Deep Dive 4

Material flow analysis of petrochemicals production in the UK

Fanran Meng
Jonathan Cullen
Rick Lupton

1 Department of Engineering,

University of Cambridge, Trumpington Street, Cambridge, CB2 1PZ, UK.

2 Department of Mechanical Engineering,

Centre for Sustainable and Circular Technologies (CSCT), University of Bath, UK

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

We use a new methodology to map the emissions and energy use of the petrochemical sector in greater detail than existing studies. An effective, targeted emissions mitigation strategy requires both a big-picture view of the sector and an understanding of the granular detail so that broad goals can be translated into specific interventions. Here, we demonstrate the use of semantic web technologies, which allow us to navigate multiple levels of detail and aggregation with the same tool. This has not been possible before now.

Figure 2 Skeleton diagram showing how processes and objects are aggregated across the chemical supply chain processes: a) overview of chemicals and plastics production; b) an example of detailed skeleton of primary chemicals in a).

the problem

The petrochemical industry is energy-intensive and generates large volumes of waste and emissions; the sector accounts for 30% of industrial energy use and 17% of industrial CO2 emissions across the globe. Digging into the details beneath these top-level figures is crucial if we want to make the sector more efficient. We need to understand precisely where and how emissions are produced so that we can design tailored policies and strategies to reduce them. Researchers have set out to track the flows of mass and energy in the industry, but so far, studies have only considered the global picture and have lacked the detail and resolution necessary to allow such targeted mitigation efforts.


We use a new methodology that allows us to map the mass flows of the petrochemical industry and visualise them at different levels of detail (aggregation), ‘zooming out’ to see the industry at a country level, or ‘zooming in’ to specific parts of the production system. As an example, we demonstrate how this new methodology can be used to ‘zoom in’ and illuminate the detail of primary chemicals production in the UK.

The mass flow map is created using semantic web technologies, which build a skeleton of the different processes and flows that make up the complex supply chains of the petrochemical sector. The skeleton is created using knowledge of the hundreds of different objects that flow through the industry and the many processes that transform them. The model is then fleshed out using process recipes, which tell us how much of each material ingredient is needed for each process and the proportions of the different outputs that are produced. The framework of objects, processes, and recipes is then combined with observational data to produce a picture of how (and how much) material flows and is transformed throughout the system.

Key Results
  • A more granular understanding of the emissions from petrochemical production is required. Our map builds this detail for the UK using capacity data from the ICIS database and knowledge of how specific production processes operate in the UK.
  • The mass flow map produced allows us to view the industry at the top level but also to delve into the details of smaller-scale processes. This enables us to identify areas where specific, practical changes can be made to increase material efficiency in line with wider commitments and goals.
  • For example, ‘zooming in’ to primary chemical production in the UK shows us that using emission-free energy to electrify steam cracking, the dominant production route, would reduce emissions from that particular process by up to 90%.
  • However, for the sector to meet its emissions targets, it will not be sufficient to simply refine existing production routes. Entirely new chemical production technologies will be required.
  • With our map of the chemical industry, we can track flows of material and energy throughout the supply chain, from chemical building blocks through to products and stocks. This is a valuable resource for developing and evaluating strategies which reduce emissions by keeping carbon in the loop.

Keep scrolling to read the full Deep Dive

Explore more Deep Dives

Deep Dive 1
Model-based energy and emission analysis of ammonia production for improved process sustainability

Ammonia production is highly unsustainable. Our model-based analysis reveals that in the US, emissions from the ammonia industry could be reduced by 10% by taking measures to enhance the efficiency of existing plants. To reduce emissions to near-zero by 2050, more fundamental changes to production pathways are necessary. Low-carbon technologies such as water electrolysis and carbon capture and storage will need to be developed and deployed to achieve this near-zero target with minimum cost.

Deep Dive 2
Global mass flow of nitrogen fertilisers and mitigation options

Nitrogen fertilisers account for 5% of global greenhouse gas (GHG) emissions. Our global mass flow model uses a whole-systems approach to explore combinations of mitigation interventions. If all the interventions we consider were applied together, GHG emissions from synthetic fertilisers could be reduced by 84% by 2050. The single most effective intervention is increasing the efficiency of nitrogen fertiliser application to plants during their use phase.

Deep Dive 3
Recycled polymers as a feedstock for chemical manufacturing supply chains in the United States: A network analysis for polyethylene pyrolysis

If 5 Mt of PE waste were chemically recycled each year, and the resulting ethylene and propylene were put to use by the chemical industry, other sources of these chemicals would be displaced. The knock-on effects of this would span a large number of products and processes. These wider repercussions need to be considered in any evaluation of chemical recycling as an emissions-reduction strategy.

