Long term contracting to support asset financing

Posted on October 24, 2016January 16, 2020 by David Duncan

Infrastructure investors are set to feature as prominent buyers of conventional supply assets, as utilities ramp up asset sales in Europe. European utilities have traditionally had the capacity to hold power stations and midstream gas assets on balance sheet. But infrastructure and private equity funds have a much greater focus on creative financing in order to maximise return on capital.

The financing of asset purchases is set to play an increasingly important role in determining transaction structures and prices. A number of assets tagged for sale come with a significant merchant (or market) exposure. Financing of these assets will depend to a large extent on the ability of buyers to secure long term contracts.

Traditionally it has been tough to secure financing without long term (10-15 year) fixed price contracts in place to cover the majority of market risk. However the interaction between Long Term Contracts (LTCs) and financing is evolving, driven by transaction deal flow, development of new assets and the refinancing of existing assets.

Today we explore the interaction between long term contracts and financing in this our final article in a series on the long term contracting of flexible supply assets.

Contracting to facilitate project finance

There are three key counterparty types that feature in a typical conventional supply asset transaction: equity investors, lenders and capacity offtakers (via LTCs). Each of these parties has conflicting interests:

Lenders are by nature risk averse and primarily focused on ensuring adequate cashflow to cover interest and principal repayment costs. They do not want the equity investors to default, but aside from that they have little interest in asset value upside.
Offtakers are primarily focused on negotiating a competitive price for capacity, while contracting as much value upside in the form of asset optionality as possible. They also have a strong incentive to push availability risk onto the asset owners (given an inability to manage it).
Equity investors: need to balance their primary goal of achieving a reasonable expected return on equity, against facilitating lenders with downside protection and managing value upside & availability risk via the structuring of offtake contracts.

LTCs facilitate asset financing via the protection they provide lenders on cashflow to debt. But contracting acts to narrow the distribution of asset earnings in order to achieve downside protection i.e. it reduces expected asset returns and limits value upside to equity. We illustrate this in Chart 1 by returning to our example of a tolling contract on a UK CCGT asset to illustrate the principles involved.

Chart 1: Illustrative annual project cashflow distribution

Source: Timera Energy

Chart 1 illustrates the impact of a long term offtake contract in narrowing the asset’s margin distribution. Lenders are focused on the left tail of this distribution. They want to be protected from market downside risk even in the case of more extreme outcomes i.e. lenders want to ensure that the left hand tail of the margin distribution does not extend down into the debt service zone.

As equity investors negotiate offtake terms, they are typically concerned with three important considerations relating to the margin distribution illustrated in Chart 1:

Expected return: The level of expected asset margin above debt service and fixed costs determines the expected return for equity investors.
Covering fixed costs: The left tail of the margin distribution determines the likelihood of equity investors suffering negative cashflows.
Upside access: The right tail of the distribution defines upside potential, with investors being particularly interested in any asymmetric value upside that acts to skew the margin distribution to the right.

The considerations set out above are driving the evolution of financing structures in European power and gas markets.

Financing innovation to incorporate market risk

Lenders have traditionally shown a strong aversity to market risk exposure. This has driven a requirement for high levels of long term contract cover to insulate asset margins from market risk, at least until the majority of debt has been paid down. But record low interest rates and a search for yield is pushing lenders to take a more open minded approach to financing structures.

Hybrid financing structures are now being developed that allow equity investors to retain a certain level of market risk. The key to these structures is giving lenders enough security on cashflow to service debt. Table 1 below sets out a high level example of terms under this hybrid approach, versus the more traditional 100% LTC model.

Table 1: Traditional vs hybrid financing structures

Traditional 100% contracted model	100% capacity sold via fixed price LT contracts (e.g. 10 year) Debt term to match contract (e.g. at an interest rate of 4%) 1.3 x Debt Service Coverage Ratio (DSCR) requirement DSCR driven by fixed long term offtake contract price Up to 60% project leverage achievable
Hybrid 50% contracted model	50% capacity sold via fixed price MT contracts (e.g. 5-7 years) Debt term to match contract (e.g. at an interest rate of 6%) 1.3x DSCR requirement DSCR tested against a downside market case 20-50% project leverage

Source: Timera Energy

The traditional model usually requires a relatively high level of intrinsic asset margin in order for equity investors to be able to secure an adequate fixed price on a long term contract. However extrinsic value makes up an important portion of overall margin for thermal power and midstream gas assets under current market conditions. This makes the traditional financing model difficult to achieve given the haircuts incurred when contracting extrinsic margin.

The hybrid structure opens up opportunities to finance assets that have lower levels of intrinsic value. This is particularly important for thermal power stations and midstream gas assets which are suffering from cyclically depressed market conditions. The issues involved are best illustrated via some practical case studies.

UK-Continental interconnector

Let’s start with an asset that does benefit from substantial intrinsic margin. The premium of UK over Continental power prices means the owners or developers of interconnectors can sell long term capacity contracts at healthy price levels. This facilitates the traditional approach to asset financing. A base tranche of capacity can be sold to provide lenders with cashflow protection to service debt. As a result interconnector projects can be financed with a high ratio of debt to equity (gearing).

New UK CCGT assets

One of the drivers of the evolution of hybrid financing structures is the requirement for new CCGT capacity in the UK power market. With system reserve margins at historical lows there is a clear requirement for new capacity. However it is difficult to source long term (10+ year) tolling contracts in the UK to support traditional financing.

Project developers are focusing on using the 15 year capacity agreements as protection to secure financing under the hybrid approach set out above. Lenders appear to be comfortable with lending up to the level of capacity payments, but have a strong preference for further margin protection above this e.g. in the form of a shorter term toll or strong equity buffer. As a result, project debt levels are likely to be relatively low. This means that CCGT financing will depend on an equity structure that is prepared to take on the remaining market risk (above and beyond the level of any toll).

Continental CCGT

We finish with an example of an asset type that is likely to require pure equity financing under current market conditions. Given the level of CCGT generation margins on the Continent, it is very difficult to secure a tolling contract at a price that supports any debt. The other key challenge in Continental markets is an absence of the capacity price support that is available in the UK.

The exception to this logic is if an asset has legacy long term contracts in place for power, fuel or steam (e.g. CHP plant with steam and onsite offtake contracts). In this case these contract cashflows can be used to service debt.

These three case studies illustrate the financing challenges facing the buyers and developers of flexible supply assets in Europe. These are relatively simple examples in what is an ongoing process of evolution of contracting and financing structures that are more tolerant towards market risk. This evolution is being fueled by the scale of European asset sales coming to market and the new sources of capital competing to invest.

Article written by David Stokes and Olly Spinks

The impact of rising coal prices

Posted on October 17, 2016January 16, 2020 by David Duncan

2016 has delivered its fair share of commodity market surprises. But none have been more unexpected than the sharp recovery in coal prices. European benchmark ARA coal prices have increased by more than two thirds since January, from below 45 $/t to more than 75 $/t.

This coal price rally has been the principal driver behind the pronounced rally in European power prices since Q1 2016. While coal prices have increased sharply, European gas prices have remained relatively weak, supporting a competitive shift across Europe towards gas-fired generation.

In this article we take a look at the drivers behind the rally, the shape of the current coal forward curve and some implications for European gas and power markets.

What is behind the rally?

Some important context precedes this year’s recovery. Coal prices halved across 2014 and 2015 as shown in Chart 1. As is often the case with the coal market, events have been evolving around China.

Chart 1: Evolution of key global coal price benchmarks

Source: Timera Energy (based on ICE futures settlement prices, forward prices as at 6th Oct)

China experienced negative coal demand growth in 2014 and 2015, as economic growth slowed and government measures were implemented to reduce air pollution. Over the same period, growing domestic production overcapacity in China helped to aggravate a global supply glut.

The sharp move higher in prices in 2016 has been supported by a reduction in production overcapacity on two fronts. Firstly, the Chinese government is taking measures to close 500 million tonnes of production capacity over the next 3-5 years (~15% decline). Secondly, lower prices have been driving a supply side market response from global producers in the form of mine closures and mothballing.

As well as these more structural drivers, there are some shorter term factors behind the price rise. Heavy rain has temporarily disrupted production in some big producer countries (e.g. Indonesia & China). There has also been a big squeeze in the coking coal market as global steel production has recovered and this has fed through into thermal coal prices.

As is often the case in commodity markets, the big spot price rally has dragged the coal forward curve higher. But Chart 1 also shows a strong current backwardation in the coal curve. This is consistent with market expectations that some of the shorter term constraints of 2016 will ease into next year.

Implications for European power and gas prices

Despite current backwardation, coal prices for 2017 still remain around 65 $/t. The rally in the coal forward curve has supported a 40% rise in 2017 power prices in Germany, which broke back above the 30 €/MWh last week (from around 21 €/MWh in Q1 this year). Similar power price rises have occurred across other Continental European power markets where marginal pricing is dominated by coal plants (e.g. France, Netherlands).

European gas hub prices on the other hand have remained relatively weak in 2016, held down by robust pipeline volumes and the global LNG glut. Up until 2016 coal and gas prices had declined in a correlated manner. But this year’s price divergence has sharply reduced the competitive advantage that coal-fired power plants have enjoyed over CCGTs for most of this decade.

This relative change in gas plant competitiveness is having an important impact on the European gas market, as well as on power markets. Gas-for-coal plant switching is playing a key role in influencing marginal hub prices in Europe. The 2016 rise in coal prices has increased the gas price levels at which switching takes place.

