Energy Transition Supply-Demand Models

Energy self-sufficiency by country and over time?

The percentage of energy provided by countries themselves and by imports for different countries.

This data-file tabulates energy self-sufficiency by country, over time, across 30 of the largest economies in the world. Among this sample, the median country generates 70% of its energy domestically, and is reliant on imports for 30% of the remainder. Energy self-sufficiency varies vastly by country.

Energy self-sufficiency can be estimated by on a gross basis by dividing total primary energy production by total primary energy consumption. Leading countries include Norway (6x more energy produced than consumed), Qatar (4x), Iraq (3x), Australia (3x), Saudi Arabia (2x), Indonesia (2x), Russia (1.7x), Canada (1.7x).

Energy self-sufficiency can also be estimated on a net basis, by summing up the share of total useful energy consumption, by resource, that is met domestically versus from import reliance. Most import-reliant countries include Singapore (99%), Japan (87%), South Korea (85%), Italy (80%), Spain (70%) and Turkey (65%).

Energy self-sufficiency in the US declined from 1982 to 2005, at which point, the US was importing over 30% of its total energy needs. Amazingly, in the past twenty years since 2005, the US has ramped up its oil production by 3x, its gas production by 2x and now also produces 5% of its energy from new energies such as wind and solar. This means the US produces 1.1x more primary energy than it needs. And, for the first time in 50-years, the US has no net energy import reliance (chart below).

Percentage of energy supplies in the US provided by imports and by self-supplies. Thanks to shale oil and gas, the US has again become self-sufficient.

Some commentators have noted a streak towards isolationism among recent US political candidates. The reality is that the US is less reliant on a stable world for its energy security than it was 20-30 years ago.

Energy self-sufficiency in Europe, on the other hand, has been a disaster, and dangerous for geopolitical security. Europe was 65% self-sufficient for its own energy supplies, 30-years ago, in 1994. Today it is 45% self-sufficient in its energy supplies (chart below).

Wind and solar do provide 7% and 3% of Europe’s energy needs, respectively. However, domestic production of gas, oil and coal have all fallen by 60-70% in the past 30-years, and countries such as Germany have enacted kamikaze energy policies such as shutting down their nuclear industries (which formerly supplied 9% of Germany’s energy needs).

Percentage of energy supplies in Europe provided by imports and by self-supplies. Due to falling fossil fuels production and other policy decisions, Europe now imports more than half of its energy needs.

Energy self-sufficiency in China is a major policy goal. Import reliance peaked in 2006, at 20% of China’s energy needs. China’s need to import oil is the largest import reliance in our entire data-set. Hence China has been electrifying its vehicle fleet faster than any other nation, while also ramping up coal production and renewables for electricity.

Percentage of energy supplies in China provided by imports and by self-supplies. China relies heavily on imported oil.

The full data-file shows our calculations, for transparency, with data into the energy self-sufficiency by country in Argentina, Australia, Brazil, Canada, China, Colombia, the EU, France, Germany, India, Indonesia, Iran, Iraq, Israel, Italy, Japan, Malaysia, Mexico, Norway, Poland, Qatar, Russia, Saudi Arabia, Singapore, South Korea, Spain, Thailand, Turkey, UAE, the United Kingdom, the US, and Vietnam. Underlying data are from the Energy Institute.

Global energy: supply-demand model?

This global energy supply-demand model combines our supply outlooks for coal, oil, gas, LNG, wind and solar, nuclear and hydro, into a build-up of useful global energy balances in 2023-30. We fear chronic under-supply if the world decarbonizes, rising to 5% shortages in 2030. Another scenario is that emerging world countries bridge the gap by ramping coal. Numbers can be stress-tested in the model.

Useful global energy demand grew at a CAGR of +2.5% per year since 1990, and +3.0% per year since 2000. Demand would ‘want’ to grow by +2% per year through 2030, due to rising populations and rising living standards (model here). We have pencilled in +1.75% pa growth to this model to be conservative.

