Industrial energy use and carbon emissions reduction in the chemicals sector: A UK perspective

Applied Energy xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

Industrial energy use and carbon emissions reduction in the chemicals sector: A UK perspective ⁎

Paul W. Griffina, , Geoffrey P. Hammonda,b, Jonathan B. Normana a b

Department of Mechanical Engineering, University of Bath, Bath BA2 7AY, United Kingdom Institute for Sustainable Energy and the Environment (I•SEE), University of Bath, Bath BA2 7AY, United Kingdom

H I G H L I G H T S decarbonisation of the UK Chemicals sector has been evaluated. • Future improvement potential of different technological interventions was assessed. • The ‘technology roadmaps’ were also developed for various alternative scenarios. • 2050 practice technologies will prompt short-term energy and CO emissions savings. • Best • The prospects for longer-term, ‘disruptive technologies’ are far more speculative. 2

A R T I C L E I N F O

A B S T R A C T

Keywords: Chemicals Industrial energy analysis Carbon accounting Enabling technologies Improvement potential United Kingdom

The opportunities and challenges to reducing industrial energy demand and carbon dioxide (CO2) emissions in the Chemicals sector are evaluated with a focus on the situation in the United Kingdom (UK), although the lessons learned are applicable across much of the industrialised world. This sector can be characterised as being heterogeneous; embracing a diverse range of products (including advanced materials, cleaning fluids, composites, dyes, paints, pharmaceuticals, plastics, and surfactants). It sits on the boundary between energy-intensive (EI) and non-energy-intensive (NEI) industrial sectors. The improvement potential of various technological interventions has been identified in terms of their energy use and greenhouse gas (GHG) emissions. Currently-available best practice technologies (BPTs) will lead to further, short-term energy and CO2 emissions savings in chemicals processing, but the prospects for the commercial exploitation of innovative technologies by mid-21st century are far more speculative. A set of industrial decarbonisation ‘technology roadmaps’ out to the mid-21st Century are also reported, based on various alternative scenarios. These yield low-carbon transition pathways that represent future projections which match short-term and long-term (2050) targets with specific technological solutions to help meet the key energy saving and decarbonisation goals. The roadmaps’ contents were built up on the basis of the improvement potentials associated with various processes employed in the chemicals industry. They help identify the steps needed to be undertaken by developers, policy makers and other stakeholders in order to ensure the decarbonisation of the UK chemicals industry. The attainment of significant falls in carbon emissions over this period will depends critically on the adoption of a small number of key technologies [e.g., carbon capture and storage (CCS), energy efficiency techniques, and bioenergy], alongside a decarbonisation of the electricity supply.

1. Introduction 1.1. Background The chemical and petrochemical industry represents the largest contributor to industrial energy demand worldwide. It accounts for about 10% of global total final energy consumption and 7% of

⁎

‘greenhouse gas’ (GHG) emissions associated with industry [1]. Chemistry provides the fundamental basis for the synthesis of core intermediate and end products in order to satisfy human needs. It supplies inputs to matter transformation chains in other industrial sectors, e.g., plastics, composite materials, industrial gases, fertilizers, and so on [2]. These products are key to the modern global economy stretching from agriculture to medicine, through fuels, plastics and synthetic textiles.

Corresponding author at: CDP - Global Environmental Reporting System, 71 Queen Victoria Street, London EC4V 4AY, United Kingdom. E-mail address: paul.griffi[email protected] (P.W. Griffin).

http://dx.doi.org/10.1016/j.apenergy.2017.08.010 Received 23 January 2017; Received in revised form 21 July 2017; Accepted 6 August 2017 0306-2619/ © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/).