The gas price levels at which switching takes place are an important factor determining European hub price support, also influencing spot LNG prices (given their linkage to European hubs). For a given level of gas prices, a rise in coal prices increases gas demand from CCGTs. Or to look at it another way, if coal prices had not risen in 2016, lower gas price levels would likely have been required to induce the CCGT load factors required to absorb surplus hub gas.

Higher prices driving coal plant out of the capacity mix

There may also be an important longer term impact of the recent reduction in coal plant competitiveness. Rising coal prices are materially eroding coal plant generation margins. This comes at a bad time for plant owners as it coincides with an increasing European policy shift against coal generation as efforts increase to tackle emissions.

The 2016 coal price rally appears to be the nail in the coffin for most of the UK coal plant fleet, which is penalised by the additional burden of the carbon price floor. France looks to set to introduce a similar carbon penalty on its coal generators in 2017 which should induce a similar result.

More broadly across Europe, accelerated coal plant closures are likely to be an important topic of discussion in utility boardrooms. If Europe wants to decarbonise, cheap gas and rising coal prices are making it easier.

Article written by Olly Spinks & David Stokes

Gas rebalancing 2: the path to price recovery

Posted on October 10, 2016January 16, 2020 by David Duncan

The first half of this decade saw an LNG investment boom. High gas prices and optimistic Asian demand projections supported a flood of Financial Investment Decisions (FIDs) in new LNG liquefaction capacity. These investment decisions are the source of the current oversupply and depressed gas prices, conditions that are set to dominate the second half of the decade.

However since 2015, investment in new supply has almost dried up as global gas prices have crashed & converged. The impact of this investment drought is being concealed by a deepening supply glut. Long delivery lead times on projects already signed off mean there is more than 150 bcma of capacity still under construction to be commissioned by 2020. But a lack of investment beyond the current pipeline of new supply is creating the conditions for a price recovery early next decade.

While investment in new supply has ground to a halt, important structural drivers continue to support global demand growth. European import dependency is increasing as domestic gas production declines. Asian demand growth, while slower than the optimistic estimates of earlier this decade, is still a force to be reckoned with. And there are tailwinds for global gas demand from a growing focus on decarbonisation and reduction in coal burn.

The global gas market may still be descending into the tunnel of oversupply. But we suspect that this tunnel is shorter than many people think. And with long delivery lead times in the gas investment cycle, light may already be visible on the other side.

3 phases of global market rebalancing

In today’s article we set out our view of the path to global gas market price recovery. We break this recovery out into three phases:

LNG glut: clearing the current global glut associated with committed new liquefaction capacity
Russian pricing power: absorbing approximately 100 bcma of ‘shut in’ Russian production capacity
New supply: A global requirement for incremental new production capacity

Last week’s article focused on a deeper dive analysis of Phase 1 and the mechanisms that can clear the current LNG glut. Today we use the same scenario framework but adjust our perspective to focus out over a longer time horizon into next decade.

Chart 1 shows the global market volume balance diagram from last week, overlaid on a schematic illustration of price dynamics for each of the three phases.

Chart 1: Schematic illustration of 3 phases of price recovery

Source: Timera Energy

Phase 1: LNG glut

Phase 1 represents the world we are already in. Downward price pressure from a global surplus of LNG is forcing the convergence of Asian and European prices towards US Henry Hub support. Regional price spreads are increasingly converging to levels driven by the variable cost differentials of moving gas between regions. Phase 1 conditions are summarised in Table 1.

Table 1: Summary of Phase 1 – LNG glut

Market balance	Volume response	Price dynamics
Global oversupply, as committed new LNG liquefaction capacity outstrips demand growth	4 market clearing mechanisms act to clear surplus LNG (see description last week). Key balancing role for European power sector switching & US LNG shut ins	Asia, Europe & US prices converged to variable transport costs. Potential for oil vs gas price divergence if oil market recovers ahead of gas.

We set out last week the 4 key market clearing mechanisms likely to provide the incremental volume response to clear the LNG glut (European switching, Asian demand response, US & Australian LNG shut ins). If you believe Russia will materially change its strategy in response to lower prices then consider this is a 5^th mechanism. The influence of these clearing mechanisms erodes the influence of long term oil-indexed gas contracts and shifts the focus towards Atlantic basin hub prices (e.g. NBP/TTF and Henry Hub).

The increasing importance of hub price signals and a rapid rise in flexible US export volumes, is set to support an increase in global gas market liquidity. Oversupply itself is a great catalyst for the development of market liquidity, given the requirement to clear surplus gas volumes via spot markets (e.g. as seen in Europe in 2008-10). These conditions set the stage for a declining influence of long term oil-indexed contracts and a rising importance of short to medium term hub linked deals.

Oversupply and maturing gas market liquidity also support the potential for a structural divergence of oil and gas prices. If the oil market stabilises and recovers before the gas glut is absorbed, then gas oversupply is likely to dampen the influence of oil-indexed contract prices in driving a parallel gas price recovery.

Phase 2: Russian pricing power

Phase 2 commences once the LNG glut has been absorbed and the world needs incremental supply. In our view that does not mean a sudden return to Long Run Marginal Cost (LRMC) driven market price signals, as many analysts assume. The reason for this is an existing surplus of Russian gas, over and above current contracted volumes. Phase 2 conditions are summarised in Table 2.

Table 2: Summary of Phase 2 – Russian pricing power

Market balance	Volume response	Price dynamics
Global gas glut absorbed Europe and/or Asia need incremental gas supply Asian demand growth may pull LNG away from Europe, to be backfilled by Russia	100+ bcma of ‘shut in’ Russian gas well placed to meet incremental demand US exports flow to balance Asia & Europe	Russian flows have a strong influence on marginal pricing, reviving the influence of gas price linkage to oil Asian & European prices remain converged but with rising short term volatility

The key dynamic of the 100+ bcma of ‘shut in’ Russian gas is that it can be flowed into Europe based on Short Run Marginal Cost (SRMC) price signals. This shut in gas is located in West Siberian gas fields developed by Gazprom in anticipation of higher European demand growth. Loss of Russian market share from Gazprom to other Russian ‘independents’ has also contributed to the volume of shut in gas.

Prior to the building of a possible future Altai pipeline, this shut in gas is entirely in Europe-facing Russian fields. But it can also indirectly satisfy Asian demand growth by allowing flexible LNG flows (e.g. US exports) to be diverted to Asia, while Europe ‘backfills’ these with incremental Russian gas.

Gazprom has historically chosen not to flow this gas at price levels below existing long term oil-indexed contract prices. To do so would act as a catalyst for hub versus contract price divergence and development of hub liquidity, both of which Gazprom considers to be against its strategic interests.

This surplus of ‘shut in’ gas puts Russia in a very strong pricing position once the current LNG glut is absorbed. As long as Russia sells this gas at a sufficient discount to new LRMC, it can block new supply (e.g. in the form of new ‘second wave’ US export projects).

It is unlikely that Gazprom will significantly undercut its existing oil-indexed contract prices. This points towards a resurgence in the influence of oil-indexed pricing on hub prices (as has been experienced across much of the last 20 years).

It also sets up the conditions where European gas prices can diverge from Henry Hub. High volumes of flexible US export flows (80+ bcma capacity by 2020) are likely to ensure Asian and European prices remain structurally converged to variable transport cost differentials. But a tightening global market may result in greater short term inter-regional price volatility.

Phase 2 represents a key step on the path from SRMC to LRMC driven market price signals. While Russia is in a position to delay new LRMC driven supply, this is only a temporary situation. 100 bcma of Russian gas is likely to satisfy only 3 to 4 years of incremental supply requirements, less if you assume more robust global demand growth. Beyond that, the LRMC of new supply is set to reassert its influence on gas pricing.

Phase 3: New supply

Phase 3 is about the transition to investment in new LNG production capacity. This may seem a long way off (mid 2020’s in our illustrative scenario). But project delivery lead times are typically around 5 years. That means producers needing to convince themselves of a price recovery to cover LRMC, 5 years in advance of new supply coming online. The dynamics of Phase 3 are summarised in Table 3.

Table 3: Summary of Phase 3 – New LNG supply

Market balance	Volume response	Price dynamics
Russian ‘shut in’ gas volume absorbed Incremental global gas supply required	LRMC competition to provide new supply 5+ year delivery lead times on new projects	Global prices rise to support new project LRMC Structural European and Asian price convergence, but with significant shorter term regional price volatility

The sources of new supply are not yet clear. But there appears to be a cluster of potential options in a 9-11 $/mmbtu LRMC range. These include ‘second wave’ US export projects, new Russian supply & Non-US LNG projects (e.g. East Africa & Canada). Wherever new gas comes from it will require a price signal from Europe and/or Asia.

As for Phase 2, market tightening is unlikely to result in structural Asian vs European price divergence (e.g. as seen post-Fukushima). By the end of this decade there will be large volumes of flexible supply that can arbitrage any structural price differences, including US export volumes and other LNG supply contracts with diversion flexibility.

Asian and European prices may diverge from Henry Hub in order to provide a market signal for new supply. But the extent of any structural diversion should be limited by the costs (fixed & variable) of developing new US export capacity.

The conditions for short term regional price volatility in a tightening market remain. There are structural lead times, often 2 to 3 weeks, for the LNG supply chain to respond to shorter term price regional price divergences. This volatility will be an important price signal for LNG portfolio supply flexibility.