Combustion energy is seen flat-lining in our net zero scenario. This includes global coal use peaking at 8.4GTpa in 2024 then gently easing to 2010 levels by 2030 (model here). It includes oil demand, rising to 102Mbpd in 2024 (data here), then plateauing (model here) as OPEC and US shale (model here) offset the decline rate impacts of conventional under-investment. It includes risked LNG supplies rising +70% from 400MTpa in 2022 to almost 700MTpa by 2030 (model here). While our roadmap to net zero would need to see global gas growing at +2.5% per year through 2050 (model here), this data-file has pencilled in flat production in 2022->30, as we think that latter scenario currently looks more likely to transpire.

Renewable energy is exploding. Our model of wind and solar capacity additions is linked here and discussed here. In our roadmap to net zero, solar more than doubles from c220GW of new adds in 2022 to 500GW by 2030, while wind rises from c100GW of new adds in 2022 to 150GW by 2030.

Other variables in the model include rising energy efficiency (note here), the need for a nuclear renaissance (note here) and other variables that can be flexed.

What is wrong with this balance is that it does not balance. The assumptions pencilled into the model see an under-supply of global energy of about 3% in 2025, rising to 5% in 2030. I.e., by 2030, the world will be “half a Europe” short of energy. The first law of thermodynamics dictates that energy demand cannot exceed supplies. So what would it take to restore the balance? Well, pick your poison…

(1) Slower demand growth could re-balance the model. Very high energy prices might mute demand growth to only +1.25% per year, although this would be the slowest pace of demand growth since the Great Depression, lower even than during the oil shocks (useful data here). Unfortunately, our view is that pricing people out of the global energy system in this way is in itself an ESG catastrophe.

(2) Ramping renewables faster could re-balance the model, although it would require an average of 1 TW pa of wind and solar capacity additions each year in 2024-30, and over 2 TW pa of wind and solar additions by 2030 itself, which is 3x higher than in our roadmap to net zero (discussion here). For perspective, this +2TWpa solution requires primary energy investment to quadruple from $1trn pa to at least $4trn pa, which all needs to be financed in a world of rising rates. It means that global wind and solar projects will consume over 200MTpa of steel, which is 2x total US steel production, and yet steel would not even qualify as a ‘top ten’ bottleneck, in our wind bill of materials or solar bill of materials. This scenario also requires a 3x faster expansion of power grids and power electronics than our base case estimates (see the links). Is any of this remotely possible?

(3) Continue ramping coal? The main source of global energy demand growth is the emerging world. The emerging world is more likely to favor cheap, dirty coal. Or worse, deforestation for firewood. Thus another way to ameliorate under-supply in our global energy supply-demand balance is if global coal continues growing, reaching a new peak of 9GTpa in 2030. Unfortunately, this scenario also sees global CO2 hitting a new peak of 54GTpa in 2030.

(4) Pragmatic gas? Another means of re-balancing the global energy system is if global gas production rises at 2.5% per year, which is the number required, and that is possible, on the TSE roadmap to net zero (model here). This scenario does see global CO2 falling by 2030. The main problem here is that pragmatic natural gas investment has become stranded in no man’s land, within a Manichean duality of fantasies and crises.

(5) Some combination? The world is complex. It is unlikely that a single lever will be pulled to resolve under-supply in our global energy supply-demand balance. In 2023, we think economic weakness will mask energy under-supply, mute energy prices, and lure many decision makers into looking at spot pricing and thinking “everything is fine”. Please download the model to stress-test the numbers, and different re-balancing solutions…

Global plastic demand: breakdown by product, region and use?

Global plastic is estimated at 470MTpa in 2022, rising to 1,000MTpa by 2050. This data-file is a breakdown of global plastic demand, by product, by region and by end use, with historical data back to 1990 and our forecasts out to 2050. Our top conclusions for plastic in the energy transition are summarized below.

Global plastic demand is estimated at 470MTpa in 2022. For perspective, the 100Mbpd global oil market equates to around 5bn tons per year of crude oil, showing that plastics comprise almost 10% of total global oil demand.

Our outlook in the energy transition sees increasing demand for polymers, most likely to 1,000MTpa by 2050. Many polymers are crucial inputs for new energies. Others are used for high-grade insulation, both thermal and electrical. Others for light-weight composites, which lower the energy consumption of transportation technologies.