Please cite this article as: Griffin, P.W., Applied Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.08.010

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

significant role in energy and GHG emission reductions going forward [7]. Cefic [8] were aided by Ecofys, the energy and sustainability consultancy, in the development of their 2050 European chemicals roadmap to a competitive, low-carbon future. They emphasised the importance of making changes to the sectoral fuel mix, particularly for heat generation and Nitrous Oxide (N2O) production. This would yield again about 15% reduction in GHG emissions by 2050, although they noted that deeper cuts could be achieved via the decarbonisation of the power system and the adoption of CCS facilities [8]. Both the latter options would be costly and necessitate technological breakthroughs. Indeed, a novel feature of the Cefic study [8] was the focus on the adverse impacts that energy and climate policy costs are likely to have on European competiveness vis-à-vis chemicals production in the USA and other regions. The ECF were likewise concerned about price competitiveness [9], albeit in the context of the so-called energy policy trilemma: competitiveness, sustainability and security of supply. They studied the transition dynamics in the chemicals industry drawing on the Cefic roadmap [8]. ECF argued [9] that substantial GHG emissions reduction could be achieved through process and energy efficiency improvements, alongside greater resource ‘circularity’ or value chain collaboration. Thus, they suggested that by seeking out “cross-process, cross-company, cross-sector, and cross country abatement opportunities” European price competiveness in the chemicals sector could be maintained. Aggregate studies of the chemicals industry on a global or region scale have their limitations. Each country has, in reality, its own distinctive historical background, structural characteristics (including access to resources), and potential for energy savings and decarbonisation. Therefore the present work seeks to draw out lessons from the chemicals sector and its development in Britain. Industry as a whole in the United Kingdom (UK) accounts for some 21% of total delivered energy and 29% of carbon emissions. There are large differences between industrial sectors in the end-use applications of energy, especially in terms of products manufactured, processes undertaken and technologies employed (see Fig. 1 [10]; where the final demand for energy by broad UK industrial sectors is depicted against various energy use categories). It is clear that the chemicals sector as seen in Fig. 1 gives rise to the highest industrial energy consumption; mainly due to low temperature heat processes (30%), electrical motors (19%), drying/ separation processes (16%), and high temperature heat processes (11%) [10]. UK industry overall has been found to consist of some 350 separate combinations of sub-sectors, devices and technologies [11,12]. Nevertheless, it is the only end-use energy demand sector in the UK that has experienced a significant fall of roughly 40% in final energy consumption since the first oil price shock of 1973/74 [11,12]. This was in spite of a rise of over 40% in industrial output in value added terms. However, the consequent aggregate reduction in energy intensity (MJ/ £ of gross value added) masks several different underlying causes: enduse efficiency {accounting for around 80% of the fall in industrial energy intensity; largely induced by the price mechanism); structural changes in industry [a move away from energy-intensive (EI) industries towards nonenergy-intensive (NEI) ones, including services [11,12]}; and fuel switching (from coal and oil to natural gas and electricity that are cleaner, more readily controllable, and arguably cheaper for the businesses concerned).

But the world landscape of the chemical industry has dramatically altered in recent years. Asian chemical companies displayed a major growth trend from 1980 onwards, driven by their domestic markets, and presently accounts for half of the global market [2]. A surge of new investment in chemical processing plant then took place in the Middle East after the turn of the Millennium, based on the natural (oil and natural gas) resources in that area [2]. However, large-scale shale gas exploration and development in the United States of America (USA), after 2006 in particular, has given their chemical sector a strong economic advantage when compared with competitors elsewhere [2–4]. The price of gas halved with an impact on competitiveness of a large part of the global chemical industry (both in terms of natural gas feedstock and fuel). Companies from across the world are locating themselves in the US to take advantage of this ‘revolution’ [3]: leading to a potential ‘Golden Age of Gas’, according to the International Energy Agency (IEA) [4]. The IEA sees shale gas as contributing about 14% to global gas production by 2035 [3,4]. The IEA [1] partnered with the International Council of Chemical Associations (ICCA) and DECHEMA (Germany’s Society for Chemical Engineering and Biotechnology) in order to produce a technology roadmap for energy and GHG reductions in the world chemicals sector out to 2050. This roadmap was developed in the context of the IEA 2 °C Degree Scenario (2DS) for global warming. It focused on the particular role of catalytic processes that account for roughly 90% of chemical processing [1]. The roadmap built on earlier studies of Best Practice Technologies (BPT) for improving energy efficiency in the sector [5] and of carbon abatement innovations. The former used mainly a top-down approach to examine some 57 processes and indicated energy saving potentials of 5–15% [5]. BPT are the most advanced, economically viable technologies on an industrial scale. ICCA commissioned the German Öko-Institut, with the support of McKinsey & Company, to critically review Carbon Life Cycle Assessment (cLCA) studies related to the chemicals industry [6]. This indicated the potential for GHG emissions {carbon dioxide equivalent (CO2e)} abatement ‘from cradle to grave’, i.e., over a value chain incorporating the extraction of feedstock material and fuels through to production, transport and distribution, product usage, and the ‘end of life’ (disposal or recycling) phases. More than one hundred cLCA studies were submitted by ICCA member companies from around the world to McKinsey for evaluation, and then the results or data were reviewed by the Öko Institut. They concluded [6] that the best option for reducing global GHG emissions could be achieved by ensuring that each life-cycle stage of the value chain yielded its optimum contribution. Otherwise, a given stage might prevent larger CO2e reductions elsewhere along the chain, and consequently not elicit net global reductions overall. Worldwide assessments of the chemicals industry have been supplemented by regional ones. Europe, or the European Union (EU) countries, in particular has instigated a number of studies of an energyefficient and low-carbon chemicals sector including, for example, those sponsored by the European Commission (EC), via their Joint Research Centre (JRC) [7], the European Chemical Industry Council (Cefic) [8], and the European Climate Foundation (ECF) [9]. The JRC report [7] used a bottom-up model to evaluate the European chemical and petrochemical sector. It assessed the current technological status of 26 basic chemical products (including fertilizers, organic and inorganic substances, polymers, etc.), as well as the associated sectoral energy use and GHG emissions out to 2050. Over this period, it was found that the EU chemicals industry would experience a 39% rise in energy consumption, but a 15% fall in GHG emissions compared with the early 2010s [7]. Around 50 Best Available Technologies (BAT), the most effective innovations presently known, were examined. The importance of replacing fossil fuel feedstocks by sustainable alternatives such as hydrogen (from electrolysis driven by renewables) and biomass was recognised. Two cross-cutting technologies [combined heat and power (CHP), already widely used in the chemicals sector, and carbon capture and storage (CCS)] were recommended as having a potentially

1.2. The issues considered The present study builds on work by Dyer et al. [11] commissioned by the UK Government Office of Science (GOS) and on a recent ‘Advanced Review’ by Griffin et al. [13]. In each case, a variety of assessment techniques for determining potential energy use and GHG reductions were discussed. Griffin et al. [13] then evaluated the wider UK industrial landscape with the aid of decomposition analysis [14] in order to identify the factors that have led to energy and carbon savings over recent decades. They then assessed the improvement potential in two 2

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

Fig. 1. Final UK energy demand by industrial sub-sector and end-use. Source: Norman [10].

Fig. 2. Primary energy intensity, percentage of costs represented by energy and water, and mean primary energy use per enterprise (reflected by the area of the data points). Source: adapted from Griffin et al. [13].

surfactants), and as sitting on the boundary between EI and NEI industries (see Fig. 2 [13]; where broad UK industrial sectors are split into an EI and NEI categories). [A high value in any of the measures shown in Fig. 2 suggests that the sub-sector is EI.] Nevertheless, it accounts for some 19% of GHG emissions from UK industry – the second largest sector after steel (see Fig. 3 [13]; where these emissions are split into broad sectors in pie chart format {both those emanating from energy use, including those indirectly emitted from electricity use, and process emissions}). The opportunities and challenges to reducing industrial energy demand and carbon dioxide (CO2) emissions (the principal GHG [12]) in the UK Chemicals sector have therefore been evaluated, although the lessons learned are applicable across much of the industrialised world. The data here has been largely extracted from an industrial Usable Energy Database (UED) that was produced for the UK Energy Research Centre (UKERC) by the present authors (see Griffin et al. [13,15,16]). A set of industrial decarbonisation ‘technology roadmaps’ out to 2050 are finally reported, based on various alternative scenarios: named Low Action (LA), Reasonable Action (RA), Reasonable Action including CCS (RA-CCS), and Radical Transition (RT) respectively. Such roadmaps represent future projections that match short-term (say out to 2035) and long-term (2050) targets with specific technological solutions to help meet the key energy saving and decarbonisation goals. Their contents were built up on the basis of the improvement potentials associated with various processes employed in the chemicals industry and embedded in the UED [13,15,16]. They help identify the steps needed to be made by industrialists, policy makers and other stakeholders in order to ensure the decarbonisation of the UK chemicals industry.