Deep Dive 5
The economic context

We need to consider the potential wider economic impacts of climate change mitigation actions taken within the petrochemical sector. Therefore, we have developed an integrated macroeconomic model that combines both material and economic flows. The model is calibrated for the US context so that it represents the real-world stocks and flows of mass, energy, population, and money.

Deep Dive 6
Why is a Network Analysis of the Petrochemical Industry useful?

A network analysis of the petrochemical industry reveals the complexity and interconnectedness of the sector and indicates how changes made in one plant or location would have implications elsewhere. This knowledge helps policymakers to target strategic locations or companies for emissions mitigation.

Read the full Deep Dive


The petrochemical sector is responsible for 30% of final industrial energy use and releases 17% of global industrial CO2 emissions. The petrochemical manufacturing system relies on hundreds of processes and produces hundreds of different types of objects. It is vital to track plastics, fertilisers and other petrochemicals from production through to end-use applications. In this study, we mapped the mass flows of petrochemicals at a country level in the year of 2018 with higher resolution than existing maps. We used a novel mass flow mapping methodology that aggregates processes and objects by drawing skeletons based on semantic technologies. The technology formalises the underlying structure of the complex chemicals production system, processes the data, and aggregates flows to enable the visualisation of results and data integration at different levels of detail. The mapping is important for linking intervention actions back to the emissions.


In recent years, demand for petrochemical products has soared. Today, the global petrochemical sector makes nearly 1 billion tonnes of chemical products, including 420 Mt of plastic products; 290 Mt of nitrogen fertilisers, fibre, and rubber; and 250 Mt of other products like solvents, additives, and explosives.1 Demand for the seven primary chemicals — ammonia, ethylene, propylene, methanol, benzene, toluene, and mixed xylenes — is anticipated to increase in the future, with many products expected to double in demand by 2050.

This anticipated growth in demand makes taking action to achieve net-zero targets both more challenging and more pressing. Detailed knowledge of the end-use applications of chemical products will be required to inform the best strategies for meeting these targets. The petrochemical sector is responsible for 30% of final industrial energy use, including 11% of global oil demand and 10% of global natural gas demand, and releases 17% of global industrial CO2 emissions2. Emissions arise from chemical reactions and high-temperature heat generation (direct process emissions); from energy conversion in the upstream energy sector (indirect emissions); and from end-of-life treatment of products. Additional emissions are released from the use phase of some petrochemicals (e.g. fertilisers) and from fugitive emissions (e.g. methane) released by upstream oil and gas operations. Additional non-GHG emissions have a significant environmental impact, such as fertiliser run-off contributing to eutrophication,3 bioaccumulation of toxic chemicals in organisms, and plastic waste in the world’s oceans4 harming sea life.5

To reduce waste and emissions from the industry, it is vital to track plastics, fertilisers, and other petrochemicals from production through to end-use applications. This knowledge will enable the identification of additional climate change mitigation options in the sector.

The chemical and pharmaceutical industry is the United Kingdom's second largest industry and is a hugely important part of the nation's economy. UK chemical sales reached £31.8 billion in 2020 and generated a share of seven percent of the European Union’s total chemical revenue.6 The value of chemical imports from EU27 countries to the UK reached £20.4 billion, while the UK's exports to EU27 countries totalled £17.3 billion. The petrochemical manufacturing system relies on hundreds of processes and produces hundreds of different types of objects.

This study describes our integrated model framework, which links data on energy and emissions directly to individual material flows across the supply chain. This framework can assess the carbon emissions and environmental impact of petrochemical products from five dimensions: time, locations, greenhouse gas types, chemical types, and life cycle stages. It is employed here to map the mass flows and estimate greenhouse gas emissions of the petrochemical supply chain in the UK using the best available chemical databases. Further research will use the framework to discuss greenhouse gas mitigation strategies on both the supply side and demand side of the petrochemical industry.

Scroll to keep reading


Current mass flow of nitrogen fertilisers and GHG emissions

We use a novel mass flow mapping methodology that uses semantic web technologies to manage the complexity of the processes and data sources in the petrochemical sector.7 The technology formalises the underlying structure of the complex chemicals production system, processes the data, and aggregates flows to enable the integration of data and the visualisation of results and at different levels of detail. We map the mass flows of petrochemicals at a country level in 2018 with higher resolution than the existing global maps.8 It consists of the following elements:

  • System structure, defining the object types (types of material, goods, or energy) and processes (activities that transform objects into different objects).
  • Observations, representing a measurement of the magnitude of flows through a part of the system, in a particular geospatial location and time period.
  • Process recipes, defining the inventory of inputs and outputs to processes while respecting mass and energy balances.

