2016 was the first year of the growth spurt we expect to see in LNG, with global supplies set to increase by around a further 30% by 2020. That is equivalent to a new LNG train coming on stream every two months until the end of this decade - quite astonishing growth.Spencer Dale, group chief economist
Stability and energy markets don’t go together - booms and busts; rebounds and reversals are the norm.
But the movements and volatility seen last year were particularly interesting since energy markets were buffeted by two separate forces: the continued adjustment to the short-run cyclical shocks that have rocked energy markets in recent years, particularly the oil market; and the growing gravitational pull of the longer-run energy transition that is under way.
In recent years the nature of the cyclical adjustments has been increasingly affected by the longer-run transition that is shaping global energy markets. On the demand side: the shift in the centre of gravity to fast-growing developing economies, led by China and India; together with a slowing in overall energy growth as it is used ever more efficiently. And on the supply side, the secular movement towards cleaner, lower carbon energy sources, led by renewable energy, driven by technological advances and environmental needs.
2016 was a year of both short-run adjustments and long-run transition, and this year’s Statistical Review shines a light on both influences.
Key features of 2016
Primary energy grew by just 1% (171 mtoe) in 2016, almost half the average rate seen over the previous 10 years.
Some of this weakness reflected short-run factors: global GDP grew by just 3% last year, its slowest rate since 2002 - other than at the time of the financial crisis - driven in part by a slowdown in industrial production, the most energy-intensive sector of the economy.
But the weakness is also indicative of the longer run trend towards slower energy growth driven by gains in energy efficiency.
This is the third consecutive year in which energy consumption has grown by 1% or less, with energy intensity - the average amount of energy needed to produce a unit of GDP - falling at historically unprecedented rates.
Growth in energy consumption was again driven by the developing economies. China (1.3%, 47 mtoe) and India (5.4%, 39 mtoe) led the way, contributing almost identical increments, and together accounting for around half of the increase in global demand.
But these similar contributions disguise sharply contrasting trends. India’s energy consumption grew at a similar rate to the recent past, underpinned by solid economic growth. In contrast, China’s energy consumption grew at less than a quarter of the rate seen over the previous 10 years.
This brake in China’s energy consumption partly reflects the gradual slowing in economic growth, but it has been greatly compounded by pronounced weakness in China’s most energy intensive sectors, particularly iron, steel and cement, which together account for around a quarter of China’s total energy consumption.
Some of the weakness in these sectors, which drove China’s rapid growth and industrialization over much of the past 15 years, reflects the structural rebalancing of the economy towards more consumer and service facing sectors.
But the scale of the slowdown - with output in iron, steel and cement below 2014 levels - suggests that some bounce-back is perhaps likely.
Short-run adjustments and long-run transition.
The story in terms of individual fuels also reflects a mix of these two forces.
Renewable energy (including biofuels) (12%, 55 mtoe) was again the fastest growing energy source, accounting for almost a third of the increase in primary energy, despite having a share of only 4%. That said, oil (1.5%, 75 mtoe) actually provided the largest contribution to growth, with the low level of oil prices boosting demand.
Natural gas (1.5%, 57 mtoe) grew at the same rate as oil, although for gas this was considerably slower than its 10-year average.
Perhaps the most striking feature across the different fuels was the continuing rapid descent of coal, with consumption (-1.7%, -53 mtoe) falling sharply for the second consecutive year and the share of coal within primary energy declining to its lowest level since 2004.
The turnaround in the fortunes of coal over the past few years is stark: it is only four years ago that coal was the largest source of energy demand growth. There may be further ups and downs in the fortunes of coal over coming years, but the weakness in recent years does seem to signal a fairly decisive break from the past.
Two years ago, 2015 was a year of thwarted adjustment for oil: strong growth in OPEC production outweighed the responses of both demand and non-OPEC production to lower prices.
In contrast, 2016 was a year of adjustment for the oil market, with oil demand again increasing robustly and production growing by less than a quarter (0.4 Mb/d) of that seen in 2015.
Global oil demand grew by 1.6 Mb/d last year. As in 2015, this strength was almost entirely due to oil importers, with both India (0.3 Mb/d) and Europe (0.3 Mb/d) posting unusually strong increases. Although, growth in China (0.4 Mb/d) and the US (0.1 Mb/d) was more subdued.
As in 2015, the strength in oil demand was most pronounced in consumer-led fuels, such as gasoline, buoyed by low prices. In contrast, diesel demand, which was more exposed to the industrial slowdown, including in the US and China, declined for the first time since 2009.
The weakness on the supply side was driven by non-OPEC production which fell by 0.8 Mb/d, its largest decline for almost 25 years. This fall was led by US tight oil, whose production fell 0.3 Mb/d, a swing of almost 1 Mb/d relative to growth in 2015. China also experienced its largest ever decline in oil production (-0.3 Mb/d).