Fig. 3. Greenhouse gas (GHG) emissions from UK industry. Source: adapted from Griffin et al. [13].

sectors: ‘Cement’ and ‘Food & Drink’, which represent the EI and NEI industrial sectors respectively. Here the ‘Chemicals’ sector of UK industry is examined in terms of energy use and GHG emissions from product sub-sectors, as well as the improvement potential of its processes. This sector can be characterised as being heterogeneous (having a diverse range of product outputs, including advanced materials, cleaning fluids, composites, dyes, paints, pharmaceuticals, plastics, and 3

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

2. The chemicals sector

caustic soda (sodium hydroxide - NaOH), and artificial fertilizers [17]. The organic chemical industry produced artificial dyestuffs from coal tar, principally located in the Manchester and Glasgow areas [18], and high explosives from compounds of nitric acid and cellulose material. By the late 19th Century, it also started to develop plastics and artificial fibres [17]. The 1914–1918 Great War had the effect of making the UK recognise that it was vulnerable to disruptions in the importation of chemicals from Germany and its allies, and from a blockage of commercial shipping by German submarines [20]. That meant that Britain had to reorganise its chemicals sector in order to produce substitute materials for explosives [from conventional nitric acid to high explosives, like nitro-glycerine and trinitrotoluene (TNT)] and fertilizers (such as nitrates; traditionally imported from Chile). Together the two world wars of the 20th Century led to much more local manufacture of chemicals and a focus, for example, on the indigenous production of high explosives, gasoline from coal and eventually synthetic polymers [20,24]. The structure of the UK chemicals industry has remained much the same from that time onwards, although companies have come and gone through mergers and acquisitions [20]. New products, such as polythene, polyvinyl chloride (PVC) and nylon, were produced in the 1950s and 1960s as a result of high levels of research and development (R & D). This arose from R & D expenditure that was nearly double that in the whole of UK manufacturing sector as a percentage of sales income [24]. A dramatic growth in the chemicals industry followed in the 1960s and early 1970s, until it became the mature processing industry that it is today.

2.1. Historical development of the chemicals industry The way in which the chemicals industry has been perceived by humanity has changed over time. In the ancient world, crafts were originally developed to meet the requirements of a domestic setting [17,18], such as the baking of bread, brewing, pottery making and leather tanning. Arabic science from about 3500 BCE {before the ‘Common Era’ (CE)}, based largely in Egypt and the Near East, led to what is now recognised as chemicals [19–22]: the early smelting of metals [especially copper, gold and mercury (or ‘quicksilver’ - Hg), as well as alloys like bronze], for example, gave rise to an understanding of the properties of their chemical compounds. Although it was only around 2000 BCE that iron was common in Egypt. Before dyes could be applied to textiles, their fibres needed to be treated with substances known as ‘mordants’ [18]. The most common of these was alum (a salt deposit), but this required purifying via a process of recrystallization. Common salt (sodium chloride - NaCl), which was initially used for the preservation of fish and meat [18], along with sal ammoniac, were also widely known minerals from ancient times. Fuller’s earth, which can act as a natural cleansing agent was found in the Near East, whilst nitre (from natural deposits in Babylonia) or saltpetre (potassium nitrate – KNO3) may well have been employed to make nitric acid (HNO3) [17]. Glass production on a significant scale in Egypt was in place by about 1370 BCE, using soda (or ‘natron’ - Na2CO310H2O) from natural deposits [17,21]. Crude soap was made from oil and alkali, possibly by burning plant material to produce potash (or ‘vegetable alkali’); any mined or manufactured salt that contains potassium (K) in a watersoluble form [20]. Elementary yellow sulphur certainly existed in natural deposits near the Red Sea, and may have been adopted to make sulphuric acid (H2SO4) or ‘oil of vitriol’ [20]. Fermentation processes for the purpose of brewing to make alcohol was in widespread use from antiquity [17,18,21]. The founder of commercial chemistry was later viewed as being Jabir ibn Hayyan (c. 721–815 CE), known in Europe as ‘Geber the Alchemist’ [13], who experimented from a workshop in Kufa (Iraq) on a range of chemical processes and initiated “the classification of materials into spirits, metals and filtration” [21]. The fruits of Arabic chemical processing progressively migrated across Europe. However, the development of chemicals as they are known today really commenced with the so-called Industrial Revolution in the UK from about 1760 CE onwards [17,19,23]. It focussed on the discovery of ways of bulk-producing acids and alkalis. This came about from a fusion of empirical ‘rules of thumb’ with the chemical sciences [21]. Chemistry itself may be traced to the first categorization of chemical elements by Robert Boyle (1627–1691) in 1661 [23]. Many practical technologies for the transformation of raw materials (or ‘feedstocks’) into useful or aesthetically pleasing artefacts were far more advanced than the understanding of the fundamental principles of applied science during that period [19]. Some of the drivers behind the expansion of the UK chemical industry came from the prolific increase in textile production (particularly cotton manufacture) that necessitated ever more bleaching agents [18], the improvement in the processing of sulphuric acid, originally by Joshua Ward (1685–1761) and subsequently by John Roebuck (1718–1794), and then its large-scale use for the production of soda in a lead-chamber process devised by the French chemist Nicolas Leblanc (1742–1806) [17,23]. UK chemical processing became centred on Merseyside (within a triangle bordered by St Helens, Warrington and Widnes), Clydeside in central belt of Scotland, and Tyneside (in the North East of England between Newcastle and North Shields) [17]. Soap production was largely based again on Merseyside, where William Lever (1851–1925) established the seeds of the multi-national conglomerate Unilever [18]. The ready and cheap availability of sulphuric acid led to the production of soda instead of potash for soap and glass manufacture [23]. Other materials associated with the heavy chemicals industry included ammonia (NH3), nitric acid,