Skeleton Hierarchy of Process

We employ a “skeleton” hierarchical structure of processes7 to serve as a common ground for linking different levels of detail on the chemical emissions and production data. The skeleton diagram in Figure 2 shows how objects are transformed from raw materials into primary/intermediate chemicals and products.

Observations - Mapping chemical production

The PRobs tools use the system structure described in the previous subsections as a basis for a linear constrained optimisation to calculate the mass flows. This is achieved by enforcing the mass balance of each process and constraining the optimisation with observations.

Observations include the production, import and export of objects and inputs/outputs of processes. The dataset used to constrain the mass flows is drawn from the ICIS Supply and Demand database,9 and the PRODCOM database of production values.10 These databases were chosen as they provide the most comprehensive coverage of production, trading, and consumption data for a wide range of chemical products. In the following sections, we use the example of primary chemicals production to explain how observations are obtained.

Steam cracking is a process for obtaining a variety of alkene and aromatic compounds from various alkane feeds. The main products of interest are primary chemicals: ethylene, propylene, butylenes, BTX (benzene, toluene and xylenes); the main feed materials are ethane, propane, butane, naphtha, and gas oil. Less valuable substances are co-produced, such as hydrogen, methane, and heavier alkanes.

Steam cracker capacities are typically defined according to the amount of ethylene they could produce in a year. This definition can be expanded to include other product capacities using a product capacity vector. Further details can be found in Levi and Cullen (2018).8

where Ft is the total quantity of all feeds in Mt yr−1; yi,n is the yield of product “i” from a unit of feed “n” and fn are the fractions each feed forms of Ft as percentages, \

; and Pi are capacities of each product (ethylene, propylene etc.) in Mt yr−1.

The feed composition of a steam cracker is obtained from ICIS database. Table 1 is obtained based on plant capacity, and we then apply the plant utilisation rate of 86% obtained from ICIS to get the real production quantities in 2018. For comparison, the long-term average capacity utilisation rate in EU27 is 81.6%.6 The yield matrix needs to be populated with typical steam cracking yield data as described in Table 1. As can be seen in the yield table, the product slate of steam crackers varies markedly between light (ethane, propane and butane) and heavy (naphtha and gas oil) feeds. Results of inputs and outputs are shown in Figure 3.

Chemicals and Plastics Production Recipe

Balancing process characterisations stoichiometrically requires the balanced process equations and the relative weights of each term in the equation. The weights of each term are used to estimate the proportions of secondary reactants and products of each process. The input requirement estimates are based on previous work and used to calculate the yield losses for processes.8 The information from the stoichiometry and IHS PEP Yearbook11 process recipes are used alongside supply and demand data from the ICIS database to estimate the quantities of each material entering and exiting each process.

Scroll to keep reading

Fig. 1 Overview of the model framework based on semantic web technologies.

Figure 2 Skeleton diagram showing how processes and objects are aggregated across the chemical supply chain processes: a) overview of chemicals and plastics production; b) an example of detailed skeleton of primary chemicals in a).

Table 1 Steam cracking yield matrix and feed composition.

Figure 3 UK steam cracking quantities in 2018 shown in a simplified Sankey diagram.

Results & Discussion

Material Flow mapping of the UK Chemicals productions

We first created a skeleton of material flows which helps us to visualise and understand how different processes and objects are connected with one another in the chemical industry. IHS' process recipes and ICIS production, trade, and consumption data were integrated into this underlying framework to produce the map of mass flows.

As shown in Figure 4, we map all the chemical flows from the refinery feedstocks through to different chemical products in the United Kingdom in 2018. The figure also shows the scale of trade between the UK and third countries. In 2018, the UK chemical production industry converted 3.29 Mt of refinery feedstocks, producing 1.82 Mt of primary chemicals, 1.26 Mt of organic chemicals, 0.89 Mt of fertilisers (importing 0.95 Mt and exporting 0.47 Mt), 1.53 Mt/year of polymers, and 0.028 Mt/year of overall waste chemicals. Trade with third countries is significant across the whole production chain, with 5.03 Mt/year of intermediate goods being imported and 4.94 Mt/year being exported.

Primary chemicals account for ~40% of total production emissions and thus there is a need to understand the technical detail about where engineering process can change.12 Steam cracking is the main source of primary chemicals. Our estimate is derived from the feed and product profile of the UK steam cracker. Capacity data for 13 steam crackers are gathered from the ICIS supply and demand database.9 A detailed map of material flows between production of primary chemicals and feedstocks can be found in Figure 5.

