One of the features of an upcoming COP is that the months prior often bring with them a slew of proposals for new global targets and initiatives. This year is no different, perhaps also driven by the extreme heat and precipitation that has been challenging the Northern Hemisphere. In July, COP 28 President, Dr. Sultan al-Jaber, put forward a proposed COP28 action plan in a letter to the Parties. Along with a number of calls for increasingly rapid transformation and improved financing for developing countries, the plan includes the following three specific energy system targets which were also widely reported on in the media:
- Reach a global tripling of renewables capacity by 2030.
- Doubling of the average global rate of energy efficiency improvements by 2030.
- A dramatic scale up of new low-carbon hydrogen production and decarbonization of existing hydrogen production to reach an overall doubling of hydrogen production.
The plan also includes emission reduction targets for the oil and gas industry to achieve by 2030 and includes a call to transform heavy-emitting sectors, including scaling up use of low-carbon hydrogen, carbon capture and storage, and carbon dioxide removal, aligned with science. However, no specific targets are included for these heavy-emitting sectors. This all represents quite a formidable ambition, but to put these three energy system targets into context the recent Shell Energy Security Scenarios provide a useful backdrop.
The scenarios comprise two different stories for the decades ahead, but both are built from the security-focused world in which we currently find ourselves. In Sky 2050, the troubles of the 2020s give way to intense global competition in a race to gain market share in the delivery of clean energy systems. Mutual interest prevails and net-zero emissions is achieved in 2050. But in the second scenario, Archipelagos, distrust, self-interest and security concerns prevail. Although a rapid transition takes place and emissions fall throughout the century, the net-zero goal isn’t realised until early in the 22nd century. Nevertheless, warming is limited to 2.2°C above the 1850-1900 baseline period.
The Energy Security Scenarios data shows global renewable capacity at around 3,500 GW in 2023, implying a 2030 goal of around 10,000 GW. The 2023 capacity is comprised of 1,200 GW hydroelectricity, 1,200 GW solar PV (commercial at 800 GW and rooftop at 400 GW) and 1,100 GW wind power. There’s also 100 GW of commercial biomass burning, but not everyone considers this as renewable energy. While hydro and biomass show only modest increases over the remainder of this decade, wind and solar increase substantially in both scenarios. In 2023, solar PV looks to be increasing by about 300 GW and wind by 100 GW, although some of this may be related to COVID delayed projects now coming online. With seven full years before the end of 2030, delivery at these scales could mean some 3,300 GW of solar in total and 1800 GW of wind in 2030, but this is well below the goal of some 10,000 GW even with hydro and biomass included. The deployment rate needs to substantially increase and that is exactly what is happening in Sky 2050.
In Sky 2050 and Archipelagos, the growth to 2035 is shown below.
In Sky 2050 the goal is reached in 2030 but, in Archipelagos, it’s not until about 2035. This isn’t about different levels of ambition across the scenarios, but instead represents the very real headwinds that rapid deployment could face, particularly in the circumstances of Archipelagos. Trade disruption is a major outcome of the story, yet there is an important dependency on trade within the solar PV story. The minerals required, the manufacture of the panels, the grid expansion (including across borders) and even the labour required for installation all depend on good trading relationships and the free movement of people for work. For a goal of 10,000+ GW, which might include 5,000 GW of solar PV, the current global manufacturing capacity and installation capability of around 300+ GW per year would have to at least triple to around 1000 GW per year, which means adding 100 GW of new production and installation capability each and every year between now and 2030. The project pipeline for new manufacturing facilities looks robust, with the limitation on deployment in many locations being grid connectivity.
The second goal is the desire to double the average global rate of energy efficiency improvements across sectors. This is quite hard to unpick without further details, but it certainly doesn’t mean ‘doubling energy efficiency’ as reported by some media outlets. The goal could be in relation to the energy intensity of the global economy, which would be measured in energy use per unit of GDP, or on the provision of energy services, i.e. a true efficiency measure, which could be in units such as tonne-km/MJ for road freight. For this discussion I will focus on the former.
In 2022 global energy use was about 4.6 GJ per US$1000 (2016) GDP. This had improved from 5.64 GJ in 2010, meaning an improvement of some 20% or about 1.5% per year. Presumably the ambition means shifting this to 3% per annum. If this were the case then energy use would fall to around 3.7 GJ per USS$1000 by 2030. Both Sky 2050 and Archipelagos are within this range, sitting respectively 0.1 GJ either side in 2030. Apart from general societal improvements in energy efficiency, which have been ongoing for decades (9.5 GJ per US$1000 (2016) GDP back in 1960), the main driver of change in this decade will be electrification of energy services. For example, using gasoline to power a car has a well-to-wheel efficiency of about 25%, but using electricity derived from solar or wind has a turbine-(or panel)-to-wheel efficiency of around 70% or more (losses in transmission and storage of electricity combined with the efficiency of the vehicle itself). This COP28 goal is a proxy for electrification, in all its forms. That means road transport, home cooking and heating and industrial use of electricity for heat. It looks quite possible, even in the slower Archipelagos transition.
Finally there is the goal relating to hydrogen production, which seems very ambitious, but perhaps this depends on how it is interpreted. The letter calls for an overall doubling of hydrogen production, without a year being given. The Financial Times reported this as ‘Double hydrogen production to 180mn tonnes per year by 2030’. This number relates to the current global production of hydrogen which is about 90 million tonnes per year, but most of this sits within industrial processes such as the Haber process to make ammonia. The Haber process starts with natural gas, and hydrogen is an intermediate product in the production. Most global hydrogen production today comes from natural gas, releasing carbon dioxide as part of the conversion process. Some also comes from coal, with significantly increased emissions as a result.
Hydrogen is also produced as a final energy product where it is then sold for other uses, such as in a fuel cell to power a truck or within a subsequent industrial process. This is where the future lies, with final energy green hydrogen replacing fuels such as diesel in trucks, Jet A1 in planes and coal in iron ore smelting. These applications amount to a small production of final energy hydrogen today, perhaps only 100,000 tonnes per annum. So doubling this number wouldn’t be ambitious at all and is almost certainly not what the letter is calling for.
If we use the current overall global production of hydrogen and wish to double that by 2030, then the COP28 target would have the world at 180 million tonnes per year (as the Financial Times noted), which presumably means about 80 million tonnes for new final energy applications such as fuel cells in transport and iron ore smelting. This is considerably in excess of even the Sky 2050 scenario, which sees 10 million tonnes in 2030 and doesn’t reach 80 million tonnes of final energy hydrogen until 2040. In Archipelagos that date moves out to 2055.
The challenge for hydrogen may not be the ability to produce it, given that electrolyser manufacturing capacity is ramping up very quickly. 80 million tonnes per annum of green hydrogen would require some 600 GW of electrolyser capacity to produce. According to the IEA, global electrolyser manufacturing capacity reached almost 11 GW per year in 2022, and based on company announcements, the global manufacturing capacity for electrolysers could reach more than 130 GW per year by 2030. This could deliver nearly 400 GW of capacity by 2030, still short of the COP28 goal but not markedly so. It’s also possible that significant electrolyser capacity could be introduced to replace the existing steam reforming of natural gas for ammonia production and this would certainly reduce emissions; but it wouldn’t create new demand for hydrogen as required by the target.
The real issue for hydrogen as final energy (or a hydrogen carrier such as ammonia which is being looked at for ships) is creating that demand. There is almost none today. There are very few trucks that use hydrogen fuel cells – no planes, no ships and almost no industrial processes (other than those where it is an intermediate product now). In natural gas-based direct reduction ironmaking hydrogen does play a role, though this is in combination with carbon. Pure hydrogen is not currently used in ironmaking applications apart from a handful of pilot plants. The 10 million tonnes of new hydrogen final energy demand in Sky 2050 by 2030 is largely split between industry and road freight. Shipping is only just emerging in 2030 and aviation not until the 2040s.
While the hydrogen goal seems almost misplaced in its ambition or perhaps has simply been misinterpreted, the package of targets still represents an important step to achieve by the end of this decade. But what is puzzling is why the package didn’t include a specific target for carbon capture and storage given that the UAE and Dr. Sultan al-Jaber have discussed the importance of the technology on many occasions. Irrespective of the objections of some environmental groups, repeated analyses by multiple organisations (including the Shell scenarios) show that without carbon capture and storage technology the 1.5°C goal cannot be realised. We also shouldn’t forget that while CCS projects can take a few years to deliver, their scale is significant at 1+ million tonnes CO2 per year per project once operational.
A substantial but not unachievable goal for 2030 would be to see the construction of 500 large scale (> 1 million tonnes per annum CO2) CCS facilities. As of late 2022, the Global Carbon Capture and Storage Institute reported 30 facilities in operation, 11 under construction and about 150 in various stages of planning.
Note: Shell Scenarios are not predictions or expectations of what will happen, or what will probably happen. They are not expressions of Shell’s strategy, and they are not Shell’s business plan; they are one of the many inputs used by Shell to stretch thinking whilst making decisions. Read more in the Definitions and Cautionary note. Scenarios are informed by data, constructed using models and contain insights from leading experts in the relevant fields. Ultimately, for all readers, scenarios are intended as an aid to making better decisions. They stretch minds, broaden horizons and explore assumptions.