Global plastic demand by region. The top billion people in the developed world comprise 12% of the world’s population, but 40% of the world’s plastic demand. We see developed world plastics demand running sideways through 2050, while emerging world demand doubles from 300MTpa to 700MTpa (charts below).

Global plastic demand CAGR? Our numbers are not aggressive and include a continued deceleration in the total global demand CAGR, from 7.5% pa growth in the twenty years ending in 1980, to 6% pa in the twenty years ended 2000, to 3.3% pa in the twenty years ended 2022 and around 2.5% pa in the period ending 2050.

Global plastic consumption per capita. Our numbers include an enormous policy push against waste in the developed world, where consumption today is around 170 kg of plastics per person per year. Conversely, plastic demand in the emerging world is 75% lower and averages 40 kg pp pa today, seen rising to 90 kg pp pa by 2050.

Upside to plastics demand? What worries us about these numbers is that they require historical trends to slow down. Historically, each $k pp pa of GDP is associated with 4kg pp pa of plastic demand. But our numbers assume slowing demand in the developed world and slowing demand in the emerging world (chart below).

Please download the model to stress-test your own variations. Underlying inputs are drawn from technical papers, OECD databases and Plastics Europe. The forecasts are our own.

Plastic demand by end use is broken down in the chart below. 35% of today’s plastics are used as packaging materials, 17% as construction materials and c10% for textiles.

The strongest growth outlook is in electrical products (currently 7-9%) and light-weighting transportation. There are few alternatives to plastics in these applications. Conversely, for packaging materials, we see more muted growth, and more substitution towards bioplastics and cellulose-based products from companies such as Stora Enso.

Upside for plastics in the energy transition is especially important, with plastics used in important components of the energy transition, from EVA encapsulants in solar panels, to the resins in wind turbine blades, to strong and light-weight components in more efficient vehicles, to polyurethane insulating materials surrounding substantively all electric components, batteries and vehicles. Of particular note are fluorinated polymers, where the C-F bond is highly inert, and helps resist degradation in electric vehicle batteries.

Plastic demand by material is broken down in the chart below. Today’s market is c20% polypropylene, c15% LDPE, and other large categories with >5% shares include HDPE, PVC, PET, Polyurethanes and Polystyrenes. We think the strongest growth will be in more recyclable plastics, materials linked to the energy transition, and next-gen materials that underpin new technologies such as additive manufacturing.

Feedstock deflation is likely to be a defining theme in the energy transition, as many polymers draw inputs from the catalytic reforming of naphtha. Today, 70% of BTX reformate is blended into gasoline, but we see this being displaced by the rise of electric vehicles, with particular upside for the polyurethane value chain.

Decarbonizing plastic production is also a topic in our research. As a starting point, our ethane cracker model is linked here, and our economic model for converting olefins into polymers is here. CO2 intensity of different plastics from different facilities is tabulated here.

Plastic recycling. We also see potential for plastic-recycling technologies to displace 8-15Mbpd of potential oil demand growth (i.e., naphtha, LPGs and ethane) by 2060, compared to a business-as-usual scenario of demand growth. All of our plastic recycling research is linked here. After a challenging pathway to de-risking this technology, we see front-runners emerging, such as Agilyx, Alterra and Plastic Energy.

Global gas: is there enough gas for energy transition?

Global gas production is forecasted to double from 400bcfd in 2023 to 800bcfd in 2050.

Our roadmap to ‘Net Zero’ requires doubling global gas production from 400bcfd to 800bcfd, as a complement to wind, solar, nuclear and other low-carbon energy. This data-file quantifies global gas production forecasts by country, what do you have to believe about global gas reserves, and is there enough gas?

Global gas production doubled in the c30 years from 1990-2019, rising at a 2.5% CAGR, which is the same trajectory that needs to be sustained to 2050 on our long-term energy market supply-demand balances.

Amazingly, from 1990-2019, global gas reserves increased from 4,000 TCF to 7,000 TCF, for a reserve replacement ratio of 190%, although the numbers have been cyclical and have fallen below 100% in recent years (chart below).

Another fascinating feature of gas markets is their flexibility, shown by plotting monthly gas production by country over time (chart below). In the Northern Hemisphere, production runs 6% higher than the annual average in December-January and 6% lower than average in June-August, as producers consciously flex their output to meet fluctuations in demand. Gas output does not show volatility, but voluntarity!