Ammonia contains nitrogen, which is fundamentally different from other primary chemicals containing embedded carbon. In the UK, 0.5 Mt is produced, ~80% of which is exclusively used for producing nitrogen fertilisers. As ammonia production is the largest contributor to Scope 1 & 2 emissions by the chemical industry, abating its emissions is critical to enable a sustainable future. This is especially pertinent because ammonia is considered the most promising pathway toward decarbonising shipping and power in specific countries.


We have developed a bottom-up model to map the key material flows of the UK petrochemical sector from fossil fuel feedstocks through to chemical products. The mapping demonstrates both the overall scale of the system and the technical detail about where engineering processes can change. Here, we have focused on primary chemical production as an example of this multi-level visualisation.

Primary chemical production is key for downstream chemical products and accounts for two-thirds of energy consumption in the petrochemical sector.13 The dominant steam cracking process requires temperatures of around 850°C to break down naphtha for further processing. If this process could be electrified using zero-emissions energy sources, CO2 emissions from steam cracking could be reduced by up to 90%. Major chemical manufacturers including BASF, SABIC, and Linde have established a consortium to jointly investigate the creation of the world’s first electrical naphtha steam crackers.14

In the chemicals industry, considerable reductions in emissions intensity have already been achieved by switching to lower-carbon fuels, improving energy efficiency, and using catalysts to reduce N2O emissions of nitrous oxide. However, these measures were refinements of existing technologies. To enable the significant, absolute GHG reductions required for climate stabilisation, entirely new chemical production technologies are needed. Our multiscalar mapping of the sector’s flows of mass, energy, and emissions will help identify where significant emissions savings can be made by improving current practice, and where these more fundamental shifts in production pathways are required.

The transition to a more sustainable chemical sector will be about ‘how’ we produce, but also about ‘what’ we produce and ‘what’ we use it for. Carbon is and will remain a core element of many chemicals. Instead of simply trying to ‘decarbonise,’ we need to consider how carbon flows through the chemical sector and focus on strategies which keep carbon circulating within the system rather than emitting it to the atmosphere; carbon is only negative for climate change when it is released to the atmosphere in greenhouse gases. Our mapping tool helps us to track the flows and stocks of material and energy throughout the entire chemical supply chain. It also clarifies the ‘how’ and ‘what’ of chemical production and helps us explore how future interventions might impact emissions. The next step is to focus on mitigation strategies that can keep carbon in the loop and foster sustainable carbon cycles.

Scroll to keep reading

Fig. 4 Map of all chemicals flows in the United Kingdom in 2018. This is a Sankey diagram where the width of the lines is proportional to the mass of the flows. The material flows from left to right are from material extraction to products. The truncated grey flows represent trade flows that either enter or leave the United Kingdom. The flows to the left represent imports and the flows to right represent exports.

Fig. 5 Map of primary chemicals flows in the United Kingdom in 2018

  1. International Energy Agency. The Future of Petrochemicals Towards more sustainable plastics and fertilisers. (2018).
  2. International Energy Agency. World Energy Balances: Overview. (2019).
  3. Schindler, D. E., Vallentyne, J. R., Brix, H. & Long, L. M. The algal bowl: overfertilization of the world’s freshwaters and estuaries Edmonton. Preprint at (2008).
  4. Jambeck, J. R. et al. Plastic waste inputs from land into the ocean. Science (1979) 347, 768–771 (2015).
  5. Van Cauwenberghe, L. & Janssen, C. R. Microplastics in bivalves cultured for human consumption. Environmental Pollution 193, 65–70 (2014).
  7. Germano, S., Saunders, C., Horrocks, I. & Lupton, R. Use of Semantic Technologies to Inform Progress Toward Zero-Carbon Economy. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 12922 LNCS, 665–681 (2021).
  8. Levi, P. G. & Cullen, J. M. Mapping Global Flows of Chemicals: From Fossil Fuel Feedstocks to Chemical Products. Environ Sci Technol 52, 1725–1734 (2018).
  9. ICIS. ICIS Supply and Demand Data Service. (2021).
  10. eurostat. Overview - Prodcom - statistics by product - Eurostat. (2022).
  11. S&P Global. Chemical Process Economics Program PEP. (2022).
  12. Meng, F. et al. Planet compatible pathways for transitioning the chemical industry. (2022) doi:10.26434/CHEMRXIV-2022-HX17H-V2.
  13. International Energy Agency. Primary chemical production in the Sustainable Development Scenario, 2000-2030. (2020).
  14. The Chemical Engineer. Greener chemicals: steam cracking could go electric by 2023 - News . (2021).

Explore more Deep Dives

Download this Deep Dive

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat.