Well, good afternoon, everyone. Thank you for those here in London who are joining us here today and everybody joining us on the global webcast. We've got thousands of people on the line as well as you here in London. I see some new faces in the room today, and many familiar ones as well. You are all very welcome to be here. I appreciate you all taking the time to be here for what is now the 68th annual edition of the BP Statistical Review. I've introduced eight or nine editions now, and I usually start by saying what a remarkable year it's been. Does that sound familiar? It's true again this year. Even more so, I think. 2018 was just astonishing. There have been some incredible developments, and not all for the better.
Spencer will work through the details in a few minutes, but first, let me start off with my three big takeaways from 2018. The first of those is by far the most unwelcome development. The world needs carbon emissions to fall dramatically, but they continue to grow. Energy-related emissions are not just growing, they are accelerating in 2018, increasing at their fastest rate for seven years. This rising trend in emissions is coupled with the biggest growth in energy consumption for nearly a decade, driven by China and India, something we've become used to. Also by demand growth in the U.S. as well. Now, that connects with the second remarkable development last year, which is the ongoing energy revolution in the U.S. The juggernaut continues to roll, with the U.S. delivering the largest ever annual increases in both oil and gas in 2018.
Those are record increases, not just for the U.S., but for any country in the 68 years of the BP Statistical Review. With the vast majority of those increases coming from onshore shale plays. That really is quite something and goes to show what an energy powerhouse the U.S. has become again. The reverberations of that are being felt well beyond the energy sector, from the standpoint of market balance, global trade, and also geopolitically. That brings me to the third big takeaway for me, which is continuing, which is the electrification of the world. In a big year for demand, energy demand growth overall, power demand increased even more strongly, which was up by 3.7%. Renewables accounted for a third of that growth and had another strong year overall, growing by 14.5%, with the increase in renewable generation just a little shy of its record increase in 2017.
Natural gas also contributed as part of a 5% increase in consumption and production. That's the biggest growth in gas since the early 1980s. Coal also took a share of the growth in power this year, with the overall consumption of coal increasing for a second year in a row. That growth follows three years of decline and is a worrying trend given that the potential to reduce emissions by increasing electrification while pushing coal out of power. It's important. When it comes to cutting emissions, electrification without decarbonization is, quite frankly, of little use. Renewable energy has a vital role to play, but even growing as fast as it is or faster still, it is unlikely to do it on its own. I'm sure it's not going to do it on its own.
As I've said before, this is not a race to renewables, but a race to reduce carbon emissions on many fronts. Looking at this year's statistical review as a whole, what I see is a reflection of all the reasons for the rising concern being shown by people around the globe, including at the doorstep here in St. James's a few weeks ago. As I said at CERAWeek in Houston earlier this year, and as our Chairman, Helge Lund, recently wrote in the Financial Times, the world is not on a sustainable path. If anything, we are moving further away from it rather than getting closer. There has to be a greater sense of urgency because the longer carbon emissions continue to rise, the harder and more costly the necessary eventual adjustments to net carbon zero emissions.
The science is telling us that, and so is the latest data in the statistical review. On that point, let me pause now. I'm going to hand it over to Spencer to take you through the details. Spencer and the team, once again, have done a remarkable piece of work, tremendous job compiling the numbers, bring them into such clear focus, along with the cooperation of governments around the world. Thank you, Spencer, and to your team. Over to you.
Thank you, Bob. Let me add my thanks to everybody for sparing the time to come to today's launch of the BP Statistical Review of World Energy 2019, both here in London, in St. James's. Bob said there are many familiar faces, welcome to you. Also for everybody watching round the world via the web. I think at last count, we had about 8,000 people registered, which is well in excess of any numbers we've ever had before. Thank you for watching online. I know it's a lot harder watching online than it is being in the room. Please stay with us. As a sort of special incentive to stay with us, we're going to do a survey. First time we've done this. We're going to do a live survey for those people watching online.
At the point when I finish the presentation, in a half an hour during the Q&A, we're going to send you a link to do a live survey, and then we're going to try and get those results back into the room before we close today. That's the plan. See if it works. Who knows? You're all very welcome. Let me also take this opportunity to thank everybody who's been working hard over the last couple of months putting the stats review together. There are a lot of data to collect, check and compile. Many thanks to the rest of the economics team, many of whom are here, others are scattered around the world watching via the web as well. Many thanks to them. Also to the team from Heriot-Watt University, led by Mark Schaffer and Erkal Ersoy.
Heriot-Watt are our longstanding partners in the production of the Stats Review, so thank you very much to Heriot-Watt as well. As Bob said, this is the 68th edition of the Stats Review. As I've traveled around the world with BP over the past few years, I've come to appreciate the esteem and affection for which the review is held. A really tangible example of this is when I meet people in different parts of the world, and they pull me to their bookcase, and they proudly show me a set of reviews going back 20 or 30 years standing on their bookcase. I must admit, at first, I'm a little perplexed by this. Why keep a copy of last year's Stats Review once a new review has been published? You can see I'm not very sentimental sort of type, especially if all the updated data is available online.
I came to realize that the commentary that's also included in the review provides a snapshot of the issues dominating the industry in that year. It's a bit like when you buy one of those birthday cards reproducing headlines, newspaper headlines, from the year in which somebody was born. They sort of jolt your memory of different events. I think that's how some people use their collection of Stats Reviews. In that context, my guess is that those people will look back at the reviews around this time, and they will observe a world in which there is a growing societal awareness and demands for urgent action on climate change, but where the actual energy data continued to move stubbornly in the wrong direction, a growing mismatch between hopes and reality.
In that context, I fear, or perhaps I should say, I hope, that 2018 will mark the peak in this mismatch. As people protested, schoolchildren went on strike, and shareholders passed resolutions, energy demand and carbon emissions in 2018 grew at their fastest rate for years. The review can't solve this mismatch, but it can provide an objective assessment of the factors driving energy developments in 2018, and their possible implications for the future. What happened last year? The headline grabbing numbers are the rapid growth in both energy demand and carbon emissions. Global primary energy, shown here in purple on the left, grew by 2.9% in 2018, the fastest growth since 2010. This, despite a backdrop of modest GDP growth and strengthening energy prices. At the same time, carbon emissions from energy use, shown here on the right in gray, grew by 2%.
The fastest expansion seen for many years, with CO2 emissions from energy use increasing by around 0.6 gigatons. To put that sort of number as 0.6 gigatons into some sort of perspective, that's roughly equivalent to the carbon emissions associated with increasing the number of passenger cars on the planet by a third, another 400 million cars. These are material increases. What drove the increases last year, and how worried should we be? Let's start first with energy consumption. As I said, energy demand grew by 2.9% last year. This growth was largely driven by China, the U.S., and India, which together, those three countries accounted for around two-thirds of the growth in energy demand.
Relative to recent trends, the most striking growth was in the U.S., where energy consumption increased by a whopping 3.5%, the fastest growth for 30 years, in sharp contrast to the decline seen over the previous decade. The strength in energy consumption was pretty much reflected across all the fuels, most of which grew more strongly than their historical averages. As you can see from this chart, this acceleration was particularly pronounced in natural gas, shown here in red, which increased by over 5%, one of its strongest growth rates for over 30 years, accounting for almost 45% of the entire growth in energy consumption last year. Growth in renewable energy, shown here in orange, eased back slightly relative to past trends, although it remained by far the world's fastest-growing source of energy.
As I hope you know by now, when we do think about renewables in the Stats review, we separate out hydro separately, shown here in blue. Renewable energy, when I'm going to talk about renewable energy today, I'm talking about wind, solar, and biofuels. In terms of why energy demand was so strong, this next chart provides a way of sort of gauging the extent of the surprise in last year's energy data. The chart compares the actual growth in energy demand, shown here in purple, against a predicted line derived from a simple framework, which uses GDP growth and changes in oil prices to predict energy growth at a country level, then aggregates to global energy.
