An opportunity for ACER to step in and complete the internal gas market

Executive summary

Cross-border tariffs for gas transmission in the EU are having an impact on price convergence and ultimately on the completion of the internal gas market. It is of utmost importance that they are set at the right level. However, there is an incentive for NRAs to overcharge neighbouring markets. The current regulatory architecture should be changed.

CRE's consultation on gas transmission tariffs for the 2020-2024 period in France (ATRT7) is about to close on 4th October 2019. The ill-informed proposal maintains abnormally high exit tariffs to Spain and Switzerland/Italy, in the former case four times higher than the ones charged to national consumption for the same service. While arguably this is not compliant with the Tariff Network Code, not cost-reflective, and clearly discriminatory, neither ACER nor any other authority have effective powers to avoid its approval. This tariff is the highest between any two EU members and is at the root of price differentials between North-West Europe and Iberian markets. The cost for Spanish and Portuguese consumers is massive.

The potentially discriminatory situation was already noted by the Italian and Spanish gas sectors in spring, including the Italian regulator, sector associations and industrial consumers in both countries, as well as by EFET, with no apparent effect on CRE's views.

ACER should nevertheless step in as soon as possible and ask for a new consultation in France with complete information, and following a proper implementation of the TAR NC. The gas internal market is what is at stake.

A European Green Deal and the Decarbonisation package

On Friday 27th September the new ACER Director, Christian Zinglersen, was designated by the Agency. His background on global climate negotiations may be signalling the emphasis that the Agency is going to place on decarbonisation. It comes at a time when European Commission’s new President, Ursula von der Leyen, has promised to propose a “European Green Deal” in her first 100 days in office.

The Agency is indeed going to face a great deal of work as regards the implementation of the Clean Energy Package and the development and later implementation of the (Gas) Decarbonisation Package, which will have to tackle the penetration of low-carbon and renewable gases, sector coupling issues, and a number of governance challenges. However, ACER must not forget the pending homework from the Third Package: the completion of the internal gas market.

The (pending) software challenge

Completing the internal gas market involves both working on the hardware (infrastructures) and on the software (regulation, notably the implementation of Network Codes). On the hardware front, some missing links remain, arguably between Iberia and the rest of Europe. Let’s forget about that for today. On the software front, the gas sector had been remarkably successful developing implementing most of measures…until it came to the Tariff Network Code (Commission Regulation (EU) 2017/460 of 16 March 2017). If its development was particularly tortuous, finally arriving to an arguably weak level of harmonisation, ACER’s analyses of NRAs tariff methodology proposals suggest that its implementation is proving to be even more challenging. And the high level of cross-border, exit tariffs is one of the concerns.

Fortunately, for the time being this challenge seems to be already present on ACER’s mind. In the recent consultation on “Public Consultation on The Bridge beyond 2025“, numerous references are made to the TAR NC, cross-border tariffs, and their impact on price differentials. The Agency acknowledges that “Implementation of the Network Code on harmonised gas tariff structures (TAR NC) reveals that there may be further room for improvement of the methodology, namely in order for cross-border tariffs to allocate properly the costs of the network used by domestic and non-domestic flows.”

But what can the Agency do?

NRAs in charge of market integration: like a fox guarding a henhouse

While one may argue that the proper application of the TAR NC should in principle remedy unduly high cross-border tariffs, under the current TAR NC there are no sufficient guarantees that National Regulatory Authorities (NRAs) will follow the TAR NC principles, and in particular ACER recommendations. In fact, ACER’s analyses of NRAs tariff methodology proposals show frequent deviations from the TAR NC requirements. In theory, once ACER analysis is notified, NRAs shall adapt their methodologies, as well as complete the missing information. But if an NRA persists on deviating from the TAR NC, ACER won’t have any opportunity to review the methodology again before it is in force. This effectively allows NRAs to deviate from the TAR NC. I.e., there is not much the Agency can do.

But, why would an NRA deviate from the principles established in a Regulation? Moreover, what is the incentive for an NRA to overcharge gas exports?

A fundamental contradiction of the governance arrangements of the internal market is that many relevant tasks have been left to NRAs instead to EU institutions. NRAs have the statutory obligation of minimising costs for national consumers, which is frequently in conflict with measures to integrate markets. In particular, they have an incentive to overcharge gas exports to minimise the cost allocated to national consumers, i.e., to introduce a cross-subsidy from foreign consumers to national consumers. In those cases, higher costs are passed to downstream markets which is not only wrong from an allocation point of view, but also detrimental for gas trading and for the well-functioning of the EU gas market.

Only a supranational institution may put a remedy to this situation, caring for the benefit of all consumers in a region. To that regard, in the European Union more executive powers on tariff setting (different from allowed revenues, by the way) should be transferred to ACER. The independence of ACER should also be reinforced in order to avoid that NRAs can block or unduly influence its decisions, or that ACER is bound by political interests.

But is this just theory, or is there any evidence that cross-border tariffs are overcharging?

The German precedent

The implementation of the TAR NC has already created tensions between Germany (BNetzA), Italy and France (CRE). In an article published by EURACTIV on 2nd April 2019, (“Italy squeals on German gas tariff reform, EU ready to step in”), it is explained how ARERA, the Italian regulator, and also the CRE, the French regulator, complained to BNetzA of the increase of export tariffs under the German regulator methodology proposal, due to the effect it would have in their markets. In the case of the Italian regulator, the article notes that ARERA argued that the reform proposed by BNetzA would have significant impacts on Italy’s wholesale gas market, since gas imported from Germany is directly setting the Italian wholesale price. Therefore, any additional cost associated to this transportation route affects the Italian gas market as a whole. Moreover, it argued that the new system would exacerbate the “pancaking effect” and distort gas price formation across different EU countries. According to ARERA, the forecast increase in transportation costs due to the German tariff reform amounts to 0,387 €/MWh, which represents an additional cost of roughly €300,000,000 per year for Italy’s gas supply.

As worrying as the prospect for the Italian market may be, the main problem faced by stakeholders affected is that the TAR NC is quite vague and, within certain margins, allows for very different outcomes. As, according to EURACTIV’s article, put by an EU spokesperson, “an increase in tariffs for exit points does not in itself constitute an infringement to EU law, as long as the underlying tariff methodology is in line with the 2017 EU regulation”. And what if it is not in line? Then, according to the EU spokeperson mentioned in the article, “the Commission may decide to launch an infringement procedure against Germany. The European Commission, as guardian of the Treaties, has to make sure that EU legislation is being followed”.

In summary, no worries: if the NRA is adopting a discriminatory decision, just wait for a couple of regulatory periods and the problem may be solved. Is this really the architecture we need to complete the internal gas market?

Persistence of pricing differentials

In a previous article published on 10th March 2016 (Mind the Gap – The Spanish gas hub challenges), I referred to the price differentials between Spain and France, and the challenges faced by MIBGAS. Three years after, Spain has done its homework: MIBGAS has shown a solid growth and gained liquidity along the curve. It is still far away from the churn rates of the most liquid hubs, but its growth has not lost momentum and new regulatory developments in Spain should further spur liquidity.

No wonder price differentials between Spain and France remain.

As a matter of fact, the exit tariff from France at VIP Pirineos is already, and by far, the highest between to EU members: 1,690.63 €/GWh. This tariff alone is more expensive than any combination of (exit+entry) tariff between two EU Member States. Adding the Spanish entry tariff, the total cost goes up to around 2 €/MWh.

The high level of this tariff is having an influence on price differentials between hubs and, consequently, in the completion of the internal market. While prices in The Netherlands (TTF, the reference hub in Europe) and France (PEG TRF) are usually aligned, gas exported from France to Spain is frequently, and in particular in winter, setting the marginal price for the Spanish market (PVB). Therefore, the price in the Spanish market is frequently the price of the French market, plus the tariff at the IP (2.05 €/MWh, which in the short-term, adding the effect of multipliers, totals 3.65 €/MWh).

While this effect is neglected by some, the correlation between price differentials and cross-border tariffs seems to be on the radar screen of the Agency.

CRE’s tariff methodology under consultation: business as usual

Tariff methodologies and tariff levels fulfilling the TAR NC should have been approved by 31st May 2019. Implementing the TAR NC can be a complex task in larger systems and, understandably, some countries are lagging behind. France, an early adopter of other NCs, is this time among the late ones, just as Spain is. CRE has nevertheless been working on the tariff methodology for some time and has already launched three consultations in 2019 on the 7th tariff period for gas transmission in France, ATRT7. The last one was launched in late July and is closing on 4th October 2019, after 10 weeks.

No surprises as regards the methodology and results: CRE is essentially maintaining what was anticipated in the second consultation in March 2019, and in fact what was already approved for the previous regulatory period, ATRT6, which started in April 2017. Tariff levels will largely remain unchanged.

But is the proposal really discriminatory, or is it just within the margins allowed by the TAR NC?

In my view, the proposal has a number of shortcomings. While CRE argues that it is formally meeting all TAR NC requirements, a number of stakeholders I have had the chance to talk believe that it doesn’t.

• A major concern is that CRE is not clarifying what methodology it is applying. The TAR NC establishes a default methodology that must be applied, the so-called Capacity-Weighted Distance (CWD) methodology. But NRAs are allowed to deviate if there is a good reason to do so. In such case, the must offer a comparison (formerly known as “counterfactual analysis”) between the methodology adopted and the default one. Surprisingly enough, CRE is neither clarifying that the chosen methodology is the CWD one, nor offering any counterfactual analysis.
• Another notable concern is the lack of information to reasonably understand or replicate the model

But the underlying problem is that a number of arbitrary decisions seem to be made just to overcharge gas exports to Italy via Switzerland (Oltingue), and to Spain (VIP Pirineos). While this might sound like an overstatement, I’ll just mention two features of the methodology, and a consequence:

• Starting with the consequence, the gas sold in Southern France, nearby the exit points to Spain, is charged around four times less transmission costs than gas exported to Spain. I.e. a gas physically arriving from the same place to the same place can pay up to four times more if it is to be exported.
• The methodology distinguishes between “transit” (any gas export, in fact) and national consumption. However the French system has limited export flows (in comparison with the entry flows), has no dedicated infrastructures for transit, and cannot physically serve a different gas to national consumers in Southern France than the one exported to Spain.
• To calculate tariffs, flow scenarios and transmission costs should be taken into account. Normally, in an entry-exit tariff system, those points which are further away from the entries are charged more, i.e. the cost of delivering gas there is higher. But this distance is based on real, physical flows in the system, or at least on a consistent model. And here is where the distinction between “transit” and national consumption plays a role. CRE has, for some unclear reason, decided that for deriving exit tariffs at VIP Pirineos, should be adopted as the entry point for 100% of that gas. Dunkerque is 1,072 km away from VIP Pirineos. This works also for Oltingue, though the distance is a few hundred km shorter. However, the flow pattern chosen for national consumption is different: the assumption is that gas always comes from the closest point until it is exhausted, and then the second closest point, and so on and so forth. This distance is, on average, 237 km (285 in winter, 170 in summer). In the previous consultation in March it was 280 km, but transportation for national consumption has apparently gotten 43 km more efficient in the meantime.

CRE amazingly argues that the unit cost per km for national consumption and “transit” is the same. In essence, the charge for VIP Pirineos is around four times higher than for national consumption, but the distance is more than four times higher as well (1,072 km vs 237 km). The difference between VIP Pirineos and French consumption in the South is due to the fact that CRE decides to set an average tariff to national consumption, for cohesion reasons, but Spain would not be better off if it didn’t, it would just be that French consumers in the North would pay several times less than in the South. It also claims that the TAR NC allows for this distinction, though the TAR NC an ENTSOG official interpretation suggest the opposite.

The situation would be debatable if gas exported to Spain and served to consumers in the South could be physically distinguished. But it is not. One cannot but think that Iberian consumers could have been luckier, then, if CRE’s assumption had been that gas exiting at VIP Pirineos was entering in Fos LNG terminals, going up to Paris, and coming back to VIP Pirineos: the distance would be 2,000 km and then the unit charge per km much lower than for French consumers. As a matter of fact, the gas exported to Spain had historically been the one produced in the South, until the fields were depleted (though it was charged from the beginning according to the theoretical cost of physical transit), and since that gas was exhausted, as for any other consumer in the South, it has been a mix of gas from Fos terminals and gas from the North.

We happy few

Claims that gas transmission exit tariffs overcharge have traditionally be made by few stakeholders. Only Enagás and Teréga did so in the ATRT6 consultation back in 2016. In the pre-TAR NC world, it was difficult to understand tariff setting or to ask for any transparency. But something changed last spring: in the second ATRT7 consultation in March, previous to the current one, no less than 10 stakeholders noted that the proposal by CRE was discriminatory or potentially discriminatory to gas exports given the information disclosed. These were:

• ANIGAS (Italian gas sector association),
• ARERA (Italian regulator),
• Confindustria (Italian industry association),
• EFET (European Federation of Energy Traders),
• Enagás (Spanish TSO),
• ENI (Italian shipper operating in several gas markets),
• GasIndustrial (Spanish industrial gas consumers association),
• Iberdrola (Spanish shipper operating in several gas markets),
• SEDIGAS (Spanish gas sector association), and
• Shell (global energy group operating in several gas markets)

While responses supporting CRE proposal clearly outnumbered the ones complaining, most of the supporters were either French industrial consumers, associations, shippers or operators, in many cases directly benefitting from the allocation of costs proposed by CRE. In summary, the whole French sector. But the 10 stakeholders questioning the methodology essentially represent the whole sectors in Spain and Italy, including a regulator, plus one of the most prominent associations in the EU.

What’s next?

Being realistic, the most probable outcome of the process is that ACER analysis arriving in early December (2 months after the consultation closes, as established in the TAR NC) will note that a number of assumptions are not TAR NC-compliant, and most probably that some information is missing so it cannot provide a proper analysis. Then the CRE will roughly maintain the same tariff levels for Oltingue and VIP Pirineos, and then…whatever comes then will take several years. In the meantime, price differentials will remain, costing hundreds of millions to Iberian consumers.

It must be recalled that the debate now is just about how to apply the TAR NC, i.e. not unduly overcharging gas exports. ACER consultation on “The bridge beyond 2025” notes the impact of tariffs on cross-border trade and suggest that in the future a gradual rebalancing away from cross-border tariffs within the EU to tariffs on external borders and/or on demand might be appropriate. There are good reasons to think that such situation is desirable, based on some works commissioned by the EC (“Quo vadis gas market regulatory framework”), and produced by the Florence School of Regulation (“Towards an Efficient and Sustainable Tariff Methodology for the European Gas Transmission Network”). But we are not yet there: this is just about implementing what is approved.

My suggestion: ACER should step in, submitting an analysis much earlier than in December 2019, asking for a new consultation in France with complete information, and following a proper implementation of the TAR NC. The EC should be supportive. The gas internal market is what is at stake.