will consume at -11 Ceq with a risk of 50%/50%. This method is used in point 2.1. Evolution of the peak demand to determine the natural gas imports and the transmission capacity needed to satisfy the consumption on the distribution network at -11 Ceq. Note that the maximum hourly consumption during a peak day could be about 20% higher than the average hourly consumption that day. This implies that a certain amount of flexibility in the natural gas network is 4 At this moment, there is not yet a legal framework to use -11 Ceq as the reference value to determine the reference winter peak. The Energy Administration is however envisaging to fix this by law as explained in chapter 5 of this document in order to have a clear reference for the dimensioning of the reference peak gas demand. 7

8 necessary to be able to balance this swing in peak demand during the day. This flexibility should be either imported, covered by the linepack within the network or provided through intraday gas trading. The figures below split out the consumption on the distribution network for L-gas and for H-gas. The consumption on the L-gas distribution network lies in the same order of magnitude as for H-gas and follows the same trend. This is normal given that the customers on the H- and L-gas network will adopt the same behaviour (i.e. consume more the cold years and less the warmer years). In 2011, the measured consumption on the H-gas distribution network reached GWh and GWh on the L-gas network. In 2010, the measured consumption was still GWh for H-gas and for L-gas or a decrease with respectively 17.72% and 19.39%. This decrease is only due to the different temperature conditions in those years. The normalised consumption figures for the distribution network on the other hand show an increase in the natural gas consumption between 2010 and The normalised natural gas consumption in 2010 reached GWh and GWh in 2011, or an increase by 4.82%. For L-gas, the normalised consumption on the distribution network stood at GWh in 2010 and grew by 3.98% to in The higher consumption on the distribution network is mainly due to the higher number of connections to the natural gas distribution grid 5. FIGURE 5: EVOLUTION OF THE CONSUMPTION OF H-GAS ON THE DISTRIBUTION NETWORK ( ) 5 In Belgium, one region has an objective of connectivity to the gas distribution grid. The Flemish region has imposed by decree to the distribution grid operator to reach a connectivity of 95% in 2015 and 99% in In 2011, the Flemish region reached a connectivity of 58.1%. 8

9 FIGURE 6: EVOLUTION OF THE CONSUMPTION OF L-GAS ON THE DISTRIBUTION NETWORK ( ) The T extreme curve in the figure above shows the calculated annual consumption for an extreme temperature profile as obtained in the year 1963 with 2964 DD. This curve gives an estimation of the energy volume that would be consumed on the distribution network in a given year. The extreme energy volume follows the same evolution as the normalised temperature profile and is about 20% higher than the normalised temperature profile. We can therefor assume that in an extreme winter condition, the suppliers on the distribution network would have to be able to increase the energy volumes significantly in order to cover the swing in the intraday demand. The evolution over the last ten years of the total natural gas consumption on the distribution network for L- gas and H-gas combined is displayed in figure 7. The measured consumption on the distribution network in total amounted to GWh in For a normalised temperature profile, the 2011 figure is GWh. The normalised consumption shows in general an increasing trend over the last ten years, except for the period between in which the natural gas consumption on the distribution network was actually decreasing. The reason for the stabilisation in the natural gas demand on the distribution network could be twofold : on the one hand a more rational energy consumption in the distribution network related to investments in energy efficiency, and on the other hand to a slowdown of the economic activities of the industrial consumers. This is even more remarkable as we notice an increase in the increasing number of connections to the natural gas grid (2.4% in the period ). In 2011, the retake in the economic activity and the increase in the number of connections to the natural gas grid on the distribution network make the natural gas consumption go up to slightly over 2007 levels ( TWh in 2011 with regard to TWh in 2007). 9

10 FIGURE 7: EVOLUTION OF THE TOTAL MEASURED CONSUMPTION ON THE DISTRIBUTION NETWORK (L+H COMBINED) ( ) Evolution of the yearly natural gas consumption by the large industrial consumers connected to the transmission network Natural gas consumption by large industrial consumers connected to the transmission network is less influenced by the outside temperature, so it is sufficient to only look at the measured consumption to assess the trend of the past years. The natural gas consumption by the large industrial consumers is more or less stable over the last ten years, with an exception for the year That year was characterized by the financial crisis which caused the natural gas demand of the large industrial consumers to decrease very sharply. This sharp decrease is noticeable in both the H-gas and L-gas network. In 2010, the natural gas consumption by the large industrial consumers went back to previous levels in line with 2008 consumption levels. This trend continued in On the L-gas network, the natural gas consumption amounted to GWh in 2011, coming from GWh in 2009 and decreasing slightly from the 2010 level of GWh. The consumption on the H-gas network increased from the strong decline to GWh in the crisis year 2009 back to previous levels ( GWh) in 2010 and reached GWh in

11 FIGURE 8: TOTAL YEARLY NATURAL GAS CONSUMPTION OF THE LARGE INDUSTRIAL CONSUMERS ( ) Evolution of the yearly natural gas consumption by the electric utilities In the L-gas network, natural gas consumption by electric utilites has been fairly limited in the last 10 years and has quasi disappeared since 2010 as, except from one small CHP plant, no more power plants are connected to the L-gas grid. The reason for this is that there is no longer an interest from the larger market players (power plants and large industrial consumers) to be connected to the L-gas grid and that the TSO always tries to connect similar clients to the H-gas network. In addition, the Belgian government has set up a task force in cooperation with TSO, DSO s, regulator and other relevant players on the gas market to study a possible conversion of certain consumers on the L-gas network to the H-gas network. We do note that at this moment, there is however no legal framework that regulates new connections to the L-gas grid. On the H-gas network, natural gas demand stemming from the electric utilities indicates an increasing trend over the period The electricity production in the years 2009 ( GWh) and 2010 ( GWh) was exceptionally high. A large part of that electricity production was destined for exportation. The large increase in 2009 which continued in 2010 could be explained by the shift from the natural gas supply surplus from the large industrial consumers that was taken up by the electric utilities. In 2011 the natural gas consumption by electric utilities fell back to previous levels of GWh. 11

12 FIGURE 9: TOTAL YEARLY NATURAL GAS CONSUMPTION BY THE ELECTRIC UTILITIES ( ) Relative share of the natural gas consumption per consumer category The breakdown of the total natural gas consumption for H-gas and L-gas emphasizes the relative importance of the natural gas demand per consumer category (for normalised temperatures). It is clear that the main source of demand on the L-gas network stems from the consumers connected to the distribution network (TD). By comparing the data for 2011 and 2008, it is clear again that there is barely any electricity production left on the L-gas network. In 2008, electricity production still represented 5% of the total normalised L-gas consumption. About 86% of the L-gas consumption is destined for consumers connected to the distribution network, which makes the consumption on the L-gas network particularly vulnerable to changes in the outside temperature. The consumption of the large industrial consumers connected to the L- gas transmission network remains stable. FIGURE 10: NATURAL GAS CONSUMPTION PER CONSUMER CATEGORY ON THE L-GAS NETWORK IN 2008 AND 2011 (GWH) Natural gas consumption 2011 (L) (GWh) Natural gas consumption 2008 (L) (GWh) % 23 0% TD L TI L TE L % % TD L TI L TE L % % On the H-gas network, natural gas demand is more or less evenly distributed between the three consumer categories: the consumers connected to the distribution network (TD) account for about 35% of the natural 12

13 gas demand after temperature normalisation, while the large industrial consumers(ti) and the electric utilities (TE) take up respectively 27% and 38%. The same situation was perceived in FIGURE 11: NATURAL GAS CONSUMPTION PER CONSUMER CATEGORY ON THE H-GAS NETWORK IN 2008 AND 2011 (GWH) Natural gas consumption 2011 (H) norm. (GWh) Natural gas consumption 2008 (H) norm. (GWh) % % TD TI TE % % TD TI TE % % The figures below demonstrate the relative importance of the distribution network in the total natural gas consumption. The distribution takes up almost half (49%) of the total yearly natural gas consumption. The large industrial consumers account for 24% of the total natural gas consumption while the remaining 27% are consumed by the electric utilities. FIGURE 12: NATURAL GAS CONSUMPTION PER CONSUMER CATEGORY IN TOTAL IN 2008 AND 2011 (GWH) Natural gas consumption 2011 (L+H) norm. (GWh) Natural gas consumption 2008 (L+H) norm. (GWh) % % TD TI TE % % TD TI TE % % Seasonality in the natural gas demand As the distribution network takes up about half of the total natural gas demand, and as the natural gas demand on the distribution network is linked to the outside temperature, a high degree of seasonality can be noticed in the natural gas consumption. Consumption on the distribution network for H-gas as well as for L-gas can increase by approximately factor eight between the summer and the winter consumption. The monthly demand pattern is quite stable over the different years. 13

14 FIGURE 13: MONTHLY NATURAL GAS DEMAND ON THE L-GAS DISTRIBUTION NETWORK AND H-GAS DISTRIBUTION NETWORK ( ) Monthly consumption TD Lnet meas. Monthly consumption TD Hnet meas GWh TD (2004) TD (2005) TD (2006) TD (2007) TD (2008) TD (2009) TD (2010) TD (2011) GWh TD (2004) H TD (2005) H TD (2006) H TD (2007) H TD (2008) H TD (2009) H TD (2010) H TD (2011) H Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec The total natural gas consumption on the distribution network on the H-gas and the L-gas distribution network in the months July and August fluctuates in general around 2 TWh and increases to about 13 TWh in the months December to January/February. This strong seasonal swing indicates again that not only the natural gas volumes are important to secure the supply, but also the seasonal flexibility in the network to be able to cope with the strongly fluctuating demand. Although for the large industrial consumers, a strong decrease in the gas consumption for the year 2009 can be noticed, there is no visible decrease of the gas consumption on the distribution network for that year. FIGURE 14: TOTAL (L + H) MONTHLY NATURAL GAS DEMAND BY THE DISTRIBUTION NETWORK ( ) Monthly consumption TD L+H meas. GWh TD L+H (2004) TD L+H (2005) TD L+H (2006) TD L+H (2007) TD L+H (2008) TD L+H (2009) TD L+H (2010) TD L+H (2011) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec For the large industrial consumers, there is hardly a seasonal swing noticeable in the monthly natural gas demand. This is because the industrial processes are not dependent on the outside temperature. There is only a very minor correlation with the outside temperature which is caused by the heating of the industrial buildings. In the monthly figures, the impact of the crisis year 2009 becomes obvious. The consumption of the large industrial consumers in 2009 per month was generally 20% lower on average than in the previous years. 14

15 FIGURE 15: EVOLUTION OF MONTHLY NATURAL GAS CONSUMPTION BY LARGE INDUSTRIAL CONSUMERS ( ) Monthly consumption TI H+Lnet meas GWh TI (2004) TI (2005) TI (2006) TI (2007) TI (2008) TI (2009) TI (2010) TI (2011) TI (2012) jan feb mar apr may jun jul aug sept oct nov dec The natural gas demand of the electric utilities is hardly linked with the outside temperature either. The significantly higher consumption by the electric utilities in the middle of 2009 and the beginning of 2010 can be noticed in the monthly data. A large part of this extra electricity production was exported to France, as the French households mainly use electricity for heating. As the flexibility in the power production in France is more limited due the composition of the production park (mainly nuclear), part of the flexibility of the Belgian electricity production is imported into France. Therefore we can state that the gas demand from the electric utilities in Belgium is to a limited extend correlated to the outside temperature. FIGURE 16: EVOLUTION OF MONTHLY NATURAL GAS CONSUMPTION OF THE ELECTRIC UTILITIES ( ) Monthly consumption TE H+Lnet meas GWh TE (2004) TE (2005) TE (2006) TE (2007) TE (2008) TE (2009) TE (2010) TE (2011) TE (2012) 0 jan feb mar apr may jun jul aug sept oct nov dec 15