What does the path to price recovery look like?

Breaking gas market rebalancing into phases of recovery helps focus in on the transition between the different drivers of global gas prices. The dominant driver of marginal pricing across the different phases is as follows:

Phase 1: Atlantic Basin hub prices (Henry Hub, NBP/TTF)
Phase 2: Russian pricing power and the revival of oil-indexation
Phase 3: LRMC based competition to provide new supply

In Chart 2 we show how the three phases come together, in an illustrative scenario of the evolution of global gas prices to 2030. It is not a price forecast. The intention of this scenario is to present a reasonable view of the relationship between LNG market balance and regional price evolution through the three phases. To the extent your views differ on key assumptions such as demand growth, market balance and Henry Hub price evolution, the scenario can provide a useful point of contrast.

Chart 2: scenario projection of global price recovery (2016-30)

Source: Timera Energy

The scenario shows the convergence of regional prices until 2020, with Henry Hub providing key global price support in an oversupplied global market. Asian and European prices then undergo a significant recovery in the early 2020s as the LNG glut is absorbed. This is driven by the ability of Russia to exert its pricing power to lift prices back towards LRMC. This is likely to coincide with a revival of the influence of oil-indexed contracts on European hub prices.

New supply is required from the mid 2020’s. In practice this means upstream investors needing to convince themselves of a price signal to cover LRMC costs of new production, 4 to 5 years in advance. This LRMC price signal may start to emerge via long term contract prices underpinning investment in new supply, in advance of spot prices returning to LRMC levels. Or it may be that producers have to bear significant price risk in anticipation of market recovery.

Long investment lead times can actually result in the boom to bust cycle of this decade working in reverse. Price recovery may seem a distant prospect from the depths of the supply glut in 2018-19. But if new supply is not FID’d shortly after, the global market may once again be very tight by the mid-2020s. These boom/bust dynamics are an inherent characteristic of long upstream delivery lead times.

The timing of phase transition is a key source of uncertainty. Our scenario in Chart 2 is based on reasonably conservative demand growth assumptions. If global demand recovers more quickly, then the phasing logic is accelerated. Similarly, if demand growth is weak and/or investment in new supply comes too early, then there may be a slower progression through the phases.

But there are some important pricing dynamics that are likely to emerge regardless of phase timing:

Europe will play a pivotal role in clearing the global gas market, given its liquid hubs, alternative supply sources, power sector switching potential and supply contract flexibility.
European and Asian prices are likely to remain structurally converged, given a rapid growth in volumes of flexible LNG supply (driven particularly by US export growth).
Short term inter-regional price volatility will not disappear, given there are structural lead times for the LNG supply chain to respond to market price signals.
Gas market maturity will erode the dominance of long term contracts, with a shift in focus to managing exposures over shorter time horizons against liquid hub price signals.
Intermarket linkages are becoming increasingly important, with growing connectivity between gas vs coal prices and US vs Europe vs Asia gas prices (and these are likely to grow in importance relative to the traditional gas vs oil price relationship).

The current gas glut is acting as a catalyst to support the evolution of these dynamics. Rebalancing, price recovery and the requirement for new upstream investment may be closer than you think. And the path to recovery is likely to drive a transformational evolution of the global gas market from the interlinked regional markets we know this decade.

Article written by David Stokes, Olly Spinks and Howard Rogers

Client briefing pack

Timera Energy has published a client briefing pack ‘Global Gas Market – the path to market recovery‘. This includes an overview of current global pricing dynamics, how the LNG glut will be absorbed and the market evolution into next decade. You can download the briefing pack by clicking on the title link above or going to Our Publications.

Gas rebalancing 1: Clearing the global gas glut

Posted on October 3, 2016January 16, 2020 by David Duncan

The global gas market has been shocked by the pace and scale of transition from boom to bust. At the beginning of 2014 suppliers were paying a lofty 20 $/mmbtu for LNG in a market that was anticipated to remain tight for years. By early 2016 producers were struggling to sell surplus cargoes for 5 $/mmbtu, 75% below the levels of just two years earlier. A bust of these epic proportions was not supposed to happen.

The initial shock has now passed and gas market consensus has digested the phenomenon of a state of global oversupply. But global demand growth remains sluggish and supply continues to grow. More than 50 bcma of new LNG liquefaction capacity has been commissioned since 2014. In addition, there is a visible pipeline of at least another 150 bcma of committed new capacity coming to market by 2020.

But as the dust from the bust starts to settle, attention is shifting to some key questions. How will the global market clear the evolving gas glut? And how will this impact the evolution of pricing dynamics and the gas market investment cycle?

Almost everyone in the gas and power industries has a vested interest in the answers to these questions, either explicitly or implicitly. In the next two articles we set out our take on the answers.

Global market rebalancing series

Rebalancing of the global gas market is a complex area with a heavy dose of uncertainty. But it is our view that the drivers behind market rebalancing can be broken down via a relatively simple logic. Over the next two weeks we set out our thesis on:

The market mechanisms that will clear the current gas glut
The path to global gas price recovery and why this may start sooner than expected

We will do this as much as possible using a practical scenario illustration of the evolution of global supply and demand volumes and regional gas prices.

In today’s article we focus on the key price/volume mechanisms that we believe will interact to clear the current market glut.

Evolution of the global market balance

In a nutshell, the current global oversupply of gas is the result of new LNG supply outpacing demand growth. Investment decisions in new liquefaction capacity earlier this decade were based on overly optimistic forecasts of demand growth, particularly in emerging Asian markets. The legacy of these decisions is still feeding through in the form of committed new supply, given relatively long delivery lead times for new liquefaction projects (around 5 years).

This overhang of surplus LNG is the primary driver behind the current global oversupply of gas. There are important domestic gas market dynamics at work behind this within regional markets across Asia, Europe and North America. But understanding how the global surplus of LNG can be cleared is a key starting point.

Chart 1 illustrates the LNG supply glut in the context of a scenario of the evolution of the global LNG supply and demand balance to 2030.

Chart 1: LNG supply glut and scenario evolution of global market balance (2016-30)

Source: Timera Energy

There is a cascading logic to the progression of the charts:

Top chart: In the top chart we show a combined LNG market balance for non-European LNG importing nations. These consist of Asian importers plus ‘other’ smaller importers, predominantly in South and Central America. ‘Asian and other’ buyers typically have first call on global LNG supply, given demand is dominated by less flexible buyers with long term contracted volumes and LNG specific import requirements.
Middle chart: The middle chart shows how LNG sits in the aggregate European gas market balance. Europe is broken out separately from other markets because it has liquid hubs, alternative supply sources and relatively flexible LNG supply contract structures. This means Europe plays an important role as a swing market to help clear the global LNG balance. Any surplus or deficit of LNG from the top chart (‘Asian and other’) can be thought of as cascading down into the European market.
Bottom chart: Supply volumes in the top two charts assume LNG liquefaction capacity operates at full target production levels. Demand volumes assume ‘business as usual’ demand based on current market price levels. Any surplus or deficit of LNG that cannot be absorbed via European market swing flexibility under these conditions, creates a global LNG imbalance (surplus or deficit) shown in the bottom chart.

For simplicity the North American gas market is not explicitly shown in these charts, primarily because it is largely gas self sufficient (i.e. a net exporter of LNG). The US market however plays a very important implicit role in clearing an oversupplied global market which we set out in more detail below.

As growth in new LNG liquefaction capacity outpaces demand growth over the next three years, a growing global surplus of LNG emerges over and above ‘business as usual’ Asian & European requirements. The surplus in this scenario, peaking at 68 bcma (49 mtpa) in 2019, is represented by the red shaded ‘LNG glut’ triangle in the second and third charts.

The scale of the LNG glut is actually relatively small (e.g. versus aggregate European demand). But even small volumes of oversupply can induce big price moves in order to induce an adequate market clearing response.

It is important to note that while this glut triangle is a representation of LNG oversupply, the market will always clear. What the triangle illustrates is the incremental supply and demand response required to allow the global market to clear (versus a business as usual scenario). But what are the market mechanisms that act to clear this LNG glut?

Clearing mechanisms at work

Clearing the glut of LNG is about inducing incremental volume response. In other words, it is about inducing market participants to adjust the volume of their production and consumption decisions in response to market price signals.

In our view there are four key mechanisms that can drive this. Two of these involve incremental demand response (as a result of lower gas prices) and two involve the ‘shut in’ of price sensitive supply. The four mechanisms are summarised in Table 1 below, along with a range of potential volume response.

Table 1: Summary of four key clearing mechanisms to clear the global LNG glut

Clearing mechanism	Potential volume	Volume dynamics	Key price relationships
European power sector switching	10-40 bcma	Power sector gas demand increases as CCGT load factors increase at lower hub prices	Relative gas vs coal price levels driving switching relationship between gas and coal plants
Asian demand response	5-20 bcma	Asian buyers (e.g. China, India) purchase incremental volumes as prices decline	Available LNG purchase price terms vs alternative energy supply terms e.g. pipeline contracts, oil
Shut in Australian exports	0-10 bcma	Australian export volumes may fall if netback spot LNG prices don’t cover feed gas costs	Netback cost of LNG sales vs variable cost of liquefaction feed gas
Shut in US exports	0-80 bcma	US export volumes may fall if netback spot LNG prices do not cover variable costs	Premium of Europe and Asian spot prices vs US Henry Hub (+ variable transport)

Source: Timera Energy

There is arguably a fifth clearing mechanism in the form of Russian production flexibility. The reason we have not specifically broken this out is that we assume that through the duration of the LNG glut, Russian acts to maintain its European market share (at around 150 bcma) rather than responding to prices. This is consistent with historical behaviour.