Global gas production by month is typically 15-20bcfd higher than average in Northern Hemisphere winter months and 15-20bcfd lower in Northern Hemisphere summer months, due to variations in heating demand

On our numbers through 2050, as part of the energy transition, a reserve replacement ratio of 107% is needed, while the ‘reserve life’ (RP ratio) will likely also decline from around 50-years today to 25-years in 2050. Please download the data-file for reserve numbers and production numbers by country.

Global gas reserves and RP ratio by country, from 1980 to 2050.RP ratio is expected to decrease from roughly 40 years today to 25 years in 2050.

Onshore resource extensions are seen primarily coming from shale, with continued upside in the US, and vast new potential in the Middle East, North Africa and possibly even European shale as a way of replacing Russian gas.

Another offshore cycle is also seen to be necessary, discovering and developing an average of 45 TCF of offshore resources each year in 2023-2050. These are big numbers, equivalent to discovering a large new gas basin (e.g., an “entire Mozambique of gas”) every 3-5 years.

Global gas consumption by region and over time is also estimated in the data-file, flatlining at 150bcfd in the developed world, but rising by 2.5x in the emerging world, with the largest gains needed in India, Africa and China (chart below).

Global gas consumption by country, from 1990 to 2050. Consumption is expected to double from 400 bcfd today to 800 bcfd by 2050 due to increased consumption from emerging markets.

Global LNG demand would also need to treble to meet this ramp-up, linking to our model of global LNG supplies. Within today’s LNG market, 25% flows to Europe, 20% to Japan, and 55% to the emerging world. By 2050, the emerging world would be attracting 80% of global LNG cargoes, with the largest growth in China and India.

Global LNG imports by country, from 1990 to 2050. Imports are expected to triple from 400 MTpa in 2023 to almost 1200 MTpa by 2050. The major importers will be China, India, and other Asian countries.

Our best guesses for how a doubling of global gas production might unfold is captured in this model of global gas forecasts by country/region. On the other hand, there is no guarantee that coal-to-gas switching will occur on the needed scale for global decarbonization, especially as 2023/24 has seen emerging world countries (India, China) ramping coal instead for energy security reasons.

Global coal production: supply outlook in energy transition?

Global coal production likely hit a new all-time peak of 8.7GTpa in 2023, of which 7.5GTpa is thermal coal and 1.2GTpa is metallurgical. The largest countries are China (4.4GTpa), India (1GTpa), other Asia (0.7GTpa), Europe (0.5GTpa) and the US (0.5GTpa). This model explores what is required to meet our energy transition aspirations.

Coal can be the cheapest thermal energy source on the planet. In normal times, coal costs $60/ton (coal mining model here) and contains 6,250 kWh/ton of thermal energy, implying a cost of 1c/kWh-th.

Coal-fired power can thus cost 2-4c/kWh (model here) and an existing coal plant is cheaper than other levelized costs of power.

Coal is the highest-carbon fossil fuel, with an average CO2 intensity of 0.37 kg/kWh-th (data here), which is 2x more than gas (note here).

The CO2 disparity is amplified further when considering coal’s Scope 1+2 emissions, as often coal mining leaks more methane than gas itself. And Rankine steam cycles fueled by coal have efficiency drawbacks (note here) and also relatively low flexibility (data here).

Hence our Roadmap to Net Zero would need to see coal consumption flat-lining from 2022, then declining at 8% pa in the 2030s and 17% pa in the 2040s, to well below 500MTpa (which in turn is abated by CCS or nature-based solutions).

This is sheer fantasy, unless wind, solar and natural gas ramp up enormously, especially in China, India and other parts of the Emerging World. Coal-to-gas switching economics are profiled here.

Some encouraging precedent come from the US, where coal production peaked at 1GTpa in the 2010s, before shale gas ramped to 80bcfd. Thus US coal declined to 500MTpa in 2021. Although questions about the continued phase-back of US coal are now being raised, due to pipeline bottlenecks from the Marcellus, and energy crisis in Europe, requiring a substitution of Russian energy supplies (oil, coal and gas).

There is always a danger of drawing lines on charts, which simply reflect aspirations, blindly projected out to 2050. The real world may not follow a straight line of pragmatic progress, but instead fluctuate between fantasy and crisis.

During times of energy crisis, such as 2022, international coal prices have spiked to $340/ton. Remarkably, this took thermal coal prices above metallurgical coal, and even above oil on a per-btu basis. Western coal producers are screened here.

Metallurgical coal may be particularly challenging to substitute. We have reviewed the costs of green steel here. We have seen some interesting but smaller-scale options in bio-coke. We have been less excited by hydrogen or syngas from gasification of coal.

Around 1GTpa of new coal projects are in planning or under construction, of which half are in China. Chinese coal production is something of a ‘wildcard’, explored in our short note here, and often defying expectations to the upside, despite rising renewables (chart below). Helping China to decarbonize might require 300bcfd of gas (roadmap here).

India’s coal use has also doubled since 2007, rising at 6% pa. It remains “the engine of global coal demand”, according to the IEA, rising +70MTpa, to 1.1GTpa in 2022, 1.3GTpa in 2023 and 1.4GTpa in 2026.

Please download the data-file to stress-test assumptions around coal mine additions, decline rates, phase-downs and coal-to-gas switching.

US shale: outlook and forecasts?

What outlook for US shale in the energy transition? This model sets out our US shale production forecasts by basin. It covers the Permian, Bakken and Eagle Ford, as a function of the rig count, drilling productivity, completion rates, well productivity and type curves. US shale likely adds +1Mbpd/year of production growth from 2023-2030, albeit flatlining in 2024, then re-accelerating on higher oil prices. Our shale outlook is also summarized below.

What outlook for shale in energy transition?

Shale is a technology paradigm where well productivity has risen by 3-7x over the past decade, through ever greater digitization. Shale economics are very strong, with 20% IRRs at $50/bbl oil on shale oil (model here) or at $2.8/mcf on shale gas (model here). We think 100bn bbls of recoverable shale resources remain in the US and ultimately, liquids production could be ramped up from 10Mbpd in 2023 to 17Mbpd by 2030 (note here), and most of this will be needed as energy shortages loom.

However the US shale industry has shifted its focus towards capital discipline and ESG. US shale averages 10kg/boe on a Scope 1 upstream basis (data here), shale oil averages 25kg/boe on a full Scope 1&2 basis running up to the refinery gate (data here) and 55kg/boe on a refined basis running up to the point of combustion (data here). The spread is wide, after comparing and contrasting 425 companies here and here. The best decarbonization opportunities for shale are mitigating flaring and methane leaks followed by electrification. Ultimately, we think the best operators could reach CO2 neutrality.

The most important questions on shale are how the resource base and well productivity will trend. This has been the topic of our shale research, and our latest views are covered in our 2024 shale outlook. Historically, we have also undertaken large reviews of the pace of shale technology progress, based on technical papers (examples here and here). There are fifty variables to optimize. And we are most excited about big data techniques, fiber optics and shale-EOR.

Modelling US shale production by basin?

Our model for US shale production looks at each of the main basins, using a factor breakdown. Total production in month T1 = Total production in month T0 + new additions – base declines. To calculate new monthly additions, we multiply (a) number of rigs running (b) wells drilled per rig per month (c) wells completed per well drilled (d) initial production of newly completed wells (IP30). And to calculate the base declines, we fit a best-fit type curve onto the new additions from past months. This model has worked quite smoothly for 6-years now, including history going back to 2011 and forecasts going out through 2030.

The Permian basin is the largest US shale oil basin, with 8Mbpd of total liquids production in 2023. Over the past six years from 2017-2023, the Permian basin has seen an average of 340 rigs running, drilling an average of 1.2 wells per rig per month, completing 1.06 wells for every well drilled (DUC drawdown) at an initial production rate of 780bpd (IP30 basis), adding +850kbpd/year of new supply to global oil markets. We still see strong growth potential, and the Permian could reach 14Mbpd of total liquids production by 2030, amidst higher activity and oil prices. All of these variables can be stress-tested in the model.