16 1.2. Natural gas supply in Natural gas imports Belgium has a well-diversified natural gas portfolio and supply network for H-gas that allows tapping into alternative supplies in case of a supply disruption. This is also shown in figure 17 below. It is however necessary to keep in mind supplies can only be re-routed if the suppliers have access to upstream capacity to enter the gas in Belgium through another entry point or to make use of the secondary capacity market. The supplies of L-gas originate mainly from the Netherlands. As the Dutch system no longer makes a difference in the gas quality and the network transmission operator converts H-gas into L-gas by ballasting it (if necessary), the availability of L-gas on the market is becoming less an issue. Natural gas can enter the country through a series of entry points on the natural gas transmission network. In 2011, a new entry point at Zelzate at the border with the Netherlands came into operation. LNG supplies, mainly from Qatar via the Zeebrugge terminal, accounted for a share of 7.5% of Belgian natural gas consumption in Zeebrugge is the main gateway to the Belgian market with a total market share of 41.20%. The new Zelzate entry point immediately accounted for 2.6% of Belgian supplies and the signs are that this share will increase. Natural gas customers who use L-gas are supplied directly from the Netherlands, or indirectly, in backhaul, via the Blaregnies interconnection point with France. For the L-gas market, substantial supplies in backhaul from Blaregnies (6.7%) on the transit flow initially intended for the French market were observed. This reflects the issue of capacity availability and allocation at the Hilvarenbeek/Poppel interconnection point on both the Dutch and the Belgian side that will be further discussed in chapter FIGURE 17. NATURAL GAS IMPORTS PER ENTRY POINT IN Source: CREG * The entry points of Blaregnies are used in reverse flow of the physical flows by making use on those points of the dominant natural gas flow. 16

17 In the risk assessment, we have already mentioned that the suppliers (H-gas) no longer base their portfolios on the demand of the consumers in a specific country, but they rather keep EU-wide portfolios. At least the major players on the natural gas market do so. On the one hand, this makes it very difficult to isolate the supply portfolio destined for the Belgian market. On the other hand, it increases the flexibility of the suppliers to deal with an emergency somewhere in the European market. Mostly they have storage available in other countries and access to flexible contracts and more diversified supply routes. This means that renominations should also be easier to handle in case of an emergency. In order to increase the flexibility of the system and to increase the security of supply, the TSO worked on an entry-exit model that has recently been approved by the national regulator and that will be introduced by the TSO as of October The main advantage for the security of supply (mainly for the H-gas network) of the Entry-Exit model is the increased flexibility that will be introduced through: only one balancing point for H-gas (imbalances over the currently three H-gas balancing zones will automatically be re-distributed to create virtually one balancing zone) and one for L-gas on which intraday balancing will be applied; introduction of a virtual trading point for H-gas; free exchange of quantities of natural gas from any interconnection point within the Belgian system to any interconnection point or domestic exit point Wholesale and retail market As mentioned before, total natural gas consumption in Belgium in 2011 amounted to TWh. Based on the energy deliveries on the transmission network, we can set out the relative market shares of all the suppliers active on the Belgian natural gas market. Distrigas (45%), GDF Suez (28%) and EDF Luminus (former SPE Luminus) (9%) make up 82% of the total energy deliveries on the natural gas market. FIGURE 18: MARKET SHARES OF NATURAL GAS SUPPLY COMPANIES ON THE TRANSMISSION NETWORK, 2011 Source: CREG 17

18 The retail market in 2011 We have to keep in mind that the retail market (and a smaller part of the wholesale market) consumes two different types of natural gas: H-gas (with high calorific value) and L-gas (with low calorific value). The L-gas market serves over half of the Belgian customers connected to the distribution grid and supplies certain regions exclusively, including the Brussels region and Antwerp city. Most of the customers connected to the L-gas grid are households. Competition on the L-gas market is more limited, especially for the large industrial consumers connected to the L-gas network. Most supplies were based on a long-term contract between the Dutch gas supplier GasTerra, Distrigas and GDF Suez. However, other operators are gradually entering this market as the issue of the gas quality (difference between H-gas and L-gas) has disappeared in the Netherlands which makes it easier for other suppliers to enter this market and to obtain access to the L-gas. In 2011, eleven shippers were importing L-gas to Belgium. However, the capacity availability as well as the capacity allocation at the Dutch side of the only cross-border interconnection point for L-gas (at Poppel/Hilvarenbeek) is still not straightforward for suppliers thatwant to enter into the Belgian L-gas market. Taking a look at the market shares on the retail market (for H-gas and L-gas combined), we see that ECS takes up more than half of the market with a total market share of 54.6%. EDF Luminus is the second largest player with a total market share of 17.9%, followed by Distrigas (9,1%). In July 2011, ENI, the owner of Distrigas announced that it would take over the activities of Nuon Belgium, Nuon Wind Belgium and Nuon Power Generation. This way, ENI would have bigger acces to the distribution market as a complement to the activities of Distrigas, which focusses mostly on the natural gas supply on the wholesale market. FIGURE 19: MARKET SHARES OF THE NATURAL GAS COMPANIES IN DISTRIBUTION NETWORK, 2011 Source: CREG,VREG, CWAPE, BRUGEL 18