Although very simple, this framework is able to explain much of the broad contours in energy demand over the past 20 years, completely fails to explain the sharp pickup in energy growth last year. Indeed, the framework suggests that energy demand should have slowed last year, reflecting the slightly weaker economic outlook and the strengthening in energy prices. What's going on? Digging into the data further, it seems that much of the surprising strength in last year's energy consumption may be related to weather effects.
In particular, there was an unusually large number of both hot and cold days last year, which led to higher energy consumption as families and businesses switched on their air conditioning and heaters. The increasing number of heating and cooling days was pretty widespread across many of the world's major demand centers, particularly in the U.S., China, and Russia, helping to explain the strong growth in energy consumption in each of those countries. To give you a little bit more context of what's going on here, this next chart shows a measure of U.S. heating and cooling days, which combines both the frequency and intensity of unusually hot and cold days. This is just for the U.S.
What was particularly unusual about the U.S. last year was there was an increase in both heating and cooling days last year, so both the blue line and the red line ticked upwards. Whereas in the past, high numbers of heating days have tended to coincide with low numbers of cooling days, or vice versa. This double whammy meant you had the increase in the combined number of hot and cold days, shown here in the green line, was at its highest level since the 1950s, boosting energy demand growth in the U.S. If we go back to our simple framework, then we now ignore or make that framework to include a measure of heating and cooling days for those countries which those data are available, this greatly reduces the extent of surprise in last year's framework.
Now my framework is now pointing to quite a material increase in energy demand growth last year. Indeed, this analysis suggests that the stimulus from weather effects in U.S., China, and Russia alone could account for around a quarter of the increase in consumption last year. Once these weather effects are included, the growth in energy demand still looks a little bit strong. The growth in purple is still a little bit higher than the orange. What's even more pronounced now is the weakness of energy growth in 2014, 2015, and 2016. Actual energy growth in 2014, 2015, and 2016, are a lot weaker than what the model and the framework would suggest. Regular followers of the stats review may recall that much of this weakness appears to stem from the pattern of Chinese economic growth during this period.
In particular, the pronounced weakness of some of China's most energy-intensive sectors, iron, steel, and cement. If you look at these three sectors, iron, steel, and cement, in the 10 years up until 2014, their average growth rates was well in excess of 10%, growing really rapidly, and they slowed exceptionally sharply in this 2014 to 2016 period, such that output in all three sectors fell in outright terms. This matters because these three sectors together account for around a quarter of China's energy consumption. This sharp slowdown in these sectors pulled down on China's energy growth and pulled down on overall global energy consumption. At the time, I speculated that some of this slowing in these sectors reflected the rebalancing of the Chinese economy towards more consumer and service-facing sectors, and so was likely to persist.
I also noticed that the scale of the slowdown suggested that some of it was likely to be cyclical and would reverse over time. Indeed, that's what's began to happen in iron and steel. It began to happen last year, in 2017, and it gathered pace even more strongly in 2018. You see this pickup in output growth in these two sectors. If we go back to our framework, including weather effects, and adjust it to also take account of movements in these key Chinese industrial factors sectors, you end up with this green predicted line.
If we take off the earlier line to make it clear, you can see the framework now is able to account pretty much for all the movements in energy growth over the past four or five years, both the weakness in energy growth in 2014 and 2016, and this strong pickup in energy demand growth last year. In answer to the question of why energy demand growth was so strong in 2018, it appears that the strength of demand was largely due to weather-related effects, especially in U.S., China, and Russia, together with a further unwinding of these cyclical industrial factors in China. The next question is, how does this strong growth in energy demand relate to the worrying acceleration in carbon emissions?
Arithmetically, the growth in carbon emissions in any given period is equal to the growth in energy, offset by any improvement in the carbon intensity of the fuel mix. In terms of this chart, the growth in carbon emissions here in gray is simply the net of the growth in energy, in purple, offset by any improvement in the fuel mix. As such, you can see it's pretty clear from this chart that the growth in carbon emissions is simply a direct consequence of the increase in energy growth. Growth in energy demand was about 1.5 percentage points higher than its five-year average, and the growth in carbon emissions was about 1.4 percentage points higher. One simply led to the other.
The pace of improvement in the fuel mix, shown here in yellow, was pretty much in line with its historical average, helped by that rapid growth in natural gas gaining share relative to coal and oil. Given the strong growth in energy, there's no puzzle surrounding the strength in carbon emissions. The final question to ask in terms of the headline growth in energy and carbon emissions last year is what might it say about the future? How worried should we be? I think this depends in large part on how you interpret the increasing number of heating and cooling days last year.
I should stress this is not my area of competence, but just in a general sense, there seem to be at least two possibilities. On the one hand, to the extent that the unusually high number of hot and cold days last year just reflects random variation, we might expect weather effects in the future to revert to more normal levels, allowing the growth in energy demand and carbon emissions to fall back. On the other hand, if there is a link between the growing levels of carbon in the atmosphere and the types of weather patterns observed last year, this would raise the possibility of a more worrying vicious circle, with increasing levels of carbon leading to more extreme weather patterns, which in turn trigger stronger growth in energy and carbon emissions as households and businesses seek to offset their effects.
As I said, there are zillions of people better qualified than I to make judgments on this, I'll just leave that to others. Even if these weather effects are short-lived, such that the growth in energy demand and carbon emissions slow over the next few years, the recent trends still feel very distant from the types of transition paths consistent with meeting the Paris climate goals, the mismatch between hopes and reality. That's what I wanted to say in terms of the headline data on energy demand and carbon emissions. My plan now is to consider the key fuels in a little bit more detail, starting first with oil.
2018 was another rollercoaster year for oil markets, with prices starting the year on a steady upward trend, reaching the dizzying heights of around $85 a barrel in October of last year before plunging recently to about $50 by the end of the year. I really hate rollercoasters. Oil demand provided a relatively stable backdrop for all this excitement, continuing to grow robustly, increasing by 1.4 million barrels a day last year. In an absolute sense, the growth in demand was dominated by the developing world, with China in dark blue and India in light blue, accounting for around two-thirds of that growth. Relative to the past 10 years or so, the big outlier was the U.S., where oil demand grew by 0.5 million barrels a day in 2018, its largest increase for well over 10 years.
This growth in U.S. demand was boosted by increased use of ethane, as new production capacity came on stream, destined ultimately for the petrochemical sector. The increased importance of petchems in driving oil demand was also evident in the global product breakdown, with products most closely related to petrochemicals, LPG, naphtha, and ethane, shown here in purple, accounting for around half of the overall growth in demand last year. Against this backdrop of solid demand growth, all the fun and excitement of the fair came from the supply side, where global production grew by a whopping 2.2 million barrels a day, more than double its historical average. You can see from this massive green in 2018, the vast majority of this growth was driven by U.S. production, which grew by over 2 million barrels a day, almost entirely due to increases in tight oil and natural gas liquids.
The increase in U.S. oil production last year was the largest ever annual increase by any country. Since 2011, and the onset of the shale revolution, U.S. oil production has increased by over 7 million barrels a day. That's broadly equivalent to Saudi Arabia's entire crude oil exports. An astonishing increase, which has transformed both the structure of the U.S. economy and the dynamics of global oil markets. In that context, there's been much talk recently about the change in structure of Lower 48 production, as big oil has increased its footprint, and the incentives for consolidation to exploit the benefits of scale and contiguous acreage have increased. Surprisingly, or at least it was to me, this process of consolidation is not yet evident in the increasing concentration of production amongst the top U.S. tight oil producers.
As shown here in the chart on the right, the share of U.S. tight oil production accounted for by the top 10 producers has declined pretty consistently over the last 10 years, so no sign of increasing concentration. That said, there is some evidence of increasing concentration in investment spending, shown here on the right, which may be a leading indicator for production. The corresponding ratio for U.S. shale gas production has increased over the past couple of years to around 55%, so quite a bit higher than its counterpart for oil. Levels of concentration may start to increase in coming years.