The Bakken is the second largest US shale oil basin, with 1.3Mbpd of total liquids production in 2023. Over the past six years from 2017-2023, the Bakken has seen an average of 40 rigs running, drilling an average of 1.9 wells per rig per month, completing 1.15 wells for every well drilled (DUC drawdown) at an initial production rate of 780bpd (IP30 basis), adding +20kbpd/year of new supply to global oil markets. We see a decline in 2024, a recovery in 2025-26 and a plateau through 2030.

The Eagle Ford is the third largest US shale oil basin, with 1.1Mbpd of liquids production in 2023. Over the past six years from 2017-2023, the Eagle Ford has seen an average of 60 liquids-focused rigs running, drilling an average of 2.1 wells per rig per month, completing 1.22 wells for every well drilled (DUC drawdown) at an initial production rate of 680bpd (IP30 basis), but liquids production has actually declined, especially during the volatility of the COVID years. We see a decline in 2024, a recovery in 2025-26 and a plateau through 2030.

Challenges and controversies for US shale?

The main revisions to our shale production models have been because of lower activity, as capital discipline has entrenched through the shale industry. The chart below shows our forecasts for activity levels at different, prior publication dates of this model. We have compiled similar charts for all of the different variables and basins, in the ‘revisions’ tab, to show how our shale numbers have changed.

Our shale outlook for 2023-2030 sees the potential for +1Mbpd of annual production growth as the industry also generates $150-200bn per year of annual free cash flow. You can stress test input variables such as oil prices in the model.

We have also modeled the Marcellus and Haynesville shale gas plays, using the same framework, in further tabs of the data-file. Amazingly, there is potential to underpin a 100-200MTpa US LNG expansion here, with just 20-50 additional rigs. Although recently we wonder whether the US blue hydrogen boom will absorb more gas and outcompete LNG, especially as the US Gulf Coast becomes the most powerful clean industrial hub on the planet (note here).

International shale? We have found it harder to get excited about international shale, but there is strong potential in other large hydrocarbon basins, if European shale is ever considered to rescue Europe from persistent gas shortages, and less so in China.

Please download the data-file to stress-test our US shale production forecasts by basin.

Japan gas and power: supply-demand model?

Japanese grid generation and CO2 intensity from 1985 to 2030. Carbon intensity could halve by 2030.

Japan’s gas and power markets are broken down by end use, traced back to 1990, and forecast forwards to 2030 in this model. Japan’s electricity demand now grows at 0.3% pa. Ramping renewables, nuclear and gas back-ups could halve Japan’s total grid CO2 intensity to below 0.25 kg/kWh by 2030.

After declining at -0.7% per year in the past decade, electricity demand will begin growing again, at +0.3% pa through 2030, due to the rise of electric vehicles and increasing electricity demand for data-centers linked to the rise of AI.

Japanese electricity generation is projected to remain almost flat to 2030.

Japan’s electricity demand is 8.1 MWH pp pa, compared with 6 MWH pp pa in Europe and 13 MWH pp pa in the US. 25% is residential, 30% is commercial, 30% is industrial.

A detailed breakdown of Japan’s electricity demand reminds us of the ubiquity of energy. 8% is used for making steel, 7% for manufacturing machinery, 5% for chemicals. Another 3% is used in supermarkets, 3% in medical facilities, 2.5% in schools, 2% in restaurants, and 1.6% for Japan’s amazing electric railway system.

Japan’s gas market is unusual in that 65% of all the gas is used for electricity generation, much higher than in the US (38% of gas use) and Europe (25% of gas use).

Japanese gas demand from 1990 to 2030. Expecting a slight decline in 2024 but then flatlining to 2030.

The key reason is not that Japan generates a large portion of its electricity from gas (only c30% of the mix), but rather that non-electricity uses of gas are underdeveloped. For example, Japanese households still generate more heat from oil products than from gas, and 60% of Japan’s 3.4 Mbpd of oil use is outside of transportation.

Our outlook for Japan’s electricity mix is that wind and solar will double from 11% to 22% of the mix by 2030, and almost 100 TWH of nuclear restarts will increase nuclear from 8% to almost 20% of the mix.