19 FIGURE 20: MARKET SHARES OF THE NATURAL GAS COMPANIES IN THE INDUSTRIAL AND ELECTRIC CONSUMER CATEGORY, The wholesale market in 2011 Wholesale market players in Belgium providenatural gas to the suppliers on the distribution network and to large customers connected to the transmission grid. In 2011, 34 suppliers are holder of a transport license delivered by the federal authority among which 17 were active and have used the licence to supply natural gas to the Belgian market. The wholesale market players sell natural gas to the suppliers on the distribution network and to about 250 large industrial end-users and power stations connected directly to the transmission grid. Distrigas is still the largest player on the wholesale market with a market share of 45.7%, which is a slight decrease compared to last year. GDF Suez is the second largest player and increased its market share to 27.4%. EDF Luminus (former SPE Luminus) was able to retain its market share of 8.6%. Those three companies together represent 81.7% of the total wholesale market. The smaller players like Wingas, Statoil, Lampiris, E.on Energy Trading and RWE Supply & Trading Nederland all have a market share between 2-4%. Six new players became active on the Belgian wholesale market in The market shares of the new players and of the other smaller existing players on the wholesale market did not exceed 2%. The merge of GDF and Suez still has its repercussions on the market shares of the large players, through which the market share of GDF Suez increases at the expense of the market share of Distrigas. Historically, natural gas was (and for a large part still is) imported through long term contracts between the producers and suppliers. The liquidity can be increased by trading on the gas hub. In the national gas trade, Zeebrugge has a key commercial role as one of Europe s major spot markets for natural gas. Huberator, the operator of the Zeebrugge Hub, facilitates over the counter trading of natural gas, while exchange-based trading is operated by APX Gas ZEE, established in It provides a spot market for trading of within-day and day-ahead natural gas contracts. The figure below gives an overview of the composition of the supply portfolios of the suppliers active on the Belgian gas market. The long term (duration of more than five years) and shorter term contracts (duration < 19

20 5 years) with the producers still make up the majority (73.4%) of the existing gas contracts. The sourcing on the wholesale market is more volatile and represented 34.1% of the total gas contracts in The reason for this volatility is that the gas sourcing on short term basis is highly dependent on the gas demand in a specific year and the dominance of the long term contracts in the portfolios of largest gas suppliers on the Belgian gas market. FIGURE 21: COMPOSITION OF THE SUPPLY PORTFOLIOS OF THE ACTIVE SHIPPERS ON THE BELGIAN MARKET IN 2011 Source: CREG Natural gas in storage Belgium has only one underground storage installation operated by Fluxys Belgium(used for commercial storage), which is the aquifer storage in Loenhout, with a useful storage capacity of 700 million m³(n). Only high calorific gas (H-gas) is stored in this facility. Short term LNG storage is also available at the Zeebrugge LNG terminal. Part of the stored natural gas is reserved by Fluxys Belgium for operational balancing of the network. The rest of the storage capacity is sold to the market for dealing with seasonal swings and situations of peak demand. Under normal winter circumstances and if the storage is filled up to its full capacity, the stored natural gas can be used during about 45 days during peak output. Natural gas injection in the storage normally starts in April and ends in September while withdrawal lasts from October until March. The storage capacity in Loenhout used to be allocated in priority to storage users who supply gas distribution networks. In 2011, there was a change in the gas act of 1965 through the law of 11 June 2011 to allow third party access to the storage facility. The law allows for storage capacity to be booked for the long (10 years), medium or short term (less than one year). The short term capacity is sold through an auction mechanism. All of the available capacity in 2011 was commercialised in a successful way and this for the long term as well as the short term. The law of 11 June 2011 also foresees that the competent authority can oblige the natural gas undertakings to liberate the natural gas in storage if an emergency situation as specified in the regulation 994/

21 occurs. This natural gas should be used in first instance to supply the customers connected to the distribution network 6. TABLE 1: NATURAL GAS STORAGE CAPACITY IN BELGIUM (H-GAS, JUNE 2010): Location Type Working capacity 7 (10 6 m³(n)) Peak output 8 (10 6 m³(n)/day) Loenhout Underground (625 k*m³(n)/h) The Zeebrugge LNG terminal also has storage capacity available (a working capacity of 228 mcm in gas and a peak send out of 42 mcm/day), but because of the high number of slots that are allocated, the LNG storage have to send out almost immediately after the LNG cargos have been unloaded. Therefore, the LNG storage tanks do not operate as storage as such but more as a very temporary buffer before sending out in the pipelines. 6 This will be further determined in the emergency plan that will be developed in the framework of the regulation n 994/2010 on the gas security of supply 7 Working gas capacity = total natural gas storage minus cushion gas 8 Peak output = the maximum rate at which gas can be withdrawn from storage 21

22 TABLE 2: GAS IN UNDERGROUND STORAGE WINTER PERIOD Date Aug Sept Oct Nov Gas storage capacity, 10 6 m³(n) (1) Gas amount in storage, 10 6 m³(n) (2) Gas stocks change, mcm (-withdrawal, + injection) - withdrawal injection Maximum withdrawal capacity, 10 6 m³(n)/day (3) Remaining days for using the stored gas (4) 41 days 44 days 47 days 47 days 43 days 35 days 29 days 18 days 40 days 43 days Dec Jan Feb Mar Apr May ) As from 15/04/2010, the Peak Shaving Plant (PSP) at Dudzele is out of service, the extension of the working volume at Loenhout has increased by 25 mcm during the storage season )Gas amount in storage is given for the 1st day of the month. 3) As from 1st of November the send-out capacity at Loenhout was increased from 12 mio m³(n)/d towards 15 mio m³(n)/d. 4) The remaining days for using the stored gas are now calculated as gas in storage divided by the Maximum withdrawal Capacity. This maximum withdrawal capacity can be sustained for maximum 3 days at full send-out regime, afterwards technical storage constraints appear and the actual withdrawal capacity is lower. 22