2.2. Structure of the modern chemicals sector Chemicals are a complex collection of many diverse and interacting sub-sectors covering a wide range of feedstocks, processes and products. Physical outputs are moved around on an international scale within or between major companies that are truly multi-national [24]. The industry is also highly focused on private R & D and protective of information, meaning that data availability is particularly poor. This high technology sector takes full advantage to modern developments in electronics and information and communications technology (ICT), such as for the automatic control of chemical process plants and automation in the use of analytical instruments [24,25]. The scale of operation of chemical firms range from quite small plants (of a few tonnes per year) in the fine chemicals area, where high purity is required, to giant ones in the petrochemical sector [24,25]. Batch production is employed by SMEs where small quantities of chemicals (up to around 100 tonnes per annum) are required. In contrast, continuous plants are typically used in cases where a single output, or related group of products, are demanded with plants of several thousands to millions of tonnes per year [24,25]. They often produce intermediates which are converted via downstream processing into a wide range of products, such as benzene, toluene and xylenes (BTX), ethylene, phenol, and PVC from petrochemical refineries or via ammonia plants [1,24]. Central to the production of organic chemicals is the steam cracking process which manufactures lower olefins, also referred to as high value chemicals (ethylene, propylene, butadiene and aromatics), from feedstocks deriving from oil or natural gas [22]. The feedstock is primarily naphtha and ethane-based, and most process fuel demand is met by fuel-grade process by-product gases. There were four steam crackers operating in the UK at the end of 2010 and three remaining in operation today. The physical unit of production for the sector is tonnes of high value chemical (thvc). Feedstock input in 2010 was about 4 Mt (according to the UK Government's Digest of UK Energy Statistics (DUKES) [26]) leading to an estimated production of 2.8 Mthvc. Based on the feedstock mix and an assessment of European cracker specific energy consumption (SEC) (according to the IEA) [27], process fuel demand was an estimated 53 PJ and electricity demand was 1 PJ [28]. Thus, direct GHG emissions were estimated to be 2.2 MtCO2e, with total emissions 4