Our scenario summary on how the four mechanisms interact to clear the LNG glut shown in Chart 2.

Chart 2: Scenario projection of how 4 clearing mechanisms clear the LNG supply glut

Source: Timera Energy

European power sector switching:

Clear evidence is emerging in 2016 of European power sector demand response to lower gas hub prices. CCGT load factors in gas dominated power markets such as the UK and Italy have risen substantially year on year. Even Continental power markets dominated by cheaper coal plants have seen switching to CCGTs over the summer.

There are two important price drivers behind this switching. A weakening of gas hub prices, caused in part by a steadily growing flow of LNG imports. Coal prices have also been recovering in 2016, lowering the gas price hurdle required to induce higher CCGT load factors.

European power sector switching is currently a key driver of marginal pricing dynamics across European hubs. It will also in our view be an important clearing mechanism as the LNG surplus grows (2016-19). We set out here why we think there may be up to 30-40 bcma of potential switching demand response in Europe. Some of this switching volume may end up being permanent, with weakening coal plant economics resulting in the closer of older coal assets (as is happening already in the UK).

Asian demand response:

In theory Asian buyers should also be able to respond to falling LNG prices by increasing import demand volumes. But the practical evidence of this is so far limited. Despite a 50% fall in Asian LNG spot prices from 2014 to 2015, Asian LNG demand actually fell 4%.

Practical opportunities for physical substitution of gas for other fuels (e.g. coal and oil) are relatively limited. Some markets also face regas import capacity constraints. But most importantly, demand response is hampered by a lack of transparent market price signal mechanisms to induce the purchase of higher LNG volumes.

Asian LNG purchasing strategies tend to be driven by national policy and domestic portfolio considerations, in markets that consist largely of captive end users. Under these conditions the emergence of LNG demand response to lower prices is likely to have both a gradual and a limited impact in clearing the LNG glut.

Australian LNG exports shut in:

There is some speculation that low spot prices may induce a reduction in export volumes from the three Queensland liquefaction terminals (QCLNG, APLNG, GLNG) that have been developed based on coal bed methane feedgas. The logic here is that exports are uneconomic if netback spot prices don’t cover feedgas costs.

The most vulnerable of the three projects to export volume reductions is the Santos GLNG terminal given a shortage of feed gas relative to export capacity. If Santos needs to purchase incremental feedgas, it has a clear cost base against which exports may be reduced. But in practice this volume is relatively small and may not exceed the level of long term contract cover (around 85%).

In our view Australian exports under long term contract which are backed off by coal bed methane feedgas resource are less likely to be shut in. Coal bed methane production cost signals are relatively complex. There is also an environmental angle to ramping down production given a powerful domestic anti-fracking lobby (led by the agricultural sector). These factors suggest to us that disruptions to targetted production levels are likely to be limited.

US LNG exports shut in:

The US is a different story to Australia. The cost base of LNG terminal feedgas is driven by a liquid and transparent price signal in the form of Henry Hub. LNG export contract structures are also highly flexible in their ability to respond to price signals.

As long as netback spot prices cover the variable costs of Henry Hub, liquefaction and transport, LNG will flow out of the US. But if netback spot prices (e.g. in Europe & Asia) fall below this hurdle then US LNG will be shut in. US shut ins are likely to be highly price responsive i.e. within a 1 $/mmbtu range in netback spot prices, there could be upwards of 80 bcma of shut in volume response by 2020.

As a result, we view US shut ins as the backstop global clearing mechanism for the evolving LNG supply glut. In other words it is US LNG flexibility that can provide whatever additional volume response is required over and above the other three clearing mechanisms.

How do clearing mechanisms interact to clear the glut

The scenario we set out above shows an LNG supply glut of 68 bcma (49 mtpa) evolving by 2019 as new liquefaction capacity outpaces demand growth. Of the four sources of incremental volume response, we anticipate European power sector switching and US shut ins interacting to play the most important role.

US shut ins are likely to play a particularly important role as the marginal clearing response mechanism in the global market given their sensitivity to price signals. If the supply glut turns out to be less severe than we show in the scenario, then a lower volume of US LNG shut ins is required. If the glut is more substantial, then higher volumes of US shut ins are needed to make way for less flexible supply sources. This balancing role of US export flows significantly increases the importance of Henry Hub in driving global gas pricing dynamics (something we come back to next week).

It is our view that the current LNG glut is more of a 5 year than a 10 year phenomenon. There are important structural factors eroding the LNG surplus, as well as the market response mechanisms we describe above. These include declining European production, emerging Asian demand growth and the current hiatus of investment in new supply.

There is currently a strong industry focus on the immediate issues of a deepening supply glut. But rebalancing and recovery may not be that far out of sight, particularly given a 5+ year lead time to develop new LNG supply.

So how does the path to recovery look? What happens to global gas prices as the market rebalances? What are the implications for portfolio exposures and the gas investment cycle? These are questions we address next week as we look beyond the current glut at the three phases of market recovery.

Article written by David Stokes, Olly Spinks and Howard Rogers

Client briefing pack

Global LNG and European gas market workshop

Timera Energy offers tailored in-house workshops exploring the evolution of the global LNG and European gas market fundamentals, pricing dynamics and the implications for asset values and commercial strategies. These involve Timera Senior Advisor Howard Rogers (also Director of the Gas Programme at the Oxford Institute for Energy Studies), who is acknowledged as a leading industry expert in the global gas market.

For more information please contact Olly Spinks.

Evidence of a 2016 recovery in gas price volatility

Posted on September 26, 2016January 16, 2020 by David Duncan

Prompt gas price volatility is the key market price signal for short term supply flexibility response. Volatility plays a key role in determining the risk that gas suppliers face in supplying customer portfolios. Volatility also has a strong influence on the value of flexible supply assets. For example, it is an important driver of capacity sales revenue for the owners of storage, pipeline & interconnector assets.

Spot volatility at European hubs has remained subdued since the start of this decade. This has been due to weak gas demand and a prolonged surplus of supply flexibility at European gas hubs. But evidence is building in 2016 of what may be the start of a more structural recovery in European gas price volatility. Recent issues with the UK’s Rough storage asset, and uncertainty over it’s long term future, are playing an important role.

We provide fairly regular updates on historical volatility. Today we also look at an alternative measure of gas price fluctuations: implied volatility. This is a benchmark “implied” from the prices of traded gas options (see boxed section below) and is becoming an increasingly useful source of information on volatility as gas options liquidity at the UK NBP and Dutch TTF hubs improve.

Implied volatility on the rise

The attraction of implied volatility as a benchmark, is the immediacy of its responsiveness to changes in market conditions. It represents a current, forward looking, market view on the level of volatility. This is in contrast to historical volatility which is by definition backward looking. For further details on the characteristics of implied vs historical volatility see the box below and a previous article on the subject.

Historical vs implied volatility

Volatility can be estimated from historic price movements (‘historic volatility’). Volatility can also be estimated or implied from traded options prices (“implied volatility”). This is done by reverse-engineering the volatility consistent with the observed option price using a standard option pricing model. Implied volatility benchmarks have the advantage over historic volatility measures in that they are forward looking and include information on market expectations of future volatility (rather than what has happened historically). Implied volatility relies on access to prices or quotes for traded options. This means that implied volatility benchmarks are less accessible than historical volatility benchmarks (which can more easily be calculated from publicly available price information).

In layman’s terms, an increase in implied volatility can be interpreted as an increase in the market risk premium priced into traded options. This can be seen in practice in Chart 1 which shows a pronounced pickup in the implied volatility of NBP gas contracts across 2016. The volatility measure displayed on the chart comes from the front month NBP ‘at the money’ call option, based on Marex Spectron* data.

Chart 1: Implied volatility of UK NBP front month “at-the-money” call option

Source: Timera Energy (data from Marex Spectron)

The rise in NBP implied volatility at the start of the 2016 was driven by rapidly declining oil and gas hub prices. Volatility rose again in Q2 2016 as the result of a pronounced short squeeze. Since June, NBP volatility levels have been elevated by the outage issues at the UK’s Rough storage facility. Brexit has also helped by increasing the impact of GBP exchange rate volatility on NBP gas prices.

Rough storage plays an important role in dampening gas price fluctuations at the UK NBP. This is because gas is withdrawn from Rough in periods of higher prices and gas is injected in lower priced summer periods. The complete injection outage on Rough this summer has been a big factor driving higher NBP volatility, as prices have fallen more than they otherwise would have without Rough to absorb surplus gas. Looking ahead to the coming winter, reduced Rough withdrawal capability is likely to continue to support volatility via more pronounced periods of higher prices.

An interesting observation on Chart 1 is that the implied volatility benchmark is itself quite volatile. This reflects the limited liquidity in the gas options market in Europe. Price swings in relatively thinly traded options contracts can translate into big fluctuations in implied volatility. This is particularly the case when demand for options rises sharply relative to available liquidity e.g. the surprise Rough outage announcement this summer. However these sharp swings in implied volatility are typically short lived as liquidity returns and the market responds to arbitrage opportunities.