A common finding across all of our research is that ramping renewables tends to displace coal more than gas (note below). Hence we think coal phase-downs could surprise, cutting coal use in power by c50% (1.5x more than official government targets).

Back up: does ramping renewables displace gas?

The result is a mild pullback in Japan’s gas consumption, from 6bcfd in 2023 to 5bcfd in 2030, freeing up 7.5MTpa of LNG.

However, the best case scenario for Japanese decarbonization would hold gas use constant, especially if LNG supplies are available, displace more coal, and this could halve the CO2 intensity of Japanese power generation by 2030.

All of the numbers and input assumptions can be stress-tested in the model. Underlying data are from Japan’s Agency for Natural Resources and Energy. The data-file is also interesting to compare with our European, US, China and India energy models.

Global population and GDP breakdown by country?

Global population and GDP are broken down in this data-file, across 10 key regions, with data back to 1960 and projections to 2050, as an input to all of our supply-demand models. Population rises at 0.7% pa from 8.0bn in 2023 to 9.7bn in 2050. Real global GDP rises at 2.5% from $105trn in 2023 to $200trn by 2050. Mega-trends are underway in demographics, manufacturing and defence.

Population and GDP growth underpin many of our supply-demand models out to 2050, such as our outlook for energy demand, plastics demand, steel demand, EV adoption and other commodities. Hence this data-file contains the key inputs, across the US, Canada, Europe, Japan, Australia, LatAM, China, India, other Asia and Africa.

Global population stood at 8.0bn people in 2023, having doubled since 1975, rising at +1.0% pa in the past ten years, decelerating to +0.7% pa to 2050, when world population will reach 9.7bn people.

The five wealthy regions — US, Canada, Europe, Australia and Japan — comprise 1.2bn people with an average 2023$ GDP per capita of $50k. They are 15% of the world’s people, 37.5% of its energy consumption, 55% of its GDP and 60% of its total household consumption. By 2050, they will only comprise 12% of the world’s people, 22% of its energy, and 45% of its GDP and household consumption.

A ‘just transition‘ also clearly cannot suppress income growth for the 4bn people with GDP per capita below $3k pp pa. Today’s global average is $13k pp pa. It rises to $20k pp pa by 2050.

Real global GDP is estimated at $105 trn in 2023, rising at 2.7% pa in the past decade, continuing at 2.6% pa through 2030 and then 2.4% pa from 2030-50. GDP is also decomposed by region and by category, across consumption, households, manufacturing, investment, defence. Long term trends have major implications.

Manufacturing, for example, comprises 15% of global GDP. The US and Europe declined from 75% of global manufacturing value add in 1965 to just 40% in 2023. Whereas China rose from almost nothing to one-third of global manufacturing value add (charts below).

As a methodology, our forecasts for global population and GDP are built-up across 218 countries, based on disclosures from the IMF and World Bank, then aggregated together, per the Countries tab, which shows which countries we categorize into which regions.

For auditability, all underlying input data are preserved in their original formatting. Whereas data-points added or categorized by TSE are in Georgia font and our usual blue text coloring/highlighting.

Underlying population data are sourced from the World Bank. Last downloaded in May-2024. GDP data and GDP growth are sourced from the IMF, last updated April-2024.

Internet energy consumption: data, models, forecasts?

This data-file forecasts the energy consumption of the internet, rising from 800 TWH in 2022 to 2,000 TWH in 2030 and 3,750 TWH by 2050. The main driver is the energy consumption of AI, plus blockchains, rising traffic, and offset by rising efficiency. Input assumptions to the model can be flexed. Underlying data are from technical papers.

Our best estimate is that the internet accounted for 800 TWH of global electricity in 2022, which is 2.5% of all global electricity. Despite this area being a kind of analytical minefield, we have attempted to construct a simple model for the future energy demands of the internet, which decision-makers can flex, based on data and assumptions (chart below).

Internet traffic has been rising at a CAGR of 30%, as shown by the data use of developed world households, rising to almost 3 TB per user per year by 2023. The scatter also shows a common theme in this data-file, which is that different estimates from different sources can vary widely.

Future internet traffic is likely to continue rising. By 2022 there were 5bn global internet users underpinning 4.7 Zettabytes (ZB) of internet traffic. Users will grow. Traffic per user will likely grow. We have pencilled in some estimates, but uncertainty is high.