I think this matters because the extent to which these concentration ratios do change over the next few years could have an important bearing on the future dynamics of U.S. tight oil production, particularly in terms of its responsiveness to both oil prices and the availability of capital. The intuition here is large companies have bigger balance sheets and so are able to smooth through variations in oil prices and availability of capital, may make production less sensitive. Watch this space for future developments. Switching from U.S. production to OPEC, shown here in yellow, OPEC production fell by 0.3 million barrels a day in 2018, with a marked increase in Saudi Arabia production offset by falls in Venezuela and Iran.
This year-on-year comparison doesn't do justice to the intra-year twists and turns in OPEC production as the rollercoaster gathered pace. The ride began in the first half of 2018 with the continuation of the OPEC+ agreement dating from the end of 2016. The OPEC+ group consistently overshot their agreed production cuts during 2017, this overshooting increased further during the first half of last year, largely reflecting continuing falls in Venezuelan output. You can see output declining, of the OPEC+ group, declining consistently through the first half of 2018, getting further and further below their target cuts. As you can see on the right, these production cuts help push OECD inventories below their five-year moving average for the first time since the oil price collapse in 2014.
The first major twist came in the middle of 2018 amid growing concerns surrounding the possible scale of future supply disruptions. Venezuelan production was continuing to fall. Moreover, the U.S. announced in May its intention to impose sanctions on all Iranian oil exports. In response, the OPEC+ group in June committed to achieving 100% compliance of their production cuts as a group as a whole. The way to think about this 100% compliance is for production to be at the target. Production being significant below the target suggested compliance of sort of 200%, it's overachieving too much. This commitment to 100% compliance contained two important signals. First, as I've just shown you, given the extent to which production was below the target level, it signaled the prospect of an immediate increase in production.
Second, perhaps even more importantly, it helped reduce the uncertainty associated with the possibility of future disruptions to either Iranian or Venezuelan output. Since the commitment to maintain 100% compliance, in essence signaled the willingness of other members of the OPEC+ group to step in and offset any lost production. That's exactly what happened between May and November of last year. Net production by the OPEC+ group increased by almost one million barrels a day, close to achieving that 100% compliance, despite production in Iran and Venezuela falling by a further one million barrels a day. Job done. Or was it? The problem with roller coasters is that as soon as you think you can relax, you get surprised by something else.
In this case, oil production by Libya and Nigeria, neither of which were part of the OPEC+ agreement, increased by almost 600,000 barrels between June and November of last year. As a result, rather than OECD inventory stabilizing, they started to grow again, reversing much of the earlier falls. The growing sense of excess supply was compounded by the U.S. announcing in November that it would grant temporary waivers for some imports of Iranian oil. This triggered yet another twist. A new OPEC+ group was formed in December of last year. This time excluding Iran and Venezuela, as well as Libya, but including Nigeria, with a commitment to reduce production by 1.2 million barrels a day relative to October 2018 levels. After a slow start, by the spring of this year, this reconstituted OPEC+ group was once again overshooting its production cuts.
These production cuts, together with further sharp falls in Venezuelan output, have been sufficient to cause inventories to fall back to around their five-year average. It's tempting to interpret these twists and turns as somehow indicative of OPEC's waning powers, but I'm not sure that's the correct interpretation. The role that the OPEC+ group played in more than offsetting the substantial falls in Iranian and Venezuelan output last year was very significant. Rather for me, the twists and turns simply reflect the difficulty of trying to stabilize global oil markets, especially at a time of record supply growth in one part of the world and heightened geopolitical tensions in other parts of the world. It feels like the roller coaster will run for some time to come. To turn next to natural gas.
As I mentioned, 2018 was a bonanza year for natural gas, with both consumption and production increasing by over 5%, one of the strongest growth rates of both demand and output for over 30 years. The main actor here was the U.S., in green, where U.S. demand accounted for around 40% of the growth in demand and production for over 45% of the global increase in U.S. gas production. U.S. gas production increased by about 86 billion cubic meters last year, driven by shale gas plays in Marcellus, Haynesville, and the Permian. As Bob said, the U.S. achieved a unique double first last year, recording the single largest ever annual increases in gas production as well as oil production. In case there was any doubt, the U.S. shale revolution is alive and kicking.
Although some of the increase in U.S. gas supplies was used to feed the three new LNG trains which came on stream in the U.S. last year, the majority was used to quench the thirst of domestic demand. U.S. gas demand increased by 78 billion cubic meters last year. To give you some sense of that's roughly the same growth as achieved over the previous six years in the U.S., or broadly equivalent to the entire gas consumption of the U.K. A big increase. You perhaps won't be surprised to hear that this exceptional strength appears to be driven by those same weather effects, with rising demand for space heating and cooling fueling increased consumption both directly in commercial and residential buildings, and more importantly, indirectly via growing power demand, shown by this purple bar.
You can see this enormous growth in gas going into the power sector as that demand for power increased. Overall, the growth in gas-fired power generation is estimated to have accounted for around half of the increase in U.S. gas consumption last year. Outside of the U.S., China's gas consumption also expanded strongly, growing by an astonishing 18%. This strength stemmed largely from a continuation of environmental policies encouraging coal to gas switching in industry and buildings in order to improve local air quality. These coal to gas switching policies have been instrumental in increasing China's gas consumption by over a third in the past two years.
Official estimates suggest that as many as 10 million households, that's roughly half the number of households here in the U.K., switched from coal to gas boilers over this two-year period, with even greater switching in the industrial sector, shown here in red. The pattern of demand, very different to that what we just saw in the U.S. In the U.S., it is very much a power story. Here, you're seeing increasing use of gas in industry and in buildings, swapping away from coal. Nothing to do with the power sector at all. Importantly, a series of improvements in import capacity, gas distribution, and demand management meant that this second successive year of rapid growth in Chinese gas consumption was achieved largely without a repeat of the price spikes and shortages which characterized the previous winter.
Global LNG supplies continued their rapid expansion last year, increasing by almost 10% last year, driven by a combination of Australia, the U.S., and Russia. I really like this chart because it's a really simple way of seeing the sort of two big cycles that we've seen in LNG expansion, LNG exports over the last 10 years. You see this first wave around the turn of the decade, driven largely by Qatar, and then more recently, this second wave, driven by Australia and the U.S. of expanding LNG supplies. For much of the past year, the strength of Asian gas demand, led by China, was sufficient to absorb these increasing supplies in 2018.
A waning in the strength of Asian demand towards the end of the year, particularly in the back of a mild winter, combined with a mini-surge in LNG exports, caused Asian spot prices to fall back. Asian prices, shown here in blue, have fallen further during the first part of this year towards the bottom of this gray band. This top of this gray band is defined by an estimate of U.S. exporters' full cycle costs, which include a return on capital, and the bottom of the band of an estimate of U.S. exporters' pure operating costs. The economics of this is that you wouldn't expect prices to fall materially below the bottom of this band for an extended period, with exporters preferring to curtail supply if they can't even cover their operating costs.
Given where prices are now, combine that with the prospect of further substantial expansion of LNG supplies over the next couple of years, means there's a possibility of the first meaningful curtailment of some LNG supplies over the next few years. If we turn next to coal. 2018 saw a further bounce back in coal, building on the slight pickup seen in the previous year, with both coal demand and output increasing at their fastest rates for five years. The growth in coal demand was the second consecutive year of increases following three previous years of decline in coal demand. The peak in global coal consumption, which many had thought had occurred in 2013, now looks less certain. Another couple of years of similar increases to that seen last year would take global consumption comfortably above 2013 levels.
India and China accounted for the vast majority of the increases in demand, with the lion's share going into the power sector. This is despite renewable energy growing by over 25% in both India and China last year. Really rapid growth in renewables, this was not sufficient to keep pace with the strong growth in power demand, with coal being sucked into the power sector as a balancing fuel. This highlights an obvious, I think really important point. Even if renewables are growing at truly exceptional rates, they were growing at over 25% in both China and India last year. The pace of growth of power demand, particularly in developing Asia, limits the speed at which the power sector can decarbonize. We can explore this point in more detail if we look directly at the power sector.