In contrast, OPEC production recorded another year of solid growth (1.2 Mb/d), with Iran (0.7 Mb/d), Iraq (0.4 Mb/d) and Saudi Arabia (0.4 Mb/d) more than accounting for the increase. Iran’s production and its share of OPEC output are now both back around pre-sanction levels.
The combination of strong demand and weak supply was sufficient to move the oil market broadly back into balance by the middle of the year.
But this was not before inventories had increased even further from their already excessive levels, such that the level of OECD inventories by the end of 2016 was around 300 Mbbls above their five-year average.
The drama and intrigue that has characterized oil markets since the price collapse in 2014 have been dominated by two principal actors: US tight oil and OPEC. What have we learnt about the behaviour of both during this cycle?
Consider first US tight oil, which didn’t exist during the last oil price cycle, and so we are learning about it in real time.
Perhaps the most important thing is that there is no such thing as the behaviour of ‘US tight oil’: the Permian is very different to Eagle Ford which is different to Bakken. So beware generalizations.
Notwithstanding that, the short-cycle nature of fracking meant activity related to US tight oil did respond far more quickly to price signals than conventional oil and, in so doing, dampened price volatility. Rigs started to fall around four to six months after oil prices peaked in June 2014 and picked up even more quickly - within three or four months - once prices started to turn at the beginning of last year.
And this lower activity fed through into slower output growth. In the first half of 2015 - so less than a year after the peak in oil prices - tight oil production grew by just 0.1 Mb/d, compared with over 0.5 Mb/d in the same period a year earlier - a swing in annualized terms of 0.8 Mb/d. Similarly, US tight oil has grown solidly in the first half of this year, following the trough in prices in the spring of 2016.
The final point to note about US tight oil is that productivity continued to rise rapidly through the cycle, with new well production per rig increasing by around 40% per year in both 2015 and 2016. Despite rigs in the Permian falling by over 75%, output continued to grow. Put differently, a rig operating in the Permian today is equivalent to more than three rigs at the end of 2014.
So that is the backstory on one of the principal actors, what about the other: OPEC?
As with many great characters in literature, OPEC took some decisive actions which caught many observers by surprise and dramatically changed the course of events. First, by not cutting production in November 2014, triggering a collapse in prices, and then last November agreeing, along with 10 non-OPEC producers, to a production cut totalling 1.8 Mb/d.
How should we think about these actions?
For me, the clearest explanation of these actions was given by HE Khalid Al-Falih, the Saudi Arabian minister for energy, industry and mineral resources at CERAWeek in March. To quote minister Al-Falih:
“OPEC remains an important catalyst to the stability and sustainability of the market…. but history has also demonstrated that intervention in response to structural shifts is largely ineffective… that’s why Saudi Arabia does not support OPEC intervening to alleviate the impacts of long-term structural imbalances, as opposed to addressing short-term aberrations….”
To unpack this a bit: OPEC’s power stems from its ability to shift oil production from one period to another. As such, it has the ability to smooth through the effects of temporary shocks to the oil market, lowering or raising production until the shock subsides.
But its ability to respond to permanent shocks is far more limited: shifting supply from one period to another makes little difference if the underlying shock persists. Consider, for example, the unsuccessful attempts by OPEC to support the oil market in the first half of the 1980s as new structural sources of production from the North Sea and Alaska came onstream.
The underlying source of the supply imbalance that emerged in 2014 was the growth of US tight oil. To use the Minister’s words, this was not a short-term aberration; it was the emergence of a new source of intra-marginal supply.
In contrast, the focus now is on increasing the pace at which the huge overhang of oil stocks is drawn down to more normal levels. This is exactly the type of temporary adjustment in which OPEC intervention can be effective - reducing supply until stocks have adjusted.
So perhaps like all the best stories, the actions of the main characters make perfect sense when seen in the right context. OPEC remains a central force, able to manage and stabilize the oil market, but the nature of that power means it is effective for short-term aberrations, not structural shifts.
Finally for oil, what has all this meant for prices?
The persistent supply imbalance and growing inventory levels caused prices to fall towards the end of 2015 and into 2016. Prices stabilized through the middle of the year as the market moved into balance and inventories levelled off, before firming somewhat towards the end of the year in the wake of the OPEC/non-OPEC agreement. Dated Brent averaged $44 per barrel in 2016, down from $52 in 2015, its lowest (nominal) average since 2004. So far, this year, prices have averaged about $53 as the OPEC cuts have started to take effect, albeit partially offset by the strong recovery in US tight oil.
Back in 2015, refiners responded to near-record high margins by increasing refinery throughput by 1.8 Mb/d, triple its 10-year average.
That led to a build-up of product inventories that dampened margins in 2016, causing refineries to reduce the growth in crude runs to just 0.6 Mb/d.
After strong growth in 2015, European refiners reduced runs by 0.2 Mb/d, while runs in Mexico, Venezuela and Brazil fell by a combined 0.4 Mb/d due to multiple refinery shutdowns.
Refining capacity expanded by just 0.4 Mb/d, less than half its 10-year average. This was the second successive year of weak growth in refining capacity, much of which can be attributed to actions by China to limit the build-up of domestic spare refining capacity.
The fortunes of coal appear to have taken a decisive break from the past. This shift largely reflects structural factors: the increasing availability and competitiveness of natural gas and renewables, combined with government and societal pressure to shift towards cleaner, lower carbon fuels.
These long-term forces in turn have given rise to near-term tensions and dynamics. This was particularly the case in China, which at the beginning of the year introduced a series of measures to reduce the scale of excess capacity in the domestic coal sector and improve the productivity and profitability of the remaining mines.
These measures were focused on reducing capacity amongst the smallest, least productive mines and encouraging greater consolidation. In addition, the government further constrained production by restricting coal mines to operate for a maximum of 276 days, down from 330 days. The impact of these measures was dramatic: domestic coal production fell sharply and prices jumped sharply higher. For 2016 as a whole, Chinese coal production fell by 7.9% (-140 mtoe), by far the largest decline on record, and through the year the price of steam coal increased by over 60%. Coal consumption also declined (-1.6%, -24 mtoe) for the third consecutive year, although by less than production, with China resuming its position as the world’s largest importer of coal.
The events in China spilled over into global coal markets, with world prices taking their cue from China. This rise in global coal prices further depressed global coal demand, particularly in power sector around the globe, with natural gas and renewable energy the main beneficiaries. Global coal consumption fell by 53 mtoe (-1.7%) and global production by a whopping 231 mtoe (-6.2%), with US production registering a second consecutive substantial fall (-19.0%, -85 mtoe).
A particularly striking example of this long-run movement away from coal was here in the UK, where the hike in global coal prices was amplified by the increase in the UK’s Carbon Price Floor in 2015. As a result, the UK’s relationship with coal almost completed an entire cycle: with the UK’s last three underground coal mines closing, consumption falling back to where it was roughly 200 years ago around the time of the industrial revolution, and the UK power sector recording its first-ever coal-free day in April of this year.
Global consumption increased by 1.5% (63 bcm), quite a bit weaker than its 10-year average (2.3%); while global gas production was essentially flat (0.3%, 21 bcm), the weakest growth in gas output for 34 years, other than in the immediate aftermath of the financial crisis.
This sub-par growth went hand-in-hand with falling gas prices - Henry Hub prices were 5% lower than in 2015, European and Asian gas markers were down 20-30% as prices continued to adjust to increased LNG supplies.
Much of the lacklustre performance can be traced back to the US, particularly on the supply side where falls in gas (and oil) prices caused US gas production (-17 bcm, -2.5%) to fall for the first time since the US shale gas revolution started in earnest in the mid-2000s.
Outside of the US, on the demand side, gas consumption in Europe rose strongly (6%, 28 bcm) helped by both the increasing competitiveness of gas relative to coal and weakness in European nuclear and renewable energy.
The Middle East (3.5%, 19 bcm) and China (16 bcm, 7.7%) both also recorded strong increases aided by improving infrastructure and availability of gas. The largest falls were in Russia (-12 bcm, -3.2%) and Brazil (-5 bcm, -12.5%) both of which benefited from strong increases in hydropower.
On the supply side, Australian production (19 bcm, 25.2%) was the standout performer as several new LNG facilities came onstream.
Looking at the growing market for LNG, although China continued to provide the main source of growth, it’s striking that the increasing availability of supplies has prompted a number of new countries, including Egypt, Pakistan and Poland, to enter the market in the last year or two. These new entrants were helped by the increased flexibility afforded by plentiful supplies of FSRUs (floating storage and regasification units).
2016 was the first year of the growth spurt we expect to see in LNG, with global supplies set to increase by around a further 30% by 2020. That is equivalent to a new LNG train coming onstream every two-to-three months for the next four years - quite astonishing growth.
As the importance of LNG trade grows, global gas markets are likely to evolve quite materially. Alongside increasing market integration, we are likely to see a shift towards a more flexible style of trading, supported by a deeper, more competitive market structure. Indeed, this shift is already apparent, with a move towards smaller and shorter contracts and an increase in the proportion of LNG trade which is not contracted and is freely traded.
A particularly interesting market in the context of the growing LNG supplies is Europe.
On the one hand, Europe’s large and increasing need for imported gas, combined with its relatively central location amongst several major LNG suppliers, means Europe is often highlighted as a natural growth market for LNG. On the other hand, Europe’s access to plentiful supplies of pipeline gas, particularly from Russia, means LNG imports are likely to face stiff competition.
In terms of this battle of competing supplies, Round 1 went to pipeline gas.
Europe’s gas imports increased markedly last year, reflecting the strong increase in demand, together with weakness in the domestic production of natural gas. But virtually the entire rise in European imports was met by pipeline gas, from a combination of Algerian and Russian supplies, with imports of LNG barely increasing.
The economic incentives in this battle of competing supplies are clear: just as with OPEC’s response to the emergence of US tight oil, Russia has a strong incentive to compete to maintain its market share in the face of growing competition from LNG supplies.
But this competitive process is complicated by possible concerns about Europe being overly dependent on a single source of supply and the energy security issues this might raise. The interesting question is whether the growth of global LNG trade, by fostering a more globallyintegrated gas market, with the optionality of being able to turn to LNG should the need arise, might mitigate those concerns.
Europe doesn’t need to consume large amounts of LNG imports in ‘normal’ times, but it has the option of doing so if the need arises.
The leading light of the energy transition is, of course, renewable power which continued to grow rapidly last year, led by wind (15.6%, 131 TWh) and solar (29.6%,77 TWh). Although the share of renewable power within primary energy edged up only slightly to 3.2%, its strong growth meant it accounted for over 30% of the increase in primary energy.
China continued to dominate renewables growth, contributing over 40% of global growth - more than the entire OECD - and surpassing the US to become the largest producer of renewable power.
One noticeable weak spot last year was the EU, where renewable power barely grew as load factors fell back from unusually high levels in 2015. This is a reminder of the variability that weather conditions can inject into renewable generation from year to year. For example, the decline in Denmark’s wind power last year was almost 5% of its total power generation.
Although wind continued to provide the lion’s share of the increase in renewable power, solar is catching up fast.
The right-hand chart considers the 67 countries that are separately tracked in the Statistical Review and records the share of those countries that, in any given year, produced a material amount of different energies. It took around 20 years for the share of countries producing a sizeable amount of wind power to increase from 15% to 75%; solar achieved the same degree of diffusion in less than half that time. In sharp contrast, nuclear energy plateaued at less than half the number of countries.
These different rates of diffusion reflect the different characteristics of the technologies: the more modular nature of solar power, together with its steeper learning curve has allowed it to spread more quickly.
Moreover, the fact that the transfer of wind and solar technology is not subject to onerous security restrictions has helped their rapid diffusion relative to nuclear power.
In terms of other non-fossil fuels: China provided the main source of world growth for both hydro (2.8%, 120 TWh) and nuclear (1.3%, 41 TWh) power. Growth in China’s hydro power has slowed sharply in recent years from the rapid rates of expansion that characterized the first part of the 2000s. In contrast, China’s nuclear programme is just beginning to ramp up: it brought on five new reactors last year - the largest ever annual increase in China’s nuclear history - and has more than 20 reactors currently under construction.
Turning finally to carbon emissions. The good news is that carbon emissions were essentially flat in 2016. This is the third consecutive year in which we have seen little or no growth in carbon emissions - in sharp contrast to the 10 years before that, in which emissions grew by almost 2.5% per year. Some of this slowdown reflects weaker GDP growth, but the majority reflects faster declines in the carbon intensity of GDP - the average amount of carbon emitted per unit of GDP - driven by accelerating improvements in both energy efficiency and the fuel mix.
The key question this raises is whether the experience of the past three years signals a decisive break from the past and a significant step towards the goals of Paris or was it largely driven by cyclical factors which are likely to unwind over time?
Long-run transition or short-run adjustment?
Looking at the factors driving this improvement, the key difference is China. China’s carbon emissions are estimated to have actually fallen over the past two years, after growing by more than 75% in the previous 10 years.
As mentioned earlier in the context of the slowdown in China’s energy consumption, there are good reasons for thinking that some of this improvement in China’s carbon emissions reflects structural factors that are likely to persist: slower economic growth; a shift in the composition of growth towards less energy-intensive sectors, and a movement away from coal. But some probably reflects cyclical factors, particularly the contractions in some of China’s most energy intensive sectors, which are unlikely to keep being repeated and may well unwind in future years.
The juxtaposition of short-run adjustments and long-run transition is likely to be a feature of energy markets for many years to come.
Group chief economist
- This is a shortened version of the presentation given at the launch of BP’s Statistical Review of World Energy in London on 13 June 2017