23 2. Level of expected future demand and available supplies This chapter gives some insights in the calculation of the future natural gas demand and checks whether the available entry capacity on the Belgian network is sufficient to fulfil the winter peak demand for the end consumers and how much of the capacity is available for border-to-border transmission to other countries. Therefore an assumption has to be made of the expected growth on the distribution network, of the large industrial consumers and of the electric utilities. The growth of the peak demand for the distribution network is the most crucial to calculate. For the large industrial consumers and the electric utilities, we use default values based on the statistical consumption with a probability of 99% Evolution of the peak demand on the distribution network In order to estimate the future consumption levels of the peak demand on the distribution network, we make an estimation of the natural gas demand during the winter peak period for a normalised temperature profile. We assume that the peak consumption is obtained at a winter peak day. The winter peak day is considered as one day with an extreme temperature occurring with a statistical probability of once in 20 years. Belgium considers the use of -11 Ceq or 27.5 Equivalent Degree Days as the reference value for the winter peak occurring once in 20 years in anticipation of the standard that will be established in the legal framework on the emergency plan.. The evolution of the natural gas demand on the distribution network at -11 Ceq (27.5 DDeq) is estimated based on a linear regression model. According to the determined parameters for this regression, we can compute: - the average value of the natural gas consumption level at -11 Ceq, this is a 50% chance that the actual natural gas demand will be higher than the calculated demand 9 ; - the natural gas consumption level at -11 Ceq with a 1% risk of exceeding the calculated natural gas demand ; - the average growth rate in an analysed period; - an estimation of the peak natural gas consumption in the coming winters. The calculations are based on the average demand on a peak day. We do have to keep in mind that this differs from the hourly peak demand which can be up to 20% higher than the average peak day demand. This difference will have to be absorbed by the flexibility in the system (e.g. through the linepack or imported through the interconnection points) Evolution of the daily average peak demand on the L-gas distribution network Based on the daily average consumption data on the L-gas network data for the winter periods of 2003/ /2012, we can calculate the parameters for the regression curves. The curve Q50% 9 A correction factor is applied to the linear regression to cover 99% of the measured data (so called 1% risks). 23

24 represents the calculated natural gas demand (in volumes) by the distribution network for the winter peak period at -11 Ceq on the L-gas network at a 50% risk of obtaining lower temperatures and therefore a higher natural gas consumption. We also calculate the consumption at a 1% risk of exceeding the calculated natural gas demand due to lower temperatures. Based on those curves, we can compute the linear regression curve that provides us with the growth rate of natural gas demand from the L-gas distribution network in the past years. We take into account the average risk value and therefore we obtain a growth rate of 1.86% per year for L-gas. The figure below shows the results. This will be our basis hypothesis to start from to calculate the future winter peak natural gas demand by the L-gas distribution network. One aspect to take into account by using a linear regression model is that it might overestimate the peak natural gas demand for the coming years. This is because possible smoothening effects that might occur in the years to come for example due to saturation and efficiency gains are not taken into account in a linear model. However, in order to calculate possible effects on the security of gas supply, it is useful to make predictions on a worst case scenario basis. Therefore we make use of the 1.86% calculation for the future peak demand simulation on the L-gas network. FIGURE 20: DAILY AVERAGE WINTER PEAK DEMAND FOR L-GAS ( ) TD L (k*m³(n)/h) - hourly average Thousands m³(n)/h Q50%(-11 C) Linear (Q1%(-11 C)) Q1%(-11 C) Linear (Q50%(-11 C)) Growth TD L-gas calculation over period Q50%(-11 C) Q1%(-11%) vector based on linear regression winter period 2003/ /2012 1,64% 1,35% 24

25 vector based on real data winter period 2003/ /2012 1,86% 1,66% vector based on real data winter period 2008/ /2012 2,50% 1,96% vector based on real data winter period 2010/ /2012 3,43% 3,38% Max. growth rate winter period 2006/ /2012 2,54% Min. growth rate winter period 2007/ /2012 1,49% The growth rate at a 50% risk based on the linear regression of the natural gas consumption over the period 2003/2004 and 2011/2012 gives us a growth rate of 1.64%. The minimum and maximum growth rates are also calculated over this period and lie between 1.49% and 2.54%. We also note that the growth rates at a 1% risk are much more stable. For the calculation of the future demand projections of the distribution network, we will take into account the growth rate based on the real measured data between 2003/2004 and 2011/2012. This gives us a growth rate of 1.86% Evolution of the daily average peak demand of the H-gas distribution network The same method as for L-gas described above is used to obtain the linear regression curves. The H- gas consumption on the H-gas distribution network at -11 Ceq is calculated in the 50% and the 99% range. Based on those curves the growth rate for H-gas demand can be calculated through a linear regression curve. The use of the linear regression model to calculate the future peak natural gas demand on the H-gas distribution network might (as for L-gas) overestimate the peak demand as it does not take into account new evolutions in the network that might have a smoothing effect on the future natural gas demand. We obtain a growth rate for peak natural gas demand of H-gas over the years of 2.44% per year. This is the growth rate in the 50% range. 25