As I have bored many of you for several years now, the power sector needs to play a central role in any transition to a low carbon energy system. It's the single largest source of carbon emissions, and it's where much of the lowest hanging fruit lie in terms for reducing carbon emissions over the next 20 years. What happened in power last year? Global power demand grew by 3.7%, which is one of the strongest growth rates for 20 years, absorbing around half of the growth in primary energy. The developing world continued to drive the vast majority of this growth, led by China and India, who together accounted for around two-thirds of this growth, again, shown by the dark blue and the light blue.
The particularly strong growth of power demand in 2018 owed much to the U.S., shown in green, where power demand grew by a bumper 3.7%, boosted by those weather effects, in sharp contrast to the slight trend decline over the past 10 years. On the supply side, the growth in power generation was led by renewable energy, here in orange, which grew by over 14%, close to its highest ever increase in power generation. Even so, this contributed only around a third of the growth in overall power generation, with also sizable contributions from both coal and natural gas. China continued to lead the way in renewables, accounting for 45% of the global growth in renewable power generation, more than the entire OECD combined.
We step back just for a moment from last year's data and look at the growth of renewables over a longer period, this histogram is based on the 78 major countries for which the review publishes individual data and shows the number of countries 20 years ago, in 1998, which had any form of renewable power. At that point, only 28 countries, around a third of our sample, were using renewables, with the vast majority of those countries only using very small amounts of renewable energy, so clustered in this range of domestic generation of 1%-5%. We fast-forward 20 years to 2018, the picture changes quite substantially. The number of countries that have adopted some form of renewable energy has increased to 58, so three-quarters of our sample, with a degree of penetration pretty much evenly spread across the whole range.
Renewable energy has come of age. To repeat a point I made last year, despite the increasing adoption and penetration of renewable power, the fuel mix in the global power system remains depressingly flat. With the 2018 shares of both non-fossil fuels, in turquoise, and coal, in black, unchanged from the levels 20 years ago. Three flat lines. This persistence in the fuel mix highlights a point which the International Energy Agency and others have stressed recently. Namely, that a shift towards greater electrification helps as a pathway to a lower carbon energy system only if it goes hand-in-hand with a decarbonization of the power sector. Electrification without decarbonizing power is of little use. In that context, I found this chart really interesting.
It breaks down the growth in carbon emissions from the power sector, shown by this brown line, into the contribution from increases in power demand, shown in purple, and shifts in the fuel mix going into the power sector in blue. In terms of the most recent data, carbon emissions from the power sector are estimated to increase by 2.7%, their fastest rate of growth for seven years, accounting for around half of the growth in global carbon emissions. Looking at the longer sweep of history, you can see for much of the past 20 years, up until the last five years or so, changes in the power sector fuel mix have been relatively small or even perverse. These blue bars were either positive, suggesting the fuel mix were getting worse, or only very slightly negative. As a result, increases in power demand fed directly into higher carbon emissions.
Electrification without decarbonization. The good news is this has improved a bit over the past five years or so. The rapid growth in renewable energy, together with an edging down in the coal share, has led to a more sustained improvement in the fuel mix. These blue bars have turned consistently negative, such that the impact of increasing power demand on carbon emissions has been partially offset. It's still only partial. Despite the rapid gains in renewable energy, the pace of growth in power demand has meant that carbon emissions from the power sector has increased pretty substantially over the past three years. You can see this positive growth in carbon emissions. It hasn't been possible to decarbonize the power sector quickly enough to offset the growth in demand.
To give a sense of this challenge, I did a simple thought experiment of the extent to which the power sector fuel mix would have needed to change over the past three years to maintain the level of carbon emissions at their 2015 level for the same growth in overall power demand. In terms of this chart, how much bigger would these fuel mix bars need to have been in terms of how much bigger negative growth from these fuel bars would needed to be to get zero growth in carbon emissions, to get the dotted line instead?
Making the same point in level space, the pace of decarbonization needed to keep the level of carbon emissions at their 2015 level rather than this upward trend level that we actually saw in the data. If we focus first on renewable energy, renewables over the past three years to achieve that big fall in those blue bars, would've needed to grow more than twice as quickly than they actually did. Rather than growing by over 800 terawatt hours over the past three years, renewable generation would've needed to have grown by over 1,800 terawatt hours. A staggering number. That additional 1,000 terawatt hours is roughly equivalent to the entire renewable generation of China and the U.S. combined in 2018.
In addition to the rapid growth in renewables we actually saw, the world would've also needed to add the entire renewable generation of China and the U.S. in just three years, just to keep carbon emissions from the power sector flat. Alternatively, the same outcome for carbon emissions could have been achieved by replacing about 10% of coal in the power sector for natural gas. The intuition here is that renewables are still a relatively small share of power generation relative to coal, and so the proportional movements in coal need to be a lot smaller. I took two points from this simple thought experiment. First, the general point that the robust growth in power demand, particularly in the developing world, greatly adds to the difficulty of decarbonizing the power sector. You have to run very fast just to stand still. Electrification with decarbonization is difficult.
Relying solely on renewables to achieve this is an almost impossible task. To be clear, rapid growth in renewable energy is absolutely essential, but it's unlikely to be sufficient. This highlights the importance of adopting a range of technologies and fuels rather than just relying on renewables. To win the race to Paris, the world is likely to require many fuels and technologies for many years to come. This includes widespread coal to gas switching, significant adoption of carbon capture, use and storage, CCUS, and increasing energy efficiency, especially in the developed world where the vast majority of people already enjoy relatively high levels of energy consumption. Let me conclude.
At a time when society is increasing its demands for an accelerated transition to a low carbon energy system, the energy data for 2018 paint a worrying picture, with both energy demand and carbon emissions growing at their fastest rates for years. As I hoped I showed, in a statistical sense, it's possible to explain this acceleration in terms of a combination of weather effects and an unwinding of cyclical movements in China. What is less clear is how much comfort we can take from this explanation. What does seem clear is the underlying picture is one in which the actual pace of progress is falling well short of the accelerated transition envisaged by the Paris Climate Goals. Last year's developments sound yet another warning that the world is on an unstable path. The mismatch between hopes and reality. It's sometimes said that history is written by the victors.
Although the Stats Review provides a history of sorts, it's a very specific near-term history, documenting the developments in the global energy system in the previous year. In that context, the only victors are the powers of comprehensive objective data. The data write the history. That's the role the review has been playing for the past 68 years. The most important thing about history is to learn from it. Looking ahead, I hope the understanding and insight provided by future statistical reviews will inform judgments and decisions as society navigates its path to a low carbon energy system. Thank you very much. We're going to go into a Q&A session, but before I do, for those people who are watching online, I just want to introduce a survey that we're going to do for you.
For those watching online, you will now be sent a link, hopefully if this technology works, to a survey. The survey's going to ask you three questions, and for each of those three questions, you have five answers ranging between strongly disagree to strongly agree. The first question was, prior to today's presentation, I was concerned about climate change. Question two was, prior to the presentation, I was concerned we would not meet the Paris Climate Goals. Question three was, having watched the presentation, the data makes me more concerned about climate change. Where between strongly disagree to strongly agree. The survey is going to be open for about 20 minutes or so.
Please, as you're listening to the Q&A, if you can go in and tick your three boxes, and then at the end, if the technology works, we'll come back and show everybody the results of how you responded to today's session. With that, I'll pass over to do a Q&A where again, we will do a mix of questions in the room and also online. If we start with some questions in the room first, and if people would just like to catch my eye and just wait for a microphone to come, we will take it from there. Who'd like to go first? If we start with some questions in the room first, and if people would just like to catch my eye and just wait for a microphone to come, we will take it from there. Who'd like to go first?