TSE's estimates for future numbers of internet users, data traffic per user, and total data traffic.

The energy intensity of internet traffic spans across data-centers, transmission networks and local networking equipment. Again, different estimates from different technical papers can vary by an order of magnitude. But a first general rule is that the numbers have declined sharply, sometimes halving every 2-3 years.

Electricity use of data centers, data transmission, and local network systems from 2009 to 2023.

The current energy intensity of the internet is thus estimated at 140 Wh/GB in our base case, broken down in the waterfall chart below, using our findings from technical papers and the spec sheets of underlying products (e.g., offered by companies such as Dell).

Energy intensity of internet processes will almost certainly decline in the future, as traffic volumes rise. Again, we have pencilled in some estimates to our models, which can be flexed.

However the energy needed for AI is now rising exponentially. Training Chat GPT-3 in 2020 used 1.3 GWH to absorb 175bn parameters. But training chat GPT-4 in 2023 used 50 GWH to absorb 1.8trn parameters. We find a 98% correlation between AI training energy and the total compute during training.

AI querying energy is also correlated with the complexity of the AI model, and thus will likely continue rising in the future. Average energy use is estimated at 3.6 Wh per query today, which is 4x more than an email (1 Wh) and 10x more than a google search (0.3 Wh).

Muting the impacts of larger data-processing volumes, we expect a 40x increase in future computer performance in GFLOPS per Watt (chart below). This yields 900 TWH of AI demand around 2030, revised up from 500 TWH in April-2023 (chart above).

Please download the model to stress-test your own estimates for the energy intensity of the internet. It is not impossible for total electricity demand to ‘go sideways’ (i.e., it does not increase). It is also possible for the electricity demand of the internet to exceed our estimates by a factor of 2-3x if the pace of productivity improvements slows down.

Global biogas production by country?

Biogas production by country from 2000 to 2023. China has now become the worlds' largest producer of biogas, though it only covers 2% of their gas demand.

Global biogas production has risen at a 10-year CAGR of 3% to reach 4.3bcfed in 2023, equivalent to 1.1% of global gas consumption. Europe accounts for half of global biogas, helped by $4-40/mcfe subsidies. This data-file aggregates global biogas production by country, plus notes into feedstock sources, uses of biogas and biomethane.

Germany has historically been the largest producer in the world, with biogas output rising to 0.8bcfd by 2015, 10% of Germany’s total gas needs, then flat-lining on the phase-back of subsidies, such as 6-25 c/kWh feed-in tariffs for biogas->power.

40-45% of Germany’s biogas feedstock is from the anaerobic digestion of crop residues (70% corn silage), 40-45% is from animal waste (80% cattle), 6% is from wastewater. 85% is produced as biogas and 15% is upgraded to biomethane. 78% is used to produce electricity. Larger listed companies include EnviTec and Verbio.

China has now overtaken Germany to become the world’s largest biogas producer, reaching 0.9bcfed in 2023, although biogas has fallen from 4% of China’s total gas use in 2013 to 2% in 2023.

The US produced 0.6bcfd of biogas in 2023, or 1% of total gas consumption, with 2,400 production sites, of which 70-80% is captured from landfills. BP acquired the US’s largest RNG producer, Archaea Energy, for $4.2bn in 2022.

Brazil arguably has most growth potential, producing 0.1bcfed, across around 1,000 production sites, 65% from agricultural wastes, and c80% is used for electricity generation.

Denmark sources the highest share of its total gas needs from biogas of any country in our database by a wide margin, at c50%. 80% is upgraded and delivered into the gas grid, encouraged by a $6.2/mcfe subsidy program for raw biogas production, and $13/mcfe for upgraded biomethane, which supports the economics in our biogas costs models.

The data-file contains underlying data into global biogas production by country, in TJ terms, in TWH terms, and in bcf of gas equivalent terms (bcfed). Backup tabs contain workings and other input data. For further data, please see our broader biogas research and biofuels research,

What outlook for shale in energy transition?

Modelling US shale production by basin?

Challenges and controversies for US shale?

Cookies?