26 FIGURE 21: DAILY AVERAGE WINTER PEAK DEMAND FOR L-GAS ( ) TD H (k*m³(n)/h) - hourly average Thousands m³(n)/h Q50%(-11 C) Linear (Q1%(-11 C)) Q1%(-11 C) Linear (Q50%(-11 C)) Growth TD H-gas calculation over period Q50%(-11 C) Q1%(-11%) vector based on linear regression winter period 2003/ /2012 2,07% 1,80% vector based on real data winter period 2003/ /2012 2,44% 2,25% vector based on real data winter period 2008/ /2012 2,45% 1,83% vector based on real data winter period 2010/ /2012 4,07% 3,38% Max. growth rate winter period 2006/ /2012 2,96% Min. growth rate winter period 2007/ /2012 1,36% The growth rate taken for H-gas on the distribution network based on the real data amounts to 2.44 % / year. The trend between 2003/ /2012 gives us a growth rate of 2.07% in the 50% risk scenario. The minimum and the maximum growth rates for this period are 1.36%/year and 2.96%. The growth rates at the 1% scenario are again more stable calculated over the different years. For the projections up to 2022, we will take a growth rate of 2.44%/year for H-gas the consumption on the distribution network. In order to interpret the consequences of this assumptions, we need to keep in mind that this is a more static assumption that does not take into account possible changes in the natural gas consumption due to a higher connectivity to the gas network, demographic changes, switching behaviour of heating sources, energy efficiency,. 26

27 2.2. Natural gas demand projections up to 2022 Based on the calculated growth rates of the peak natural gas demand for L-gas and H-gas on the distribution network, it is possible to evaluate the future total natural gas demand for the coming 10 years. Therefore we will have to add the estimated consumption of the distribution network to the estimated consumption of the large industrial consumers and the electric utilities. For the distribution network, we use as hypothesis a growth rate of 1.86%/year for L-gas and of 2.44% for H-gas. Those growth rates are obtained through a regression analysis as explained in the chapter above. The estimations of the natural gas demand by the large industrial consumers and the electric utilities come from the Transmission System Operator (TSO). For each large industrial consumer, power plant or cogeneration unit, the TSO determines a peak capacity perspective of the hourly consumption (default value) that represents the real natural gas needs. The calculation of the default value is based on a statistical analysis of the highest hourly consumption over the three previous years. Non-representative data can be excluded. Therefore, weekends, official holidays, abnormal peaks (test phase, incident, ) or abnormally low consumption will not be taken into account. The statistical analysis also identifies, as a sort of warning mechanism, each drastic change in the consumption profile of the consumer over a certain period. Based on those alerts, the TSO, in cooperation with the customer, can analyse the underlying reasons for such alerts and determine whether future investments in the network will be necessary. In 2012, the default value for the consumption of the customers connected to the gas transmission network reduced by 4.73% compared to the previous year. The TSO has also information available about the future expansion, construction or closure of new industries or power plants that will be connected or taken off the network for the next five years. New or expanding industries or electrical power plants give an estimation of the capacity they would need for natural gas supply within a certain time delay. Based on the provided figures, a low case, medium (or reference) case and high case scenario can be drafted for the large industrial consumers and the electric utilities. In the low case scenario, we assume hardly any new power plants or large industrial clients coming on line. The medium or reference case takes up all the new or expansions of large industrial clients or power plants for which a final investment decision has been taken and the high case scenario takes up all the projects that have been introduced but for which no final investment decision has been taken. Also the conversion from certain large industrial consumers on the L-gas network to the H-gas network has been taken into account. In order to make the calculations, we assume that the planned conversion will take place in the year This will create a decrease of about 63 k.m 3 (n)/h on the L-gas network and an increase in the industrial consumption on the H-gas network of 53 k.m 3 (n)/h. 10 This is an assumption in order to facilitate the calculations. The exact dates of the conversions are not known yet, so for the sake of convenience it is assumed that the entire conversion will take place in

28 On this base case scenario, we add the new demand for transmission capacity coming online in the coming ten years in the low case, medium case and high case scenario for L-gas and for H-gas Projections of the hourly daily average L-gas demand and entry capacity Figure 21 below gives an overview of the projected total natural gas demand on the L-gas network in the three scenario s, the available entry capacity and the share of capacity that will be available for natural gas transmission through Belgium from border to border in the coming years. This is an estimation of the daily average that is based on the hourly data of the consumption during the winter peak period. The natural gas demand is based on the natural gas demand from the distribution network to which the calculated growth rate for the peak demand was applied over the coming 10 years and the increased consumption from the large industrial consumers and the electric utilities in the low, medium and high case scenario. Based on the assumptions explained above, the total average peak demand would increase from 2119 k.m 3 (n)/h in the winter to 2450 k.m 3 (n)/h by 2022 in the low case scenario, to 2458 k.m 3 (n)/h in the medium case scenario, and to 2459 k*m 3 (n)/h in the high case scenario. In spite of the slight difference, we can consider that the three scenarios lead to identical conclusions. This would be an increase of about 16%. The increase in the peak demand is mainly due to the increase of the peak consumption by the distribution network. The growth of the large industrial consumers will be very limited and there are no new power plants that will be implemented on the L-gas network. It is to note that the L-gas market is already rather saturated, it will be likely to see a stagnation in the consumption on the L-gas network. We compare the expected demand with the available entry capacity to evaluate the remaining capacity in the network to get an estimation of the available capacity for natural gas transmission from border to border to the neighbouring countries. The available entry capacity is calculated based on the lowest offered firm capacity on the borders with the adjacent network operators. On the Dutch border, the offered firm capacity will decrease next year by 90 k.m 3 (n)/h (2 575 k*m³(n)/h during winter 2013/14 et 2014/15) and will increase again by 40 k.m 3 (n)/h by the year 2015 (2 615 k.m³(n)/h during winter 2015/16) 11. As we have no insight in the results of the future open seasons for the next winter periods, we assume that no other changes will occur in the available capacity (2 615 k*m³(n)/h). The exit capacity available on the Dutch side of the border plays a crucial role in the security of L-gas supply in Belgium. Therefore it would be useful to have an insight of the future developments of the exit capacity on the Dutch side of the border in order to move towards a match with the capacity available at the Belgian side which is sized according the expected needs of the Belgian L-gas market (and transmission needs to supply France with L-gas). On top of the L-gas imports from the Netherlands, Belgium is able to convert H-gas into L-gas thanks to the conversion units in Lillo (production capacity of L-gas: 300 k.m³(n)/h) and in Loenhout 11 The information on the available capacity is based on the current statements of GTS and are subject to changes over the coming years. 28