Historical volatility tells a consistent story

One key limitation of implied volatility as a benchmark is the relatively short available historical data set. It is only over the last three to four years that gas options liquidity has evolved to the point that supports sensible estimation of implied volatility. To understand the evolution of NBP price behaviour over a longer 15 year time horizon we return to historical volatility in Chart 2.

Chart 2 Evolution of NBP historical volatility (2001-16)

Source: Timera Energy

The recovery in historical volatility in 2016 (Chart 2) is not as pronounced as the one in implied volatility (Chart 1). But the perspective of a 15 year horizon also appears to suggest that 2016 may be marking the start of a volatility recovery.
Recovering volatility is consistent with several important fundamental drivers at work in the European gas market:

Ageing & closing supply flexibility infrastructure (e.g. Rough and Groningen)
A lack of investment in new flexible supply infrastructure this decade given weak market price signals
Increasing swing demand from gas-fired power stations as load factors recover
Increasing intermittency as development of wind and solar capacity continues

The fact that implied and historical volatility benchmarks are both pointing towards a recovery is an important sign that these fundamental drivers are starting to make an impact.

Article written by Olly Spinks & David Stokes

*For more information about the Marex Spectron implied volatility data please contact Richard Frape.

The impact of declining offshore wind costs

Posted on September 19, 2016January 16, 2020 by David Duncan

Competition between European utilities to deliver offshore wind projects is hotting up as companies refocus their business models towards renewable generation development. The low carbon generation technology landscape has been redefined by the results of two offshore wind tenders over the last month.

DONG won a July tender to develop two 350MW projects off the Dutch coast at a jaw dropping 72.70 €/MWh. Only to be out done by Vattenfall who won a 350 MW Danish offshore tender last week for 63.80 €/MWh. These numbers have smashed a cost target set by DONG in 2012 to reduce offshore costs below 100 €/MWh by 2020 (at the time costs were around 160 €/MWh).

Costs reductions are being driven by factors such as technology innovation, larger turbine sizes, falling raw materials prices and lower financing costs. Low tender prices in the Dutch and Danish auctions also reflect governments de-risking tendering processes by easing consenting, providing relevant wind & ocean data and excluding grid connection costs.

Falling offshore wind costs come at an important time for the European renewables industry. Consumers and governments across Europe have become more sensitive to rising low carbon subsidy related charges on electricity bills. But the combination of falling offshore wind costs and lower commodity prices is helping to alleviate consumer pain.

In this article we set out five factors to consider in relation to falling offshore wind costs and their impact on the evolution of European energy markets.

1. Offshore wind is an increasingly competitive technology

Because of its rapidly declining cost curve, offshore wind now appears to be gaining a significant advantage in the low carbon technology race. Onshore wind sites are increasingly hard to find and technology costs are maturing. The costs and delivery risk associated with nuclear plant are rising rather than falling. The lifecycle environmental and sustainability benefits of large scale biomass look to be increasingly dubious. And large scale solar is a difficult prospect in most of Europe given winter peak loads and lower load factors.

The UK National Audit Office published a report in July 2016 which contained analysis of different low carbon technology costs. Chart 1 shows how mature (2025) technology costs estimates have evolved over the last six years. The downwards trend is clear and this analysis does not account for the shock cost reductions this summer which are clear evidence of faster cost decline rates for offshore winds.

Chart 1: Forecast of levelised costs in 2025 for renewable technologies in the UK

Source: Nuclear Power in the U.K, National Audit Office, July 2016

2. Support and competition is driving down offshore wind costs

The scale and cost of European support for low carbon technology has not been without its critics. On a global scale, the direct emissions reduction benefits from European renewable generation development are relatively small. But a key argument supporting consumer-borne subsidies has been that they spur technology cost reduction benefits that can be enjoyed globally.

Support for offshore wind appears to be paying clear dividends in the form of falling costs. This has been assisted by genuine benefits from using competitive allocation mechanisms for development support, as opposed to governments trying to negotiate bilateral agreements. The recent Dutch offshore tender also illustrates the benefits of improved quality and transparency of information on project value drivers such as project permitting, wind speed data and sea floor characteristics which benefit both project developers and investors who need to incorporate lower risk premiums into their project evaluation.

The same cannot be said for nuclear technology. Falling offshore wind costs make last week’s decision by the UK government to proceed with the Hinkley Point C project, and its 35 year £92.50 (108 €/MWh) guaranteed inflation adjusted fixed price, even more controversial. The government has indicated it intends to try and renegotiate terms with investors to reduce the estimated cost burden of the project (~£30 billion). But as it stands, Hinkley Point looks to be an increasingly bad deal for consumers, particularly in a world of lower commodity prices and declining offshore wind costs.

3. Higher volumes of offshore wind development

Falling costs and the increasingly competitive nature of offshore wind are likely to provide a boost to the volume of capacity developed across Europe. This may be particularly important in countries that are constrained in their ability to develop onshore wind.

Some of the big hurdles that faced offshore developers at the start of this decade are also starting to fall away. Regulatory frameworks are being developed to deal with the complexity and cost of offshore transmission networks. Governments are implementing streamlined and de-risked tender processes to ease the burden on project developers. Investors are getting more comfortable with project financing. Wind farm sizes are also increasing in parallel with the development of larger turbines (with unit sizes now approaching 10MW) leading to increasing economies of scale. These factors combine to suggest higher capacity build rates in offshore wind.

4. Knock on cost implications to maintain security of supply

Unfortunately the headline tender price levels for offshore wind do not tell the whole story. Part of the reason that the recent Dutch and Danish tender results look so low is that the system costs that result from offshore wind development are being pushed onto the consumer via other charges.

Transmission connection costs and risk vary significantly by project based on wind turbine distance from existing networks and the ability of multiple closely located projects to share new connection costs. And the direct costs of connection are only part of the cost burden that offshore wind places on electricity markets.

There can be significant knock on cost requirements to upgrade the onshore transmission network to accommodate wind. This is particularly the case if the development of offshore wind is focused in areas that are isolated from customer load centres, such as those off Scotland in the UK. Higher wind volumes also act to increase system balancing costs given the intermittency of wind output.

The intermittent nature of wind also means it needs to be backed up by flexible generation capacity particularly where there is insufficient grid interconnection to allow market forces to balance volume variability within and across national borders. This means higher costs in the form of capacity payments to support adequate flexible capacity.

Large scale electricity storage may one day reduce the requirement for flexible generation, but current costs and capabilities of utility-scale storage suggest this will not be for many years into the future. Demand-side response offers further potential to balance intermittent generation, but similarly is dependent on smart grid and development of new incentive mechanisms.

5. Higher power price volatility

There is set to be a strong relationship between increasing volumes of European wind capacity and higher power price volatility. Volatility comes both from price rises when wind levels are low and increasing periods of low/negative prices during high wind periods.

There are sometimes geographical diversification benefits in wind patterns that can help smooth the impact of intermittency as interconnection between European markets improves. But the speed and scale of volume swings in wind output will continue to feed through into higher prompt volatility.

While increasing volatility may make governments nervous, it is an important price signal for the developers of flexible generation capacity. Developers and owners of gas-fired power plants are adapting as intermittent generation increases in order to focus more on returns from flexible operation in response to volatile power prices.

Wind likely to be backed by development of new gas-fired plants

Europe is going to need to replace large volumes of conventional generation capacity over the next decade. Coal plants are facing closure across Europe as governments tackle emissions. The European nuclear fleet is set to shrink significantly next decade with net closures scheduled in many countries (e.g. Germany, Sweden, UK, Belgium and Switzerland). Many of the first generation of European CCGTs are also approaching the end of their economic lifetimes.

Offshore wind can play an important role in helping to plug the capacity gap caused by conventional plant closures. This will need to be supported by substantial increases in pan-European interconnector capacity to help manage intermittency. But before the arrival of competitive large scale storage technologies, it looks like another generation of gas-fired plants will be developed in order to maintain adequate system flexibility.

Article written by David Stokes and Olly Spinks

UK CCGT margins take off, coal plants bleed

Posted on September 12, 2016January 16, 2020 by David Duncan

Some important structural changes are taking place in the UK power market supply stack. 2016 has seen a pronounced shift in favour of gas-fired generators. The flipside of the recovery in CCGT margins and load factors is that UK coal plants are being driven out of merit.

In last week’s article we looked at the changing fuel price relationships behind a recovery in the fortunes of CCGT plants across Europe. These same drivers are at work in the UK but are being magnified by three additional factors:

A very tight UK system capacity margin
The dominant role of CCGTs in setting power prices
The UK carbon price floor

Excluding emergency reserve (SBR) capacity, the UK capacity reserve margin as measured by National Grid is now zero (0.1% if you want to be pedantic). In other words there is a significant possibility that system stress this winter will require Grid to call emergency reserve capacity to maintain security of supply. That prospect is now feeding through into forward spark spreads.

The price setting role of CCGTs in the UK has meant that falling gas hub prices in 2016 have had a direct impact in reducing power prices. The combination of falling power prices and the UK carbon price floor has sent spot UK coal plant margins well into negative territory.

Coal generators are now structurally out of merit in the UK. But coal units remain critical for UK security of supply until the Capacity Market starts to deliver large scale new gas-fired capacity from 2020. This leaves the UK’s coal units setting prices in the year-ahead capacity auctions at levels that allow them to cover fixed costs and avoid closure.