Sir, you win the race. You are first up. Then we have three questions on my left.
Thank you very much. Very interesting. It's Tom Brown from ICIS News. First off, looking at the trends about weather, and that impact on demand and consumption, it seems like climate change projections say that extreme weather is going to be much more prevalent going forward. If that's going to, as you say, the vicious cycle you mentioned, does that mean that the Paris Agreement's goal of 1.5 to 2 degrees centigrade temperature rise limit is probably impossible? What else could be done, and at what level? Secondly, do you expect petchems to continue to drive energy demand going forward? Thanks a lot.
In terms of weather, as I said, I am not an expert on weather effects. In terms of my amateurish reading, and it is a very amateurish reading, my understanding of where the consensus at the moment is, as a process of global warming takes place, you can think of the distribution of weather shifting slowly to the right as the world gets hotter and hotter. That suggests the number of hot days will increase. Other things equal, the number of cold days will tend to decrease. There are some debates about the extent to which the number of cold days will decrease. I think the consensus is that shifting distribution will give you more sort of hot days, so more cooling, where you need to turn your air conditioning on, but less where you need to turn the heating on.
What we saw last year, particularly in the U.S., where you saw both this increasing number of both hot and cold days, isn't so easy to square with that type of reading of the literature. That's now exhausted entirely my knowledge of weather effects. Beyond that, others should try. You should speak to experts. In terms of petrochemicals, I think in the energy outlook, we highlighted the important role of petrochemicals, particularly in driving oil demand over the next 20 years. Our expectation over the next 20 years is the so-called non-combusted use of oil. The use of oil particularly as a feedstock in the petrochemical sector will be the sort of dominant source of the growth of oil demand over the next 20 years. I've got a cluster of four questions here, so they can fight it out amongst themselves.
These two gentlemen here and one gentleman there.
Start to fight.
It's Morten Frisch. Again, thank you very much, Spencer, for a very good presentation. Regarding the carbon emissions, in particular from gas, when you calculate this, do you take into account what I've seen over the last two years, an increasing amount of flaring? In the Middle East, Iran, Iraq the worst, but all Gulf Cooperation Council countries still flare, although many of them are short of gas. Russia fall in the same category. Lately, very worrying trend in the United States, part of the U.S. energy revolution in Bakken and in the Permian. They haven't managed to build out the pipeline systems, the gas processing quick enough, and we see rising gas oil ratios from the oil production. Is this included, or could you see something about how the industry looked upon this, and how this is going to be handled?
Clearly, that's very important in the picture.
Sort of the technical answer and the bigger answer. The technical answer, what we are calculating here is purely carbon emissions from energy use. It does not take account other forms of greenhouse gases, so it does not take account of methane. That's just technically because that's what we're able to do, and that's pretty much the industry standard. If you look at the IEA's estimates of carbon emissions, they also do just carbon emissions from energy use and do not try to take account of broader greenhouse gases. In that sense, those methane emissions associated with gas and flaring would not be captured in these numbers. Let me not speak for the industry. Let me speak to a little bit in terms of BP.
I think in terms of BP's view, we're recognizing that trying to understand better measuring and controlling any form of methane emissions is a key part for natural gas to play a key role going forward. BP has committed to try and being an industry leader here. We set our own standards for methane emissions, which many others around the industry have adopted. We are spending a lot of time and effort trying to improve the way we both measure methane emissions and try and control methane emissions. I think it is absolutely critical if natural gas is to play its role during this energy transition, that we can better measure and better control methane emissions. I think it's a very significant issue, and certainly we view it as a very significant issue.
Thank you.
It's great that BP is now aligning its plans to the Paris Agreement. When do you expect BP to achieve net zero carbon emissions from both your operations and your products? What plans have you for that?
I think the difficulty with that is, it's very difficult for BP to do that in isolation. If we look at the Paris Agreement, it's clear that they have a target to achieve net zero carbon emissions. If you look at the pledges the government signed up to at the Paris Agreement, they were consistent with carbon emissions continuing to rise. It's very hard, so I don't know. What do you think? Because it's very hard. BP won't be able to do this on its own. When governments at the moment, where we sum up the pledges of all those NDCs and we say, "Well, what does that add up to?" That doesn't add up to carbon emissions falling. Certainly not to zero, but not even falling. It adds up to carbon emissions continuing to rise.
In some sense, I think the question would better be asked of those governments that made those pledges. When do you think the targets they are going to sign up to will be consistent with getting the world to net zero? I have one more question here, and then I'm going to take a couple from this screen.
Spencer, thank you. John Cooper, Fuels Europe, the European refining industry. A question about the power sector and coal versus renewables. We see the cost of renewables, wind and solar, keep falling. You today have told us the story of how coal is actually being built further. Do you have any insights as to why coal fire power stations are being built instead of wind and solar in several parts of the world?
I think it's just literally the pace at which renewable energy can expand. In those countries, China and India, two perfect examples here. They accounted for the vast majority of the increase in coal consumption last year. Renewable generation increased by over a quarter in one year last year. They're not not growing. It's growing enormously rapidly. Power demand is growing even more rapidly. At that point, unless you can build out renewables even more quickly than 25% a year, something else needs to be sucked in as a balancing fuel. That's the way I think we should think about coal.
The share of coal in both of those power sectors fell last year, its relative importance fell in both China and India, but it continued to grow literally just because it's very hard to grow renewable energy sufficiently quickly to meet all that growth in power demand. That was the point I was trying to make in that thought experiment. If we are going to do electrification, we should only be doing electrification if it goes hand in hand with decarbonizing the power sector. Just relying on renewables to do that is an almost impossible task. It's essential, that rapid growth in renewable energy, but unless it can break all sorts of speed limits, unless it can grow. Think about the growth we've seen over the last few years. Astonishing.
You'd have to more than double that to be able to do that, just to keep carbon emissions flat. That's an enormously difficult task. I think that's the way to think of coal. It's just being sucked in as a balancing fuel in those. It's the cheapest fuel, and it's the most readily available fuel, and that's why it is coming in in that way. I'm going to do a couple of questions from online, since there's 150 people in the room, there's about 8,000 people watching online, so it's a bit of democracy. Which is always very hard, because you have to now work out what the. Let me do the top question. It's easier. The question is, what are the top three surprises in terms of the data last year? What were you most surprised about? Good questions.
I guess the first thing I was surprised about was just the strength of energy demand and carbon emissions. I had not been expecting that number. When you lived through last year, I wasn't conscious of it being a particularly extreme weather event in those countries. When we first started seeing the data, we were entirely perplexed about what was going on, and spent a lot of time just trying to understand what was going on. I think number one is that. Perhaps the second biggest surprise was just the enormous growth of U.S. tight oil and shale gas production last year. To achieve the largest ever increase in one is quite an amazing feat. To achieve both at a time when, in terms of oil, there was already increasing bottlenecks, I think was a very significant surprise.
That would perhaps be the second thing I would point out. I think perhaps if I had to do a third, it would be this interaction of the power markets with renewables, where it was the first time I just sat down and done that arithmetic about just how hard would it be to decarbonize the power sector relying solely on renewables. I was quite struck by just the enormity of that challenge. Just in that sense, it was less the data, but the data provoked a question, and I was surprised by the answer to that question. Perhaps let's do one more from online, and it's a topical one, which is sort of how is the U.S. trade war going to impact the future of energy markets?
Trouble when you say these questions, you've got to try and think of the answer at the same time as hearing the question. I think there's two things going on with the trade war. First of all, we can see that the economic momentum in terms of just the near-term economy has slowed quite materially this year. People like the IMF are estimating economic growth are around half percentage points weaker this year than we saw a year ago. That will feed back on both oil demand and gas demand. How much that will come through, I think will be a function in part about what sort of policy responses we see, both in terms of monetary policy in the West, some of the fiscal expansions we're seeing in China.
Just the extent of economic slowing this year and into next year could have a significant impact on oil demand and gas demand. One of the features we've seen of oil demand over pretty much, well, every year since the price collapse in 2014, is oil demand has grown significantly above its trend rate. We've seen growth rates well in excess of 1 million barrels a day growth. If we see a very sharp slowing in economic growth, that could then become more challenging. Over the longer run, in terms of these trade wars, in the Energy Outlook, we sort of posed this question about what impact could a world of greater trade wars have on the global energy system if they were to carry on persisting and escalating over time.
This wasn't a comment on any particular trade war, but just a world in which there was greater protectionism and greater trade war. The punchline from that impact was, in that world, people may worry more about their energy security. Some of those very large oil energy importing countries, China and India, which at the moment import very large amounts of oil and gas, may at the margin say, "Actually, I may be less keen to import oil and gas for energy security reasons, and I may prefer to substitute into other domestic forms of energy. For those countries, a combination of renewables and coal. What we found is for relatively small assumptions or relatively reasonable assumptions, the impact that could have on the global oil and gas demand are really quite significant.
I think there's a near-term impact from these trade wars associated with economic momentum, but the longer-term impact is it could reduce the demand for traded oil and gas as countries prefer to use their domestic energy in a world where their energy security concerns become bigger. We should go to this side of the room. Lord Howe has his arm up and this gentleman here, we have two, one after the other.
This is all fascinating and thank you. I'm just wondering whether we're getting China right. You said it's because there's a great resurgence or appears to be, in the big power using steel and cement and so on, there's been a surge upwards. The general view is, we've been told that China is growing again, because of increased consumer production and also home production rather than imports. I read somewhere that the Chinese economy is now 10 times greater than it was in 1990. 10 times. That means 1% growth today is equivalent of 10% back then. Are we really getting the China scene right? Are you really telling us that it's consumer plus a gigantic increase in infrastructure and capital investment as well? If it is, then all bets are really off about the impact on carbon emissions.
It's going to continue to be absolutely enormous and dominating over everything else.
I think I'm somewhere in the middle of that. What I was saying was, do I expect to go back to that pattern of growth, which we saw during the massive industrialization phase, where those sectors were growing at 10, 12, 14% a year and were driving economic growth, driving energy growth, and driving coal consumption in particular? No, I think we have seen a structural rebalancing away from that industrial-led growth towards more consumer and service-facing growth. Do I think you can carry on growing an economy the size of China at 5 or 6% a year with output of iron, steel, and cement falling in outright terms? No, I didn't think that was very likely either. I always expected some sort of cyclical bounce back, and I think what we've seen is some degree of a cyclical bounce back in those sectors.
I don't expect to go back to where we were, I think that that particularly weak growth we saw in China's energy consumption in 2014, 2015, and 2016, which then fed through into very weak growth in global energy demand, and these three years when global carbon emissions were essentially flat for three years, I don't think all of that was structural. Because that coincided with a period when those three sectors' output was falling in outright terms. I don't think that's likely to persist going forward as well.
I think there was a structural element to it, which will persist and will carry on persisting, but I think there was a cyclical element which has unwound over the last couple of years, and that's one of the reasons why we've seen energy growth pick up from that weakness in 2014, 2015, and 2016 to back up to where we are now. The gentleman right in front of you.
Thank you very much. It's Patrick Heron, retired price reporter, I think you would call it, or benchmark publisher. I may have got this wrong, but I think on one of your oil slides, you showed a substantial increase last year in aviation or jet fuel consumption. As jet fuel aviation generally is a segment which perhaps not too large, but very difficult to substitute for the current fuel. How do you see that going over the next few years?
Yes, in terms of the 2018 data, we saw quite a significant rise in jet fuel. In our BP Energy Outlook, we see the growth in aviation demand as being one of the most significant source of the growth for oil demand. Partly because that sector efficiency trends are less significant in that sector and the ability to substitute other forms of energy, be it electricity or gas into that sector is also less. We see that as a sort of significant part of oil demand. Then as we go to a net zero carbon world, that suggests to me that aviation is sort of the poster child for why we will need significant increase in biofuels of one form or another, because we're not going to be able to electrify planes. I'm going to get in a plane to go to the U.S. tonight.
We're not going to have electrified planes crossing transatlantic anytime soon. My health and safety experts suggest we're not going to have hydrogen put onto planes. We will need increased use of biofuels of one form or another in aviation. That seems to be absolutely critical. I quite like biofuels because it's one sort of sector, it may not be the only one, but it's that one sector you're pretty confident that if you can produce more at an economic rate, the world will demand more of it. You do not have to market economic biofuels. The world will demand it if only you can produce them. This is purely a supply side issue. The more economic biofuels of one form or another the world can produce, the more the demand will be there.
I think aviation is a sort of key part of that story. Oh dear. Okay. How am I doing here? We have a couple more over here, and then you guys have got to wave your hands more frantically on the other side. We're doing okay. Okay. Yeah. There's a couple questions here. This gentleman at the front and then
Thank you. Just a quick one, and by the way, it was really nice hearing you in person. I've always watched the recordings online.
Those of you watching at home, in our offices, you should come down to St. James and
It is great. I'm just wondering. There are lots of talk about decarbonizing heat and also about having more electric vehicles. Clearly, these two policies will lead to an increase in energy demand. You have said that energy demand has accounted for the greatest increases in carbon emissions. Then you ended your presentation by pretty much making the case that perhaps you could transition the coal power plants from using coal to probably using gas. Do you think it might be a better policy to maybe look at these coal power plants and say transition to gas or even biomass, like what they're doing here in the U.K. at the Drax power plant? Do you think that it's still a good idea to continue to decarbonize heat as well as talk about electric vehicles? Is this a case where you have worst before better?
The point, I guess two points. One is, if we are going to move to net zero carbon world, we will need to decarbonize the transportation system. We would need to decarbonize heat. Eventually, electric vehicles or one form or another, either direct electric battery vehicles or hydrogen cars with an electric powertrain, will have to be a central part of the transportation system. My point here, in terms of switching from internal combustion engine car to an electric car in itself doesn't increase the demand for energy, just switches from one form of energy to another form of energy. In many respects, that electric car is more efficient than a comparable internal combustion engine car. We'll have savings in terms of the energy it's using.
My only point here is, if all we do is electrification without, at the same time, focusing hard on decarbonizing the power sector, then we're not achieving very much. What I put out is a bit of a repeat in terms of the chart I showed last year, and I showed again this year. That power sector, the most important sector we need to decarbonize. The fuel mix over the last 20 years is three flat lines. The share of coal in the global power sector in 2018 was the same as it was 20 years ago. The share of non-fossil fuels in 2018 was the same as it was 20 years ago. We've got electrification essentially without any decarbonization. What we need is to focus heavily.
Let's carry on thinking about electrifying the things we can electrify, but let's really think hard about trying to improve and decarbonize the power sector, because that's where lots of the lowest hanging fruit lie in terms of reducing carbon emissions, but it's where we don't seem to get much action. If you go back and you think about the newspaper articles you've read over the last year, how many newspaper articles you've read about the importance of EVs, and how many newspaper articles you've read about the importance of decarbonizing the power sector? You may be struggling on the second one, and that's the point. It matters a lot, but it does not get in the policy attention or focus that I think it deserves. Martin is behind.
Yeah. Hi, hello. It's Martin Russell, Morgan Stanley. You showed a slide about oil demand growth, which showed 2018 growing not too far from what it was globally in the previous years. It did show that pretty much all of the growth was China, India, and the U.S., and it implied quite a sharp deceleration in pretty much all of the rest of the world. That stood out to me. I was wondering if you had any views on what was driving that and also combined with that, if you have any views on how oil demand is progressing so far this year.
Yeah, the first one's a good question. I had exactly the same question about three days ago. I was looking at that chart thinking, if you remember, the way the chart works is this sort of other bit, and that other bit was growing quite significantly and then just declined to nothing. Arithmetically, there's a couple of things going on there. One, essentially, oil demand growth in the OECD, outside of America, had been growing strongly in the last couple of years, boosted by low oil prices. That impact from low oil prices is gradually there we are. Thank you. Thank you very much, whoever's on the box. The point that Martin was making is, what's happening to these orange bars here? Why are they going away? Partly it's OECD outside of America, had been being boosted by low oil prices. That effect's gradually waning.
We saw about a 30% increase in oil prices in 2018 average relative to 2017. That boost. Essentially outside of the U.S., the trend decline in OECD oil consumption sort of continued or sort of resumed in those countries. In addition, there's a whole series of countries in Latin America and the Middle East where there's sort of idiosyncratic effects on each of those countries where it had been being boosted in 2016 and 2017, and it fell away in 2018.
There's no sort of global narrative, but there's a whole set of countries which you can look at and say, "There's that one in Latin America, that one in the Middle East, which were boosting it, and then they just fell away." Partly it's a price story, and partly it's a sort of set of sort of idiosyncratic countries in those parts of the world. I have one more here, and then we'll go to the left-hand side, and I'll try and pick some in the middle as well.
Ben Coombs, PwC. I just got a question. I like your title about an unsustainable path. I just wonder, lots of people in this room have been working in this area for a number of years and some decades. Yet the unsustainable path and the sustainable path are getting wider. We always say, "Oh, well, governments can't bridge it. Individual companies can't bridge it. We all need to come together." The public seem to be getting more and more active in this space. What are your concerns around how that might play out if we aren't able to get onto a sustainable path in the next two, three, five years, and what that might mean in terms of a disconnect with the public and consumers in the U.K. and beyond?
Okay. Let me just give a personal response to that rather than a BP response to that. My personal response is, I understand and completely empathize with what they're feeling, with their impatience. It's clear that we need to get on a path in terms of an accelerated transition, something aligned to Paris. It's clear that the sum of the policies currently being introduced by governments, and we've pledged in Paris, don't add up to anything like that. That frustration is rising, and we share that frustration, as Bob said. Helge Lund in the Financial Times, our chairman, wrote a very clear op-ed where he said an accelerated transition path to Paris is not only the best thing for society, it's the best thing for companies like us. We share that frustration. I guess I have two things.
Partly, when you see some of these demonstrations, there's a sort of important but obvious point. One point being, if those demonstrations start to threaten health and safety and the ability of people to carry on their normal lives, then I worry about it in terms of health and safety. The second point, which is a sort of deeper point and a greater form of frustration, is if we are going to solve this problem, we do not need people to be trying to polarize views. We need to be coming together, like you said. I've just spent 10 or 15 minutes trying to explain why relying on renewables on its own won't work, just as a matter of arithmetic. We will need many technologies, many fuels for many years to make this work. If we have demonstrators trying to sort of say, "You're good, I'm bad.
You're right, I'm wrong," the world's more complicated than that. I share the frustration, and the only thing I worry about is sometimes it's a little bit simplistic about who's good and who's bad, because I think if we are actually going to seriously solve this, then we will all need to come together, and we will need a whole range of fuels and technologies to solve this. Perhaps their arithmetic is different to mine, but my arithmetic is pretty clear. It's published today, and they can explain to me why my arithmetic is wrong. We'll go to this side.
David Fyfe, Argus Media. Thank you very much, Spencer, and the whole BP economics team. The Statistical Review is a great resource for the whole market. I just wonder, with the discussion we've just had, where we seem to be very much focusing on policy prescriptions that deal with decarbonizing the fuel mix much more than decarbonizing the impact of growing energy use. Do you think there are policy levers that could try and reflect that imbalance a bit better going forward?
Trying to respond to the manifestations of climate change. I'm now out of my depth, David. In a sense, I think this strays into science and technology far more than economics. I am slightly nervous about those types of technologies, whether they are sustainable. They may well provide us some time while we address the underlying issue. What I would be nervous about if people think of them as a substitute for addressing the underlying issues. I know there are others in BP who are far more expert than I. Sort of always slightly worry when people push this too much, because in some sense, the underlying problems are there, and there's a limit to which you can just try and sort of mitigate the problems of it without solving the underlying issue.
I guess the subtext really is, what role does carbon pricing play?
Oh, I see. Oh, sorry, I missed it. Oh, sorry, I misunderstood the question. In terms of how to solve this problem, I remain certain that carbon pricing has to play an absolute central role in dealing with this issue. We know from Economics 101, we have hundreds of years of analysis of this. If you don't like something, you want to ration something, you put a price on it. We know that's how it works. The reason why that works is because it provides incentives for everybody, producers, consumers, innovators, entrepreneurs, to find the best efficient way of reducing carbon emissions. They can keep on adapting and changing over time in a way that governments, when they do regulation, in essence, have to pick winners and losers. That's what regulation is.
I used to be a policymaker for five or six years, picking one interest rate for one month at a time. I'm sure I got that wrong many, many times. The idea of picking regulation for the global energy system for the next 20 years is frightening. Put a price on it and let the world find a way of doing it, rather than let governments pick winners and losers. Let me pick a couple of questions from online. The first one, the question here, I'm not sure what the answer is, but what would be your best guess for the trend line in 2019, and what will we be looking back to in 12 months' time? I guess the question there is, what's the big uncertainties we face at the moment? Let's put weather to one side, as we've seen the impact of weather.
I think perhaps two or three big uncertainties. One is, particularly in the oil markets, is we have very significant supply uncertainties in terms of oil markets. How big will the falls and disruptions be in Venezuela? How successful will the U.S. administration be in reducing Iranian oil exports even more? On the other side, just how rapid a growth will we see in U.S. tight oil this year as some of the constraints on pipelines and takeaways are reduced. Sort of very significant uncertainties, I think, on oil supply. I think a second big uncertainty is what will happen to the global economy. We have seen a material slowing in the momentum of economic growth. At the moment, that's showing up more in financial market indicators and sentiment indicators like consumer confidence.
It's a bit too early yet to see it showing up in the actual real hard data. How much of it will come through? How much of that will be offset by government policies, be it monetary policy and fiscal policy? I think will have a big bearing on just the economic environment. I think the third uncertainty is just how will society be changing and viewing energy. It feels to me this is changing, in my relatively short time in energy over the last four or five years, each year it feels like the tone in which these conversations are happening are changing really in quite discrete, significant periods. How will this year pan out in terms of how society is thinking about energy and in terms of the environment we'll be looking in?
I think they are the three things I'll be looking at going forward. One other question from here and then we go across, and I'm looking at my control room. I think in about two more minutes, for those of you online who haven't finished your survey, you have to quickly put in your survey responses because my people who are adding them up are very slow. They have a sort of abacus to add up all the answers. They said they need at least 10 minutes. I don't know why, but they need at least 10 minutes. In about one minute then the survey will be taken down so they can get their abacuses out and start adding everything up. The question here is, what impact did electric vehicles have on global oil consumption last year?
Electric cars grew by about 2 million cars last year. The extent those 2 million cars, instead of increasing to overall cars, that was 2 million less internal combustion engine cars. The impact that would have in terms of oil consumption is trivially small. It's sort of 0.1 or 0.2 million barrels a day. Just to put this into perspective, it's not trying to be critical of electric cars, it's just trying to put the scale of this. 2 million electric cars grown last year. That increase in overall carbon emissions equivalent to increase in the number of cars on the planet by 400 million passenger cars.
Just gives you a sense about, yes, electric cars will be important, but they're not going to change the world on their own, and we need to have far bigger. This is the whole point again, we need to think of this as a holistic thing. 400 million cars is roughly equivalent to what that increase in carbon emissions of 0.6 gigatons. Just the scale of the problem that we saw in terms of carbon emissions last year. I think the survey now is officially closed for those online. Thank you very much. We'll come back to the answers in a minute. I think it was this side of the room. I saw some hands up in this side of the room. We have one in the middle. We have a question right at the back.