29 (production capacity of L-gas: 100 k.m³(n)/h). Those conversion units from H-gas to L-gas increases the entry capacity by 400 k.m 3 (n)/h. The conversion plant allows conversion of H-gas into L-gas if there is a shortage for L-gas in any given period. Total entry capacity for L-gas is about 3000 k.m 3 (n)/h (e.g. in winter 2015/2016, entry capacity : 2615 k*m³(n)/h and conversion capacity 400 k.m³(n)/h). The difference between the entry capacity and the projected demand results in the maximum available exit capacity to France (Exit Taisnières/Blaregnies). This is represented by the light blue curve. As the demand for L-gas increases and the entry capacity remains approximately unchanged and limited, the available exit capacity to France decreases from 946 k.m 3 (n)/h in 2012 to 556 k.m 3 (n)/h by Up to 2016, we have to consider whether there is enough capacity to meet both the French and Belgian gas demand, taking into account the importance L-gas storage of Gournay for the French market in the supply chain. FIGURE 22: L-GAS DEMAND AND ENTRY CAPACITY PROJECTIONS FROM (IN K*M 3 (N)/H) L gas flow hourly daily average (k*m³(n)/h) Low Scenario Medium Scenario High Scenario Entry Capacity Conversion cap. Entry + Conv. cap Max. Exit cap. Taisn. exit cap Projections of the daily average H-gas demand and entry capacity The same calculations as for L-gas are made for H-gas and presented in figure 22. The projections for the total demand in the low, medium and high scenario are represented by the green, orange and red curve. 29

30 The growth of the consumption at the winter peak demand on the H-gas distribution network was set at 2.44%/year. The total average peak demand would increase from 3890 k*m 3 (n)/h in the winter to 4663 k*m 3 (n)/h by 2022 in the low case scenario or an increase of 19.9%. In the medium case scenario, the average peak consumption would increase by 26.4% from 3891k*m 3 (n)/h in 2012 to 4919 k*m 3 (n)/h in 2022 and in the high case scenario, we would see an increase by 36.8% from 3895 k*m 3 (n)/h to 5330 k*m 3 (n)/h in The increase in the peak demand by 2022 is mainly due to the increase of the importance of the electric utilities in the gas consumption. Also the higher peak consumption on the distribution network (+30.4% compared to 2012 calculated data) contributes significantly to the total increase.. Concerning the electric utilities in the low case scenario, we assume that no new power plants will be built until In the medium case scenario, three new natural gas fired power plants will come online after 2016 and in the high case scenario six natural gas fired power plants (10 units in total with a total capacity of about 4600 MW) will come online as from 2014 based on the current capacity commitments that have been made to Fluxys Belgium. Currently the six projects are all in study phase; the realisation of the project will only start after the written notice and capacity booking send by the customer to Fluxys Belgium. The electric utilities contribute a great deal to the increase in the total H-gas consumption in the high case scenario. It is however assumed in the high case scenario (projects that have been introduced but no final investment decision has been taken) that the first new natural gas fired power plant will be online by 2014 while the likelihood of this happening is rather low. In addition, it is to be noted that the lead time to construct a gas fired power plant is about three years and the lead time for connecting a possible new power plant on the Fluxys Belgium network will depend on the location of this potential power plant (near or far from an existing pipeline). In some cases, location of a possible power plant is close from an existing pipeline or an existing site and the connection can be performed rather quickly; in some other cases, if there are no pipeline near the chosen location, it can take until five years to connect the power plant to the Fluxys network. A steady but more limited growth is recorded for the large industrial consumers connected to the transmission grid. The maximum consumption of the new potentials coming online in the category of the large industrial consumers in the next ten years does not exceed 131 k*m 3 (n)/h (in the high case scenario). No significant increase in the market share of the large industrial consumer category in the total natural gas demand is noticeable. The entry capacity will increase in 2015 with 750 k*m 3 (n)/h 12 which can be attributed to new pipeline connecting Dunkirk to Zeebrugge (new entry in Alveringem at the French-Belgian border). This can be seen in the graph on the capacity. Contrary to L-gas, we need to subtract the capacity (360 k*m 3 (n)/h of Hgas) used by the conversion plan to produce L-gas if needed (e.g. during the winter peak). The blue lines show the exit capacity that is available in a low, medium or high scenario. For the H-gas network, we can assume that at the moment, there is sufficient capacity to meet the needs of the domestic Belgian market and for border-to-border transmission through Belgium, but that the capacity available for either the Belgian market and the border-to-border 12 The capacity of the new pipeline was set at 750 k*m 3 (n)/h in this calculation as the exact amount of capacity that will be available was not known at the time of the calculation. 30