What is happening with UK generation margins?

Chart 1 shows the historical evolution of spot CCGT and coal plant generation margins (CSS and CDS), as well as current spark and dark spread forward curves.

Chart 1: UK historical spot and forward CSS & CDS

Source Timera Energy (49% HHV CCGT efficiency, 36% coal efficiency)

A sharp uptick in CCGT margins (spot CSS) can be seen in Summer 2016. This has been driven by falling gas prices given oversupply across European hubs (as explained last week). This oversupply has been particularly pronounced in the UK gas market because of the Rough storage injection outage.

Summer maintenance on Norwegian export lines to the Continent has seen gas diverted to the UK rather than have production shut in. Exports from the UK to the Continent in summer often run at maximum interconnector capacity anyway, and without Rough to soak up the resulting surplus, NBP prices have fallen to clear the UK market.

While these dynamics are an interesting short term phenomenon, there is a more important structural change taking place with spark spreads. The recovery in spot CSS across the summer has been accompanied by an increase in forward CSS, particularly for the coming winter (16/17). Winter CSS is now trading around 9 £/MWh, with average forward spark spreads over the next 12 months rising above 6 £/MWh. These levels represent a return to CCGT margins not seen since early this decade.

The events of 2016 have crushed coal plant generation margins. Spot baseload CDS has fallen into negative territory over the summer as gas prices have weakened. The CDS forward curve has so far just managed to remain in positive territory when measured on a market basis (i.e. only including fuel and carbon costs). But once variable coal transport and system costs are accounted for, forward baseload coal generation margins are also in negative territory.

As a result, the more efficient and flexible UK coal units have been relegated to a peaking role, eking out small margins from within-day price shape. Older coal units are sitting idle as reserve capacity.

Driving coal-fired generators out of the capacity mix was one of the key reasons for the UK government introducing the carbon price floor. The price floor in combination with shifting commodity prices has done that job well. But this has left the UK power market with a major security of supply issue.

Winter 16/17 – CCGT margins may recover further

Winter sparkspreads are foreshadowing what could be a more significant jump in spot outturn CSS this winter. The UK may be insulated from the risk of rolling blackouts by the 3.5 GW of Supplemental Balancing Reserve (SBR) capacity which the system operator (National Grid) has contracted over the coming winter. But SBR capacity does not participate in the generation supply stack and should only be called by Grid under threat of system emergency. This leaves room for the zero system capacity margin (ex-SBR) to drive significant increases in power prices and volatility to the benefit of CCGT margins.

The other factor that is helping UK CCGTs is the fact that coal plants are becoming more expensive on a variable cost basis. As well as covering variable fuel, carbon and system costs, coal plants also need to recover relatively high start costs. Because coal plant generation margins (CDS) are so weak, power prices are needing to rise in peak periods to incentivise coal plants to run. These factors act to drive increasing CCGT margin rents.

The UK will likely scrape through this winter without flickering lights given a 5% system reserve margin including SBR capacity. But it is the non-SBR zero reserve margin that will drive wholesale market pricing dynamics. The extent to which this translates into market fireworks will of course depend on variables such as cold weather, wind conditions, outages and performance of the debilitated Rough storage facility. But Winter 2016/17 is set to be the tightest period of the UK power market’s two decade history.

2017 and beyond

SBR will be discontinued from Summer 2017, with the new year-ahead 2017/18 capacity auction the mechanism used to procure an adequate system reserve buffer. This will have an important impact on wholesale pricing dynamics. Unlike SBR, capacity contracted in the year-ahead auctions will participate in the supply stack. In other words the units that make up the system reserve buffer will compete with other generators on a variable cost basis.

All other factors being equal, this should act to dampen power price levels and volatility when compared to the SBR scenario of Winter 16/17. But the relative behaviour of commodity prices will also play an important role. If weaker gas prices mean coal plants remain out of merit, then CCGTs may continue to earn healthy rents as coal unit variable & start costs act to increase power prices above CCGT variable costs.

Coal units will also play a key role in setting capacity prices in the next three year-ahead auctions (17/18, 18/19 & 19/20) before new supply comes online in 2020. Plants such as Fiddlers Ferry West Burton & Cottam are currently only recovering a fraction of fixed costs in the wholesale market. This means they will need to bid at capacity prices levels that allow fixed cost recovery in the year-ahead auctions. Competition to supply incremental megawatts will be particularly limited in 18/19 and 19/20 T-1 auctions given most generators already have capacity agreements from the four year ahead (T-4) auctions. Don’t be surprised if these auctions clear at a significant premium to previous ones.

Implications for asset owners & investors

The UK government is targeting the closure of all coal plants by 2025. It is unlikely that they will need to wait that long. The carbon price floor and the shifting relationship between gas and coal prices appear to be sending coal plant generation margins into terminal decline.

The UK may lose one or two more coal stations by 2018. But the remainder of the coal fleet is likely to remain on capacity market life support until 2020 when new gas plants come online. Despite being driven out of merit, coal units are set to retain an important influence on marginal pricing in both the capacity and wholesale energy markets over the next 3-4 years.

Coal plant woes are good news for UK CCGT owners. Structural gas price weakness and a tight system reserve margin are driving CCGT margin recovery. The relegation of coal in the UK merit order also means higher CCGT load factors, higher average efficiency and lower start costs. These factors contribute additional value over the headline spark spread recovery.

The recovery in the fortunes of UK CCGT plants is attracting plenty of investor interest. Utilities are eyeing the sale of existing CCGTs as a way of raising capital to sure up their balance sheets. And a bolstered capacity target for the December 2016 four year ahead auction has seen a frenzy of activity around CCGT development projects.

There have been a number of false dawns to the CCGT recovery story which has been anticipated for several years. Conditions for UK CCGT have improved steadily since gas prices started to weaken in 2014. But 2016 looks to be confirming the start of a more structural recovery.

Continental generation margins: Gas is back

Posted on September 5, 2016January 16, 2020 by David Duncan

‘Positive baseload generation margins for German CCGTs? That’s not possible!’

Exclamations such as this have echoed across European power markets this summer. At the start of 2016, the concept of CCGTs displacing coal plants in the merit order was off the radar. Raising such an idea in front of a Risk or Investment Committee in January would have raised sceptical eyebrows. Yet CCGTs all over Europe are back in action this summer… for the moment anyway.

We set out the logic behind why we thought gas vs coal switching was going to be a key mechanism for clearing surplus gas at European hubs in an article this spring. This summer’s revival in European CCGT load factors is an illustration of gas vs coal switching in action. Across Europe we estimate up to 40 bcma of potential CCGT switching demand as gas prices fall.

Germany is arguably Europe’s most hostile power market for CCGTs given the dominance of cheap coal and renewables. So in today’s article we focus on the German market to illustrate the drivers behind the sudden shift in competitive balance that has taken place between gas and coal plants. But the logic extends across all European power markets.

Plant margins & fuel prices: a summer 2016 case study

Enormous energy is exhausted attempting to conquer the fundamental modelling of European power markets. The basic techniques of supply stack modelling have hardly evolved over the last 20 years. But the evolution of processing power and data management have enabled the complexity and granularity of power market modelling to increase exponentially.

These developments have led to the insightful analysis of factors such as the impact of a strong breeze at 03:30 on Sunday the 29^thJanuary 2037. But behind the overheating processers and ever expanding databases a simple fact remains: power plant margins are predominantly driven by the relative behaviour of fuel prices.

Summer 2016 is a good example of this and you don’t need a complicated power market model to understand it. Chart 1 illustrates the recovery in baseload German clean spark spreads (CSS) so far this summer. Baseload CSS has risen from around -8 €/MWh in May to over 3 €/MWh by the end of August. That is an increase of more than 10 €/MWh in the space of three months. Over this period, coal plant generation margins (CDS) have remained at close to zero levels as coal plants dominate marginal price setting.

On a forward basis baseload spark spreads decline back into negative territory. However peak German spreads remain positive across the coming winter.

Chart 1: German baseload coal and CCGT generation margins (CDS and CSS)

Source: Timera Energy (gas efficiency 49% HHV, coal efficiency 36%)

Chart 2 shows the primary cause of the recovery in CCGT margins, a pronounced shift in relative gas versus coal prices.

Chart 2: ARA coal prices vs German NCG gas hub prices

Source: Timera Energy

Let’s consider this relative fuel price shift a leg at a time. European gas hub prices have come under renewed pressure across the last quarter. Owners of oil-indexed gas contracts have been strongly incentivised to take high volumes of gas given 6-9 month price lags which are capturing the low Brent prices from the start of 2016. In addition, summer gas demand for storage injection has been relatively subdued, with the UK’s Rough storage out of action and healthy Continental storage inventories. A gradual flow of LNG cargoes continues into North West European hubs, although Europe has so far not yet felt the fuller impact of surplus LNG as ramp up issues have continued with several new LNG liquefaction projects.

In sharp contrast, global coal prices have undergone a pronounced recovery across Q2 and Q3 2016. As usual with the coal market, China is in focus. There have been ongoing cutbacks in Chinese coal production in response to overcapacity. These have been reinforced by temporary weather related issues such as unusually hot weather (boosting coal-fired power demand) and heavy rain (impacting coal production). A tightening in the Pacific Basin coal market has boosted Atlantic Basin prices as European buyers need to compete for supply.

The shift in relative fuel prices has been reinforced by poor nuclear availability in the French power market this summer, which has supported healthy German exports.

Looking beyond this summer – what is happening with forward prices?

The recovery in prompt CSS has not lifted forward baseload spark spreads into positive territory (forward CSS can be seen around -5€/MWh across Winter 16/17 in Chart 1). The current downward slope of the forward spark spread curve is driven by two factors:

Gas curve contango: there is a pronounced upward slope to European gas hub curves, in sympathy with Brent curve contango
Coal curve backwardation: the coal curve remains in backwardation, reflecting an anticipated easing in the conditions that have tightened supply over the last few months

Gas curve contango reflects a market anticipation that the pronounced oversupply of Summer 2016 is a temporary phenomenon that will ease into the coming winter. This is reinforced by the upward pull that oil-indexed gas contracts exert on European gas forward curves, as the 2016 oil price recovery feeds through into contract prices and the Brent curve remains in contango. Coal curve backwardation on the other hand reflects an anticipated easing in the conditions that have tightened supply over the last few months.

It is however important to note that the forward spark spread curve is not a good forecast of future spot spreads. If you are not convinced you can find evidence in animation here and an explanation of the logic behind this here.

What factors are likely to determine the evolution of European generation margins going forward?

One important factor is that coal plants remain the dominant setter of Continental power prices. That can be seen in Chart 1 via the relative stability in German generation margins (CDS) across 2016 i.e. power prices are moving in a correlated fashion with coal prices. This means that further falls in gas hub prices (or rises in coal prices) will translate into higher spark spreads. So if spot gas hub prices remain weak (rather than recovering in line with gas curve contango), this will support the CCGT margin recovery story.

There are two drivers worth watching as an indication of further weakness in European gas hub prices over the next 2 – 3 years:

LNG imports: The volume of surplus LNG that flows into Europe as new global liquefaction capacity ramps up, will play an important role in determining whether hub prices recover back towards oil-indexed benchmarks.
Henry Hub: The US gas market represents important price support for European hubs, with Henry Hub dynamics increasingly linked to European hubs by the hedging and optimisation of US LNG export volumes.

The recovery in CCGT margins and load factors will likely ease into the coming winter given seasonal gas price shape. But the events of Summer 2016 illustrate the growing importance of gas vs coal switching in clearing an oversupplied European gas market.

We have focused on Germany this week as the last cab off the rank when it comes to CCGT margin recovery. The story for CCGTs in other Continental power markets improves with the power price spread over Germany. But the UK power market is a clear first cab off the rank. The impact of relative fuel price shifts on UK CCGT margins has been magnified by an anaemic system reserve margin and gas plants returns have taken off this year. We return to focus specifically on the situation for generation margins in the UK next week.

Article written by David Stokes and Olly Spinks

Rough storage issues remain a structural threat

Posted on August 29, 2016January 16, 2020 by David Duncan

The UK gas market was hit by a major shock in July in the form of an extended outage at its largest gas storage facility. The immediate impact of Centrica Storage Limited’s (CSL) suspension of injections at the Rough facility was a spike in winter gas prices. Winter prices eased again last week as CSL announced that it would bring back the majority of Rough wells online, allowing existing gas in store to be withdrawn from November.

Rough accounts for more than 70% of the UK’s working gas volume. It is a seasonal storage asset with a relatively low cycling rate. But the scale of storage working gas volume means Rough plays a key role in providing supply flexibility to the UK. Rough is also large enough to be important in a broader North West European market context.

The curtailment of Rough does not create a significant seasonal flexibility issue in the UK market. But it does exacerbate the UK’s gas deliverability issues. During periods of high demand or supply outages, the UK can face constraints in delivering enough gas into the network to meet demand. It is the loss of Rough’s deliverability, rather than working volume that poses a problem for the UK market.

The market may have breathed a sigh of relief last week as CSL’s announcement alleviated worst case fears for the coming winter. But the potential loss of Rough capacity remains a structural threat for the UK gas market.

An update on events leading into this winter

Rough is a partially depleted offshore field that was converted to storage operations in 1985 so has been operating for over 30 years. A number of issues have surfaced around the ageing Rough facility over the last 18 months which have impacted storage operations:

18 March 2015: routine inspection identified ’a potential issue with well integrity’ and resulted in a 25% potential reduction in working gas volume from 3.7 bcm to 3.1 bcm.
22 June 2016: a full shutdown was announced for 42 days due to a containment envelope failure in one of the wells at a lower pressure than expected. This was discovered during the testing programme prompted by the March 2015 outage.
15 July 2016: the full shut down was extended until March / April 2017. CSL stopped selling capacity (SBUs) for the 2017/18 storage year but is investigating the possibility of making some wells and limited withdrawal capacity available over winter 2016.
4 August 2016: scenario curves released by CSL indicate that it believes that Rough will be able to meet a minimum withdrawal rate of 6 mcm/day (vs a 42 mcm/day flow rate at full output).
22 August 2016: CSL announced that it expects 20 wells (of a total of 30) to return to service by Nov 2016, likely to support a maximum deliverability of around 35 mcm/day.

Chart 1 illustrates the UK’s total storage deliverability over the coming winter under three different scenarios for Rough availability:

100% Rough availability i.e. assuming full injection would have been possible over the summer – blue shaded area
A ‘best case’ scenario for the restrictions on Rough coming into the current winter (based on a constant 35 mcm/day withdrawal) – solid red line
A scenario without the Rough storage facility (i.e. 0% availability) – dashed red line

Chart 1: Impact of Rough availability on total UK gas market storage deliverability

Source: Timera Energy (storage facility data from National Grid)

The chart illustrates the daily volume of gas deliverability (vertical access) assuming maximum rate withdrawal from full inventory over time (horizontal access). Rough as a slow cycle seasonal facility sits at the bottom of the chart. Despite Rough’s dominance of UK working gas volume, its slow withdrawal rate means that it contributes proportionally less to UK deliverability (and takes a long time to empty). Faster cycle storage facilities (such as Holford, Stublach & Aldbrough) sit higher up in the chart. These provide much higher deliverability relative to working gas volume, but empty more quickly.

There are some important points to note about CSL bringing wells back online for this winter. Estimates of potential deliverability (around 35 mcm/day) may actually be significantly lower in practice across the winter. Rough withdrawals will likely be impacted by:

Lower delivery rates as inventory volumes fall
More pronounced profiling of Rough volumes into higher priced periods given the relatively low level of inventory (e.g. in Q1 2017 when the UK is more vulnerable to cold weather and supply disruptions)
Ongoing caution by CSL in how they operate the facility

So estimates of headline withdrawal rates (and the solid red line in Chart 1) are likely to overstate the deliverability that Rough will be able to provide this winter.

Rough is also constrained this winter by the fact that current inventory levels are a third of normal levels (1.26 bcm vs 3.73 bcm at the start of last winter), due to the injection restrictions. This leaves the UK market significantly exposed to any more prolonged periods of system stress (e.g. as was seen in 2013).

Out of the frying pan into the fire – Rough remains a structural problem

The partial availability of Rough for the coming winter may have calmed the market temporarily. But the bigger issue facing CSL is that revenue from current seasonal price spreads is unlikely to support the lifetime renewal capex required to maintain Rough operation. This leaves the UK market contemplating the red-dashed line in Chart 1.

Visually the biggest impact of the loss of Rough capacity is the reduction in the number of days of gas in store (comparing the blue shaded region with the red-dashed line). But the UK market’s vulnerability is actually focused on the volume of deliverability over a shorter 2-3 week period, i.e. on the left hand side of the chart.

A loss of Rough capacity decreases the UK’s daily deliverability of gas from storage by almost 25%. This can be seen via the red dashed line (0% Rough) intersecting the vertical access 42 mcm/day below the blue shaded region (100% Rough). And it is this threat of loss of deliverability that is of most concern to the UK market given its vulnerability to periods of system stress e.g. a cold snap or infrastructure outage.

The logic can be summarised as follows:

Well interconnected: There is more than adequate import capacity to bring gas into the UK (e.g. via the Norwegian Continental shelf, the BBL and IUK interconnectors and LNG terminals). But there can be significant supply chain lead times to attract adequate volumes of gas imports, particularly for LNG.

Deliverability squeeze: The UK market is most vulnerable to a shortage of deliverability over a 2-3 week horizon, as a result of high demand and/or supply disruptions. A classic example of this was in Mar/Apr 2013 when cold weather, field and interconnector outages caused an NBP gas price spike for several weeks, eventually alleviated by the diversion of LNG cargoes in response to higher prices.

If the UK loses Rough capacity it will have a knock on impact for the utilisation and value of other storage assets. Other storage facilities will likely operate to a more seasonal pattern to backfill loss of Rough capacity. This in turn reduces the volume of deliverability flexibility that the UK market has to dampen price fluctuations. In other words it supports prompt gas price volatility and the value of faster cycle storage capacity. This is a dynamic that is likely to support gas volatility across the European gas market.

Rough is not alone. It is only one example of ageing European flexible gas supply infrastructure. Around 5% of European storage capacity has been closed this decade. There is also the prospect of significant declines in supply flexibility from maturing gas fields (e.g. Groningen and Troll). As these factors come into play they are likely to drive a recovery in the market price signals required to support the economics of incremental flexible supply infrastructure, particularly prompt gas price volatility.

It is unlikely that large volumes of new seasonal storage capacity will be developed. The supply flexibility requirement in the European gas market is evolving with increasing interconnection, growing market liquidity and the more flexible use of gas-fired power plants to support intermittency. These factors all point towards the increasing importance of deliverability (versus seasonal flexibility). And this is likely to mean a shift in focus towards faster cycle storage assets, LNG import infrastructure and demand side response.

Article written by David Stokes and Olly Spinks

French Carbon Price Floor

Posted on August 22, 2016January 16, 2020 by David Duncan

As the European carbon price continues to languish below 5 €/t, France has decided to take action. The French government announced in April that it would follow the UK in unilaterally implementing a carbon price floor targeted at 20-30 €/t. Subsequently in July, an advisory committee commissioned by the government released a report with a strong recommendation to implement a domestic mechanism to increase the carbon cost for coal plants.

France’s unilateral action reflects its frustration at the absence of any meaningful European initiatives to strengthen carbon pricing. But unlike the UK, France is also focused on spurring other European countries into action in an attempt to build some momentum behind the COP21 agreement. The French government are actively promoting the idea of a EU wide ‘carbon price corridor’ to maintain an ETS carbon price above 20 €/t from 2020.

The French are right in challenging the status quo. The European ETS carbon price signal has spent the last three years hovering around 5 €/t, a level which is essentially meaningless for inducing significant emissions reduction. But the impact of implementing a price floor in France versus more broadly across Europe are two very different things. In this article we explore the impact of (i) higher carbon prices in France and (ii) higher carbon prices across Europe, focusing on the potential impact of these scenarios on power market pricing and asset value dynamics.

A summary of the French proposal

The French government initially proposed a uniform carbon price floor to operate on a ‘top up’ basis, similar to the existing UK carbon price floor. In other words a defined domestic price premium would be added to the EUA ETS price.

However, a French government advisory committee, which reported back in July, recommended an alternative approach to specifically target and penalise coal (to incentivise a shift to lower carbon intensity gas-fired generation). This could either be done via an increase taxation on coal-fired power plants or by imposing tighter carbon emissions standards. The French environment minister, Ségolène Royal, has not yet made a formal policy announcement but has reiterated the Governments commitment to introduce some form of domestic policy measure from January 2017.

Royal has also indicated that France will propose that Europe introduces a European ‘carbon price corridor’ that will allow EU nations a degree of flexibility to set their own carbon pricing terms within a broader European framework. The current French proposal is for a European carbon floor price of between 20 €/t and 30 €/t in 2020 with annual increases of 5 – 10% with a target of 50 €t by 2030. The proposed price ceiling would start at 50 €/t in 2020 and increase at a similar annual rate to the floor price.

Impact on the French power market

The primary effect of a uniform carbon price floor policy would be its impact in increasing the variable costs of coal and gas fired generators. Chart 1 illustrates the incremental cost increase for a CCGT 50% HHV and a coal plant 40% HHV.

Chart 1: Short run marginal cost impact of 30 €/t carbon price floor (Cal17 on 10th Aug 16)

Source: Timera Energy

This will clearly have an adverse margin impact on French thermal generators, particularly coal plants. However an attractive aspect of implementing a carbon price floor from a French perspective is that it is likely to have a limited impact on wholesale power prices. This is because domestic thermal (coal and gas fired) power plants play a relatively small role in setting marginal prices in the French market.

France has low installed volumes of coal (3GW) and gas (6.5GW) fired generation capacity. And the pricing impact of this capacity is limited, given interconnector flows play a dominant role in setting French power prices, particularly on the German border.

The carbon price floor is most likely to impact winter and peak prices when French thermal plant can influence marginal power prices. The policy should have little impact on summer and offpeak prices which are currently driven by coal priced imports from Germany.

If France imposes a carbon price measure on coal plant only as recommended by the advisory committee report, then the wholesale power market impact will be even more muted. One of the reasons the committee made this recommendation was to prevent the closure of gas-fired plant required for security of supply. Gas plants would benefit from somewhat higher load factors if the government pursued a coal only carbon charge.

Any increase in domestic (or neighbouring) power prices will be warmly welcomed by the French nuclear giant EDF, given this will flow straight to the bottom line of their nuclear power portfolio. The benefits of the price floor policy for EDF may not be a coincidence either. The French government owns a majority stake in EDF and is in the process of raising capital to support the utility’s nuclear development ambitions. Increasing wholesale power prices would certainly help to ease EDF’s balance sheet issues.

Possibility of a broader European roll out?

The French push to implement broader carbon price support across Europe is likely to come up against much stronger headwinds. But if successful, it would also have a more meaningful impact on wholesale power market pricing dynamics.

Germany is key. The German electorate is certainly open to proactive environmental policy. But given coal plants dominate wholesale price setting in the German power market, higher carbon prices will be passed through directly into higher wholesale prices.

German policy makers are already contending with a strong industry lobby pushing against the impact of renewable support in raising power bills and eroding German competitiveness. This lobby is unlikely to welcome measures to support a higher carbon price. The strong German coal and lignite industry lobby will also be hostile. However the German Energy Ministry so far appears to be at least engaging in a conceptual discussion of a policy shift to support to a minimum carbon price, even if so far it is from behind closed doors.

Several other European countries may be open minded in considering implementation of a carbon price support policy (e.g. Ireland, Netherlands, Belgium, and Norway). The degree to which carbon price support gains traction across other EU nations will likely be an important factor in influencing the German debate.

What does a broader price floor rollout mean for power market pricing dynamics?

Let’s leave aside the policy debate and focus on the pan-European impact of higher carbon prices. This could be driven by a broader implementation of carbon price support mechanisms or by measures to support the EU ETS price (a much cleaner way of addressing the problem).

The relative impact of higher carbon prices is greater for coal plants than for gas plants, given the carbon emissions intensity for coal plants is more than double or CCGTs (0.85 t/MWh for 40% a coal plant vs 0.4 t/MWh for a 50% CCGT), as illustrated in Chart 1. This drives several knock on effects:

Short run competitiveness: the relative increase in gas plant competitiveness vs coal on an SRMC basis, supports gas vs coal plant switching (or substitution of gas for more expensive coal plant generation).
Long run competitiveness: The erosion of coal generation margins, combined with other regulatory hurdles that coal plants are facing, would likely bring forward asset closure dates.
Gas burn: Power sector gas demand in Europe would be positively impacted, by 1. in the short term (given gas for coal switching) and by 2. in the longer term.

The UK power market provides a useful case study for implementation of a carbon price support policy. All three factors can be observed in progress.

The impact of higher carbon prices on absolute power prices and generation margins is more complicated. This depends on the degree to which carbon prices are passed through via the marginal plants setting wholesale power prices. Chart 2 however illustrates some important upper and lower bounds to consider, based on an example of a 30 €/t price floor in the German power market.

Chart 2: bounding the impact of a 30 €/t carbon price floor in Germany

Source: Timera Energy (Cal17 prices on 10th Aug 16)

The impact of higher carbon prices on wholesale power prices depends on the marginal (or price setting) generation unit. The two key asset classes that dominate power price setting in Europe are:

CCGTs: which generate CO2 at an intensity of approx. 0.40 t/MWh of power produced
Coal plants: which generate CO2 at an intensity of approx. 0.85 t/MWh of power produced

Because coal plants have a much higher carbon intensity than CCGTs, the pass through of carbon prices to wholesale power prices is greater when coal plants are marginal (setting prices). But a carbon price floor would also change the dynamics of price setting plant. It would act to push coal plants out of merit, increasingly the influence of CCGTs in setting marginal prices.

Chart 2 illustrates the impact of higher carbon prices on German Calendar 2017 power prices and generation margins. The chart shows the impact on (i) the Clean Spark Spread (CSS) generation margin of CCGTs (left hand column) and (ii) the Clean Dark Spread (CDS) generation margin of coal plants (right hand column). Two scenarios for carbon price pass through are considered:

100% coal pass through (upper bound): If coal plant is setting prices, then the higher carbon intensity of coal plants translates into:
- CSS: a significant rise in CCGT generation margins (from -6 €/MWh to more than 7 €/MWh), given power prices rise by more than the increase in CCGT variable costs (left hand column)
- CDS: no change in coal plant generation margins, given power prices rise by the same as the increase in coal plant variable costs (right hand column)
100% CCGT pass through (lower bound): If CCGTs are setting prices, then the lower carbon intensity of CCGTs translates into:
- CSS: no change in CCGT generation margins, given power prices rise by the increase in CCGT variable costs (left hand column)
- CDS: a significant fall in coal plant generation margins (from 4€/MWh to -9 €/MWh), given power prices only rise by the much lower increase in CCGT variable costs (right hand column)

The dominance of coal plants in setting power prices in Germany means that higher carbon prices translate into three important effects:

Higher power prices: The influence of the greater carbon intensity of coal generation feeds through to create a significant uplift in wholesale power prices.
Gas for coal switching: Higher carbon prices act to shift the competitive balance towards gas plants, increasing CCGT load factors and pushing coal plants up the merit order. This to some extent counteracts the impact of coal plant pass through of higher carbon costs to power prices.
Higher CCGT generation margins: The increase in power prices acts to significantly increase CCGT margins (given a lower CCGT carbon intensity).

A unilateral French carbon price floor is likely to have a limited impact on the European power market landscape. But a floor price implemented in Germany could have a transformational effect on thermal asset generation margins and coal plant closures across Europe.

Article written by Olly Spinks and David Stokes