Jonathan Oxley from the World Energy Council. Thanks once again to the team. It's great to see BP continuing to contribute proactively to this discussion and debate. Spencer, you mentioned several times about the prospects or need for carbon pricing and the economics that need to go with that. What fundamental levers do we need to see change in order to make this happen? We've seen examples of that in Canada, where from province to province you see different takes on this, and it's running into trouble. We've seen the gilets jaunes in Paris. How are we going to move beyond this question of implementing carbon pricing and seeing it take place?
There's a sort of technocratic answer to this and a political answer to this. The technocratic answer to this is political will. It is not difficult to do this. If there's a political will to introduce carbon pricing, be it via carbon tax or be it via cap and trade, these are soluble problems. Relative to the difficult economic problems in terms of economic policy design, designing a carbon price system that works isn't that difficult. If you want as a first step, look at what's happened here in the U.K. In the U.K., we just introduced a carbon price floor about four or five years ago. The impact that has had in terms of reducing coal within the U.K. power sector has been very significant indeed. I think technically it is not difficult to do.
The subtext of your point was a broader one, which was: How do I win political will to achieve this? To be clear, remember, you have to impose a carbon price whatever way you do it. Either you do it explicitly via carbon price, or you do it via regulation, and there's an implied carbon price, but there's a price here. I think this is, if you like, how do I square these two things? We have people campaigning saying we need a transition, we need to do more on climate change. You have to create the political will that people are willing to accept that and trying to do that.
I think perhaps the way to do that is trying to combine these two forces to recognize to people that when we're doing this, we are responding to this sort of challenge that they face both for their generation, but even more importantly, for their children and their children's children. That's a hard thing for a politician to do, particularly a politician which is elected on a five-year cycle. It's just a tough thing to do. In some sense, we need to take the energy, which we were talking about earlier, about people demanding urgent action and connecting it with what these policies actually look like. I'm glad I'm not a government official anymore, because I think that's a difficult task to do. We have a question at the end.
Hi, Harry Spencer from Medano here. You talked a lot about the importance of decarbonizing the power sector. I just wanted to ask, were there any notable trends in terms of nuclear power generation?
Yes. Nuclear power grew by about 2.5% last year. A little bit less than overall energy demand, but pretty strong. 75% of that was in China. We saw an increase of 60 terawatt hours, 45 of which came from China. In terms of capacity, we saw about 11 or 12 gigawatts of new nuclear power capacity introduced last year. Nine came from China. In some sense, there was other sort of encouraging bits in nuclear in the sense we saw another number of nuclear reactors, which had been offline in Japan, come back online. They're up to about 10 of those nuclear reactors now in Japan are back online. The big growth story here, it is China.
I think there's sort of, if you like, looking forward, there's two big forces going on, and it's unclear which of the two forces are going to dominate. One force is that in the conventional world, in the sort of large conventional markets, the U.S., parts of Europe, there, the fleet is aging, so the average nuclear power plant age, it varies across countries are sort of somewhere between sort of low to late 30 years old. Some in sort of mid to late 30 years old. Nuclear power plants can be start commissioning, it varies across between 40, 50, 60 years. Over the next 10 years, you're going to see increasing number of those power plants decommissioned. You're not seeing investment going into that nuclear sector.
You'll see nuclear capacity and generation in most of the developing markets, my hunch, will be starting to decline. On the other side, you see this extraordinary growth happening in China. I think the interesting question is, does that just stay in China? Or if that rapid growth, is that generating learning curve, so you can start moving down the learning curve again? Some of those improving economics of nuclear spills over into some other parts of the world, be it Africa, or the Middle East, and that sort of means you've got an upside part to nuclear.
I always say when we do the Energy Outlook, and my team are going to really cross with me now, but I always say when we do the Energy Outlook, one of the benefits of doing the Energy Outlook is you sort of just crowdsource which bits did we get right, which bits we get wrong, keep challenging ourselves. We published the Energy Outlook in February of this year. One of the questions we were most pushed on, I've been most pushed on over the last sort of three or four months is, what role could nuclear play in terms of a rapid transition? Why aren't you thinking harder about the role of nuclear in a rapid transition?
That's certainly one thing I think my learning is I want to go back and challenge myself and the team, sorry guys, a little bit harder on that to say whether we should be pushing harder on that. I know, for example, if you look at the IEA, they have a stronger profile for nuclear than we do, and I think that's a sort of good challenge. I think it's sort of at the moment, there's these two big forces going on. I'm not quite sure which ones are going to dominate going forward. I think I have time for one more question in the room, and then we're going to throw up the-- How's the abacus going? We got a tick for the abacus? We've got a tick for the abacus.
One more in the room, then this lady here. Then we'll put up the answers.
Good afternoon. Thank you. Amy Bowe from McKinsey. I'm going to ask a question that is a bit of an extension on one that a gentleman asked earlier. Maybe it's a bit too soon to have an answer to this. In the same vein as the question that you've asked the audience online, to what extent do the increasing trends in energy demand and carbon emissions that we've seen over the last couple years influence or perhaps change BP's thinking about your own climate change strategy balanced against the need to meet this increasing energy demand?
I think that's a very good question. In terms of just how recent developments have shifted us, I mean, I think couple of points, I guess. We saw three years of really exceptional progress, what looked like exceptional progress in 2014, 2015, and 2016, where energy demand growth was very weak and carbon emissions was essentially flat for those three years. It felt like, wow, perhaps this is a point of inflection. Perhaps we really are making progress. There's improvements we're making. A lot of this was energy intensity. GDP was still growing, but the amount of energy needed was slowing far more quickly than energy growth was. Strong growth in renewables. Coal was falling for three years in succession. It felt perhaps we're at a point of inflection.
At the time, and I can sort of say this, it sounds a bit per se, but I did say at the time, some of this is not going to stick. Some of this relates to those sectors falling in outright terms. I don't think it's going to stick. I think what we've seen over the last couple of years has reinforced that some of it didn't stick. Perhaps some of that progress we thought we were making isn't coming through. Isn't going to stick. This challenge is very significant. I think over the last three or four years, we've continuously been surprised by just how the pace of progress on renewables. We've continued to be surprised about how quickly costs have fallen and how quickly renewable generation increased. That's a good sign.
As I was saying earlier, I don't think that fundamentally shifts this thing because renewable energy is absolutely essential to this process, but it's not going to be sufficient. We're going to need other fuels, and I don't think that's changed that component to it. I think recognize that important role of that, but recognizing that we will need other things. coal to gas switching, carbon capture, use and storage, pushing harder and harder on energy efficiency, recognizing the growing use of hydrogen. There's a whole series of different technologies I think are becoming somewhat clearer as we go through. Okay, I'm conscious of time. This will be interesting. Okay. To remind you, we had three questions. This was sort of a little bit of fun, but it was interesting to see what everybody had been waiting silently at home.
I think we got nearly about 1,500 answers to the questions. That's pretty cool, in terms of people hanging on and actually moving away and answering. The first question was, prior to the question, I was concerned about climate change, and not surprisingly, or perhaps surprisingly or not surprisingly, but 80% odd, 85% were either agreed or strongly agreed that they were worried about climate change. Were they also similarly worried about their ability to hit the Paris climate goals? Less so, which I think is good, because we were debating these in the office, and the point would be, I'm worried about climate change, but I'm also confident that governments will do something about it. Now, they're not that confident that governments are going to do something about it, but they're a little bit confident about something to do with it.
The share has declined a little bit from that 85% number to about 80% number. We've got a little shift in the distribution upwards. Okay, and what impact did today's presentation have? Oh, not much. Okay. Is that a compliment or what? I don't know. I don't know how to take that. Pretty much they're unchanged in terms of that. Thank you very much. Well, thank you everyone for sparing the time for today's presentation. Those of you here in London, but also for those of you who have stayed on via the web. It's a hard thing to do to listen for an hour and a half. Thank you very much and for taking part in the poll. Thank you very much, everyone. Thank you.