31 transmission will nonetheless decrease with 11.6% in the high case scenario in the period 2012/2022. FIGURE 23: H-GAS DEMAND AND ENTRY CAPACITY PROJECTIONS FROM (IN K*M 3 (N)/H) H gas flow (k*m³(n)/h) - hourly daily average Low Scenario Medium Scenario High Scenario Entry Capacity Conv. H->L Entry - Conv. Exit avail. (Low case) Exit avail. (Medium Case) Exit avail. (High case) Projections of the H-gas demand at the hourly peak and the available entry capacity In addition of the calculations made on a daily average basis, the same is made for the hourly peak value as this will have its implications on the flexibility in the network. We assume that the hourly peak equals the average peak increased with 20%. The data for the increase in the natural gas consumption in the three scenario s (low, medium and high case) remain the same, so the main conclusion is that the remaining capacity available for exports during the hourly winter peak will be 20% lower during the peak than in the scenario described above. The total H-gas demand on the hourly winter peak will reach 5098 k*m 3 (n)/h in the low case scenario, 5354 k*m³(n)/h in the medium case scenario and 5765 k*m 3 (n)/h in the high case scenario. The entry capacities are the same as mentioned above. This decreases the available border-toborder transmission capacity or the capacity necessary to supply the Belgian market by 5780 k*m 3 (n)/h in the low case scenario and by 5113 k*m 3 (n)/h in the high case scenario. 31

32 FIGURE 24: H-GAS DEMAND AND ENTRY CAPACITY PROJECTIONS FROM (IN K*M 3 (N)/H) H gas flow (k*m³(n)/h) Low Scenario Medium Scenario High Scenario Entry Capacity Conv. H->L Entry - Conv. Exit avail. (Low Case) Exit avail. (Medium Case) Exit avail.(high case) 4. Natural gas infrastructure 4.1. Existing and new infrastructure Fluxys Belgium, Belgium s transmission system operator, has a network of more than 4000 kilometres of pipelines with 18 interconnection points and five compression stations. The 9 cross border pipelines connect the Belgian natural gas market directly to Norway, UK, Germany, the Netherlands, France and Luxembourg. The Belgian network disposes of five compressor stations that are all (except for Weelde) entirely electrically driven (like elsewhere in the EU: more and more compressor stations are electrically instead of natural gas driven), which again confirms the growing convergence between electricity and natural gas security of supply which deserves particular attention. The five compressor stations are located in: Weelde: The compression station in Weelde was upgraded in 2010 to increase the pressure of lowcalorific natural gas in the pipeline from Poppel on the Dutch border to Blaregnies on the French border. 32

33 Winksele. The compressor station at Winksele lies at the crossroads of major North/South and East/West transmission axes. The existing station serves to maintain pressure levels in the North/South L-gaspipeline. Due to increasing demand for capacity (H-gas) for both the domestic market and border-to-border transmission together with the increased flexibility needs to introduce the entry-exit system, it was decided to build new compressor facilities in Winksele for the East/West H-gas transmission axis as well (vtn-rtr pipeline). In this way, capacity can be enhanced on both the East/West and North/South transmission axes. The aim is to commission the new compressor facilities in late 2012 as it will be needed to support the changes that will be brought about the Entry-Exit model that will come into force by October Berneau. Since 2010, work has been under way on a project to build additional compressor facilities at the compressor station in Berneau. These upgrade works will extend the options for combining natural gas flows on the East/West and North/South axes and also fit in with, inter alia, the transition to an entry/exit system and the introduction of a virtual trading point. The aim is to bring the additional compression into service in late Work is also under way at the compressor station to replace two existing compressors on the s-gravenvoeren-blaregnies pipeline. Sinsin: on the H-gas pipeline to Luxembourg. This station will be taken out of service after the enhancement of the Luxemburg pipeline in Zelzate: The Zelzate compressor station came on line at the end of 2008 to create additional capacity for the overall rise in demand of the Belgian domestic market and enables larger volumes to be transported to and from the underground storage facility in Loenhout. Since 2010 it is also used to recompress the cross border natural gas delivery at Zelzate from the Dutch GTS grid. The Fluxys Belgium transmission grid is also connected to other facilities operated by Fluxys Belgium or its subsidiaries: the Loenhout underground storage facility and the LNG terminal located in Zeebrugge. The Loenhout underground storage facility is aquifer storage for high calorific natural gas that combines mainly seasonal storage with high flexibility of usage. The Zeebrugge LNG terminal is used to load and unload ships carrying liquefied natural gas (LNG). LNG is temporarily kept in storage tanks at the facility as a buffer before re-gasifying the LNG and injecting it into the grid for transmission or loading the LNG back onto LNG ships. 33

34 FIGURE 25: THE BELGIAN TRANSMISSION NETWORK (2012) Source: Fluxys Belgium The need for new infrastructure is evaluated every year by Fluxys Belgium in the ten-year indicative investment program. These updates take into account the changes in requirements in terms of natural gas supply, request for new connections and the changing needs of grid users identified through subscription periods and international market consultations among other things. Several simulations based on the winter peak (January at -11 Ceq) are being set up to calculate the effects on the network. In the latest investment plan, the calculations were already based on the entry-exit model that will be effective by October Current investments are underway or realized: In January 2010, Fluxys Belgium launched the VTN2-RTR2 project which involves laying a second pipeline next to the existing VTN1-RTR1 pipeline. This project was finalized in 2011 and in service since end This creates additional physical reverse flow on the east west axis. This project consolidates border-to-border capacity and helps considerably in bolstering security of supply in Belgium and northwest Europe. Therefore the VTN2-RTR2 project was one of the priority TEN-E projects listed by the European Commission Enhancement in the pipeline to the Grand Duchy of Luxembourg: between Ben-Ahin and Bras the transmission axis to Luxembourg is being replaced by a larger section pipeline. Normally, the works should be finished by Merelbeke-Zwijnaarde: The pipeline between Merelbeke and Zwijnaarde was rerouted over a distance of more than 5 km to allow a tramline to be extended in The newly 34