Gazprom Executive Discusses Proposals for Tariff Network Code

Amidst unprecedented difficulties in garnering consensus on a looming change to European legislation, Natural Gas Europe had the pleasure of interviewing Alex Barnes, Head of Regulatory Affairs at Gazprom Marketing & Trading Limited.

In the interview, we dealt with the inefficiencies related to the proposals for the new Tariff Network Code. According to Barnes, "with floating tariffs shippers are effectively signing a blank cheque whereby the final tariff they pay for capacity is based not just on their own bookings, but on how much capacity other shippers in the system book, and at what tariff."

Barnes also advocates that TSOs should be required to publish the models they use to calculate the tariffs at the entry and exit points on their system. This, combined with more certainty around the reserve price, would improve the situation, making the system better functioning. Barnes suggested that solving these inefficiencies could help creating an internal European energy market. 

At a recent conference, you presented evidence regarding gas transportation tariffs and some inefficiencies related to the current system. You basically said that shippers don’t know the price they will pay for capacity when bidding for capacity in auctions. What are the reasons for this? 

This stems from the concept of “floating tariffs” in the proposed draft code so that the price that TSOs charge for capacity will change from year to year, in order that the TSO earns its “allowed revenue” (the sum of money that it is allowed to earn by the regulator based on the value of its pipeline network, a regulated rate of return and other costs such as Opex and Depreciation). The main way that a TSO earns revenue is by the sale of capacity; therefore the revenue it earns each year is a product of the capacity it sells and the price at which it sells the capacity. Different capacity products sell for different prices. When setting the price for capacity the TSO will make an estimate of how much capacity they will sell and at what price. If they get this wrong, they will not earn the allowed revenue they expect. Floating tariffs mean that if they earn less than they expect in Year 1, they can increase the tariffs for capacity in year 2 to compensate. This tariff increase applies not only to capacity yet to be booked, but also to capacity that was booked prior to Year 2, but for use in Year 2. This could mean that capacity which was booked several years ago, long before the development of the Tariff Network Code was even thought about, will cost more than was originally envisaged. However those shippers who made such bookings will not be able to change how much capacity they have booked, and their consequent financial liabilities, even though the terms of their contract with the TSO may have changed significantly. With floating tariffs shippers are effectively signing a “blank cheque” whereby the final tariff they pay for capacity is based not just on their own bookings, but on how much capacity other shippers in the system book, and at what tariff.

The issue I explained at the European Autumn Gas Conference was related to this, but is more absurd. With floating tariffs it is difficult to forecast capacity bookings and hence year on year changes to tariffs several years ahead. This is because shippers now have more choices as to what type of capacity they book – annual, quarterly, monthly, daily – and can now book capacity on a more short term basis so that bookings can match expected flows (profiling). This is as a result of the Capacity Allocation Mechanisms (CAM) network code.

However, based on ACER’s Framework Guidelines and the current drafting of the Network Code, shippers will not even know what the tariff is for capacity for the next year. This is because the auctions for capacity for the next gas year (starting October) are held in March if you want annual capacity strips, or June if you want quarterly strips, but the final tariffs which will apply do not have to be published until August or September. The tariffs that TSOs publish are the reserve price in the auction. If the auction clears at the reserve price you pay the tariff; if it clears above the reserve price you pay the reserve price plus a premium. However, as the final tariff is not set until after the auctions have taken place, you will not know what the final reserve price is, and therefore the total tariff that you will pay for the capacity..

The regulators have claimed that tariffs will not vary much from year to year - at the recent Madrid Forum it was claimed that they would only vary by 2% year on year. However it is not clear on what information this is based as tariff increases for this year (2014) compared to 2013 have already shown much greater variation, and it also assumes that shippers’ booking behaviour will not change as CAM is fully implemented. It is a reasonable expectation in economics that if you change either the way a good is priced, or its attributes, or a combination of both, then how much of it is bought may also change.

How can regulators promote more transparency and more rational reserve prices? How can regulators make the system more efficient?

There are two key ways that regulators and TSOs can improve matters - although it will not solve the underlying uncertainty of floating tariffs.

Firstly, TSOs should be required to publish the models they use to calculate the tariffs at the entry and exit points on their system. Shippers could then use these models, together with their own assumptions as to what levels of capacity bookings will be made and how much revenue the TSOs will be allowed to earn, to make their own forecasts of how tariffs will evolve over time. This will not give certainty, but it may give shippers sufficient comfort to book capacity on more than a month ahead basis.  

Secondly, TSOs and regulators could ensure that the reserve price that is published before the auctions is the one that remains in place for the forthcoming capacity year. This would mean TSOs and regulators calculating the final reserve price earlier, or changing the auction timetable, or a combination of the two. The TSOs and regulators have been reluctant to do the first as they say that they will be basing the reserve price on estimates of revenues – the later they calculate the reserve price, the more accurate the estimates will be. This is true, but has to be weighed against the problems caused to the efficient function of the auctions by not having a clear reserve price.  

Also it is often the case that TSOs and regulators are not sure of their final revenues when setting tariffs or price controls. It is common to use so called “correction factors” which enable additional revenue to be earned in future periods to make up for less than expected revenue when tariffs are set. In this case it would be possible to use a rolling approach whereby a best estimate of revenue in Year 0 would be used to set the reserve price in Year 1. For Year 2 the reserve price would be based on an adjustment to allow for a difference between the Year 0 estimate and the final Year 0 revenue, and an estimate of the Year 1 revenue. This would continue into future years with Year 3 reserve price based on any adjustment due to Year 1 revenues, and an estimate of Year 2, and so on. 

How do these inefficiencies impact competition? Do they lead to barriers for the entrance of new market players? 

The whole idea of having capacity auctions is so that those shippers who value capacity the most will be allocated it if they pay the highest price in the auction. Obviously shippers’ bidding strategies depend on the reserve price for the capacity, and how much they are prepared to bid above it if demand for capacity exceeds supply. Even if demand does not exceed supply, and therefore the auction will be expected to clear at the reserve price, it is difficult to see how a shipper can decide how much capacity he will want to buy, as he does not know what his final bill will be. You cannot value something, and make a rationale decision as to how much to buy, if you don’t know the price you will be paying. Not knowing the reserve price for the following year, let alone the payable price for the years after that, severely undermines the rationale for capacity auctions and hence competition in the European internal gas market.

It does create barriers to entry as new entrants, those who may buy capacity at only one or two points, face the problem of greater uncertainty about their tariffs compared to, for example, large players who book capacity across a greater number of points. The latter group may benefit from an averaging effect (i.e. some tariffs go up, some go down) whereas this will not be the case for those who have capacity at only one or two points. Furthermore, the lack of predictability of tariffs will make it more difficult to enter a market compared to a company that already operates there. In the former, imagine trying to obtain board approval to enter a market if you do not know what it will cost to do so, and there is no means of forecasting what costs (i.e. capacity tariffs) might be. For companies already in a market it is more a question of whether they wish to continue to participate when there is such uncertainty about costs, or how they can change their approach. 

Keeping in mind that tariff changes are already volatile with significant tariff increases observed at the beginning of the year, will these inefficiencies have an impact on the gas industry as a whole? Will it potentially have a backlash on gas as a transition fuel? 

Tariff uncertainties could have a number of potential impacts, depending on how shippers react to the new rules. Some of the regulators have assumed that shippers booking behaviour will not change, but I believe this to be unrealistic and not borne out by experience, for example in Germany and the UK. Where there is plenty of capacity available there is a tendency to book more on short term basis (i.e. booking capacity just before you need it) and to profile capacity. Given that a lot of gas demand is driven by weather, or by other variable such as the spark spread, this could lead to more variable levels of capacity booking over the year, and hence more volatile earnings for TSOs. However there is little incentive for shippers to book longer term because of the uncertainty of floating tariffs. Lack of long term bookings may also lead to regulators questioning the viability of TSOs’ asset bases, on the grounds that if there are not long term bookings, there is perhaps no need for the capacity.

The factors determining gas’ role as a transition fuel are different from tariffs and are more related to EU energy policy on renewable targets and the price of carbon as revealed by the ETS. However it is fair to say that the inefficiencies in the proposed tariff code will not help create a properly functioning internal European energy market. 

Both regulators and shippers have an interest in promoting longer term bookings. Multipliers, which are part of the network code for tariff harmonisation, should help in this process. What is the main problem with multipliers? What to expect in the next months?

Multipliers help promote long term bookings because they make short term capacity bookings (e.g. daily or monthly) more expensive than long term bookings e.g. annual. The economic justification is easy to understand in the following way. Imagine a world where there is one network and one shipper. The shipper books capacity for one day and pays 1/365 of the price of an annual strip of capacity. However, even though the shipper only needs capacity for one day, the network has to be there for the other 364 days - generally speaking pipelines cannot be built and dismantled over night. Therefore the true cost, and hence true tariff for the shipper should be the same as the annual tariff in this example. 

In the real world pipelines are built to be able to serve winter peak demand which inevitably means that capacity which is needed for the winter is often not booked or used during the rest of the year. If shippers who book capacity for only a few peak months a year pay a tariff based on the number of days booked multiplied by 1/365 (i.e. a multiplier of one), they are effectively underpaying for the capacity. However the difficulty in setting the multipliers is knowing how much capacity may be booked, for example during off peak periods of the year. Weather is not the only determinant; spreads between markets will also influence capacity bookings, and we have already seen changes in flow patterns for gas as traders move gas between differently priced markets. If TSOs set multipliers too low they potentially favour companies which book capacity short term over those who book long term, and exacerbate the floating tariff problem because they under-recover their allowed revenue. If they set multipliers too high they favour shippers which book long term, and potentially deter cross border arbitrage between markets by setting the cost of capacity too high.

In general, there is much confusion around the tariff network code which should be defined by ENTSOG by the end of the year. What are the other bones of contention - apart from multipliers? Do you expect ENTSOG to complete its task by the end of the year? 

ENTSOG have to send a draft code to ACER by the end of the year as this timetable is dictated by the legislation. However that is not the same as saying that ENTSOG will be able to deliver a code that is fit for purpose. The main problem is that the ACER Framework Guidelines which determine the scope of the code were very prescriptive, laying down not just the areas which ENTSOG should cover but also what the code should say. Although ACER did hold consultations on the Framework Guidelines, tariffs is a very complex subject with many trade-offs. The best way to choose between these trade-offs, which are not the same for all the different countries which will be covered by the code, is to work through all the problems and see which combination of trade-offs best fits the objectives of the code. No-one, whether it be shippers, TSOs or regulators, has a monopoly of understanding, and the Stakeholder Workshops held by ENTSOG earlier this year were meant to be a means for working through the issues. Unfortunately ACER declared at the start that a number of key aspects of the code were not open for discussion. The problem with the reserve price being set after the auction has taken place is a classic example of something that could have been avoided if proper discussion and development had been encouraged by ACER.

There are four key problem areas. Firstly, the problems that the code is trying to solve are not clearly defined and therefore it is not possible to know if the solutions are the best ones. For example ACER has said that the code is needed to improve cross border trade, and therefore we need greater harmonisation of tariffs across the different countries. However North West Europe has lots of cross border trade and contains the most liquid European traded hubs. It does not however have harmonized tariffs, nor does it rely on floating capacity tariffs. The question therefore is why ACER believes its approach is best, when the North West European experience would seem to indicate that other factors may be more relevant.

The second problem?

ACER’s insistence on floating tariffs. As discussed above the lack of predictability of the actual price shippers will pay for capacity make it less likely that they will book long term, which in turn exacerbates the problem of floating tariffs. Furthermore floating capacity tariffs have a more pernicious effect than the use of a “top up” commodity charge (i.e. charge based on flow of gas rather than capacity booked) as a means of ensuring the TSO reaches its allowed revenue. With a capacity charge the only way you can mitigate an unexpected increase in the capacity charge (which makes  the cost of transporting gas greater than the spread between the markets), is not to book capacity until the last minute. With a commodity charge approach you can book the capacity (thereby giving useful signals to TSOs that such capacity is needed) and decide on the day whether or not to flow gas depending on the relative cost of the commodity charge and the market spread (as the capacity charge is a sunk cost). 

The regulators’ thinking may be influenced by concerns that commodity charges discourage cross border flows when they exceed the spread of prices across markets. By focusing all recovery of TSOs revenue via a capacity charge the thinking is that, because capacity costs are sunk costs, then cross border flows will take place so long as there is a positive spread, no matter how small. This is true, but ignores the effect it has on shippers’ willingness to book capacity long term and thereby give appropriate economic signals as to future requirements for capacity. There is a risk that, if TSOs do not have long term bookings, regulators may regard the capacity as no longer necessary, and therefore exclude it from the TSOs asset base used to calculate allowed revenue. Also some regulators seem to ignore the fact that commodity charges are just another way of charging the cost of the network to those who use it. It is not a bad economic result if trade does not take place if the cost of transporting a good is greater than the additional value that is generated by selling that good in a different market. Floating capacity charges simply put more risk on those who book capacity across several years.  

There is also the argument that networks should be paid for via capacity charges because the vast bulk of the network costs are fixed and are based on the size (capacity) of the network, not on variable (commodity charges). This is true but it ignores the fact that floating capacity charges and the ability to book capacity short term have the effect of “commoditizing” capacity and making it more like a variable cost.

ACER has presented the choice as between floating capacity charges and nominal fixed capacity charges, and say that floating capacity charges are best. There are merits for floating capacity charges but the ACER comparison is too simplistic. They have assumed that the value of a network will always be growing which may not be the case in a world where EU policy is explicitly designed to reduce fossil fuel use. Furthermore they have not considered options such as capacity charges fixed in real (as opposed to nominal) terms, the use of premia for fixed charges (as suggested by ENTSOG), or the use of commodity charges. 

The third? 

Transparency. However charges for use of the network are levied on users, it is important that users can understand how tariffs are calculated and how they can change over time. For this users need to have access to the same models that TSOs use to calculate tariffs so that users can input their own assumptions about capacity bookings and TSOs allowed revenues. This will enable users to have their own views on tariff evolution and make informed decisions when booking capacity. At the moment far too many TSOs tariff calculations are made via a “black box” and all shippers know are the tariffs themselves when they are published, often at short notice. However ENTSOG did announce at the Madrid Forum that they will be working to improve transparency so it is more in line with that practised by National Grid in the UK which is already very transparent.

There is an additional problem that the code may not apply to certain charges that make up the final total tariff that users pay. For example some countries charge additional levies on top of the usual capacity and commodity charges. In Germany users are charged for the transformation of German Low Cal gas networks to be able to take High Cal gas. However this charge is levied on all exit capacity, both H Cal and L Cal, including those users who transport H Cal gas across Germany to other markets such as France. This is clearly against the Third Package which says that tariffs must be cost reflective. Users who transport H Cal gas, and especially those who transport it to other countries, are not the ones creating the need to convert the L Cal networks. It is a pity that ACER and the EU Commission do not focus more time and effort on anomalies such as this one. 

The fourth?

Tariff Methodologies. The apparent aim of these is to ensure all TSOs calculate their tariffs in a standardized way to ensure that the internal market works as it should. However there are a number of problems with the approach taken. Firstly, there has been no proper analysis of how the proposed tariff methodologies will change what TSOs currently do. All that has happened is that local NRAs and TSOs have said that their current methodologies fit with the proposed standard methodologies. Given the lack of transparency as to how most tariffs are calculated (see above) shippers have no way of gauging what the impact of the new code will be. Secondly it begs the question as to the value of the Code methodologies since, on the face of it, they are not going to change anything. In any case the methodologies in the code are so general it is not clear what value they add. Thirdly the draft code requires that TSOs show that the methodology they choose is the best one by comparing it to one of the other methodologies (the “counterfactual”). However TSOs can choose which counterfactual they use, which means that there will be no standard comparison across different TSOs. TSO A can choose Methodology 1 as his counterfactual, whilst TSO B could choose Methodology 3 as his counterfactual. If TSOs can set their own metrics like this, it devalues the utility of the counterfactual test as they can simply choose the counterfactual which makes their existing methodology look best. 

The Third Energy Package includes 12 binding network codes on gas. These codes were supposed to be defined and implemented by 2014, but only 4 will actually be defined or implemented by the end of the year. Does it create further confusion? What are the consequences of the delay?

Given the time it takes to develop network codes, and the fact that the codes have been developed sequentially, it was never likely that the codes would be all ready by the end of 2014. However this does not mean that significant progress has not been made already. For example the PRISMA platform for capacity auctions is already very active, and a number of countries are aiming to implement the codes ahead of the deadlines.

The real cause for confusion is poorly developed codes. For example the Congestion Management Procedures (CMP) include definitions of congestion which cannot distinguish between circumstances where demand for capacity genuinely exceeds supply, and situations where capacity is being hoarded. Also the CMP allowed Germany and Austria to choose different measures from their neighbours which has made the overall system less efficient where cross border capacity is bundled. Whilst the EU Commission has tried to address these issues via a Guidance note the problem is that it is very hard to change the legislation - which includes the codes - once it is passed. The stakes for the Tariff network code are much higher because its potential impact is greater. 

Taking a step back, on 14 October 2013, the Commission adopted the first EU-wide gas network code on the allocation of capacity in gas pipelines (CAM). What will be the impact of the new code that will apply from 1 November 2015?

CAM has already made an impact as a number of countries have implemented all or part of the CAM via capacity auctions on PRISMA. the implementation of CAM enables shippers to profile their capacity bookings to match their expected flows which leads to a change in their booking behaviour. 

The European Commission said the network code aims at ensuring more efficient allocation of the capacity. It also said it will promote transparent third party access. Do you agree? What are the main problems with the CAM? What will be the difficulties for operators in the next 12 months?

CAM does create a level playing field for capacity allocation and does promote third party access. Overall CAM works but it has some unnecessary flaws. The principal problem is the idea of mandatory capacity bundling whereby shippers book capacity from hub to hub, rather than booking separate exit and entry capacity between market zones. In theory bundling should make it easier to trade across borders because it means shippers only have to book one product rather than two. However in practice shippers are still booking two different products (exit and entry capacity) with two different TSOs and with two different sets of contractual terms and conditions. All that happens with bundling is that you book both products at the same time. However the failure to harmonise the underlying products means that bundling doesn’t help that much. Furthermore the way that different countries have implemented third party access (CMP, different technical capacity levels on either side of the border) means that bundling forces neigbouring countries to take a “lowest common denominator” approach. On balance it is not clear that bundling has improved things much - after all North West Europe trading is succeeding without bundling - and it actually makes it more difficult to solve other problems. A better approach would be to enable bundling where the market demands it but not to force the issue where there are other problems. 

The other problem with CAM is that it has no mechanism to allow new capacity to be built where an auction shows that demand exceeds supply. The Incremental amendment is aimed at solving this, and could work were it not for floating tariffs. Floating tariffs mean that shippers will not know what they will have to pay when they make the long term bookings necessary to trigger new investment via the Incremental process. This makes it less likely they will make the long term bookings without which the new Incremental process will not work. 

In conclusion, is there any good news for operators? Do you expect any significant progress in the coming months? Do you agree with experts seeing an increased appetite for long term investments in the European gas industry? 

If you judge the Third Energy Package on the EU Commission’s criterion as a means to ensure a properly functioning energy market, overall it is good news. Trading and cross border flows are increasing and the new rules such as CAM and CMP will make this easier. There have also been improvements in terms of transparency of information which is necessary for market players to make informed commercial decisions. Gas definitely has a very important part to play in Europe’s energy future, and therefore the industry should be attractive for long term investors. The issue is more about developing the rules which work as well as possible to make it as easy as possible for the market to work properly, and for players to invest. 

Sergio Matalucci 

Sergio Matalucci is an Associate Partner at Natural Gas Europe. Follow him on Twitter: @SergioMatalucci