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

was about 2.4 MtCO2e. Another key activity of the chemicals sector is ammonia (NH3) production. Ammonia is synthesised from hydrogen which is primarily manufactured by steam reforming of methane from natural gas. The process is exothermic and heat is used to raise steam for other ancillary processes. It also releases CO2 as a by-product. Two UK sites produce ammonia by conventional steam reforming [28]. Fuel and feedstock input in 2010 was 10 PJ and 19 PJ of natural gas respectively [29] for a production of about 857 ktNH3 [30]. Accordingly, process emissions were about 1 MtCO2e, whilst direct combustion emissions were 0.5 MtCO2e. Assuming an electricity requirement of 0.8 GJ/NH3 [18], total emissions from the sector were an estimated 1.6 MtCO2e. Baseline process efficiencies and plant load factors adopted in the present study were largely based on information covering the region of Western Europe. Nevertheless, the focus was on divisions 20–21 of the UK Standard Industrial Classification (SIC 2007) for economic activities, i.e., ‘chemicals and chemical products’ (20) and ‘basic pharmaceutical products and pharmaceutical preparations’ (21). The breakdown of the structure of the chemicals sector (SIC 20, 21), plus energy for petroleum feedstock production and CHP plant at integrated refineries, includes [28]:-

Table 1 Ultimate yields for steam crackers. Source: adapted from Griffin [28]. European data collated by Neelis et al. [31]. Chemical feedstock (kg product per tonne)

High-value chemicals Ethylene Propylene Butadiene Aromatics Fuel products/ backflows Hydrogen Methane Other C4 components C5 and C6 components C7 and nonaromatics < 430 °C > 430 °C Losses Process energy Backflows to refineries Share of total:

• Petrochemical processes • •

– Steam cracking (of ethylene, propylene, etc.). – Petrochemical feedstock production (of naphtha, ethane, etc.). Other chemical processes and miscellaneous – Identified chemical process (steam reforming, polymerisation, etc.). – Other (unidentified or ancillary processes, boiler plant, energy overheads, etc.). CHP and non-CHP power plant {including major power producers (MPP)} – Chemicals (includes one petrochemicals site and ‘other generation’). – Refineries (includes two integrated petrochemicals sites).

a

Naphtha

Gas oil

Ethane

Propane

Butane

Othera

Total

645

569

842

638

635

645

688

324 168 50 104 355

250 144 50 124 431

803 16 23 0 157

465 125 48 0 362

441 151 44 0 365

324 168 50 104 355

489 117 42 40 312

11 139 62

8 114 40

60 61 6

15 267 12

14 204 33

11 139 62

24 157 32

40

21

26

63

108

40

51

12

21

0

0

0

12

5

52 34 5 264 91

26 196 5 261 170

0 0 5 314 0

0 0 5 249 113

0 0 5 242 123

52 34 5 264 91

19 18 5 270 81

26%

3%

25%

22%

15%

8%

100%

Assumed identical to naphtha based steam cracking.

that account for the bulk of energy use in the sector: a Pareto-like approach. The rest of the improvement potential in the chemical industry was assessed by determining the possible impact of key cross-cutting technologies [13,15,16]. 2.3. Steam cracking of (lower) olefins

The baseline was modelled as 27 different sub-sector processes, each with its own physical or product output; as displayed in Fig. 4 [15,28]. Here direct net energy demands (fuel, steam, and electricity) are shown in terms of their current {or ‘baseline’} and BPT levels for chemical processes in the UK. However, due to data restrictions, only a few of the sub-processes identified were analysed in terms of their improvement potential. It may be argued that this level of disaggregation represents a useful split of the UK Chemicals industrial structure that others may find beneficial for future assessments. The procedure adopted was therefore to focus principally on a limited set of five chemical products/processes

This process is at the heart of the UK petrochemicals sub-sector and produces olefins, which form the basis of many other chemicals (including plastics). Olefins such as ethylene, propylene and butadiene are derived from gas and petroleum feedstocks like ethane, propane, butane, naphtha and gas oil. It is by far the largest energy-demanding subsector in terms of both fuel and feedstock within the chemicals sector. DUKES [26] provides data for the consumption of petrochemical feedstock from which it is possible to model the UK steam cracking process [28]. The typical yield of products deriving from each feedstock, along Fig. 4. Direct net energy demand (fuel, steam, and electricity) of the baseline and best practice technology (BPT) levels for chemical processes in the UK. Source: Griffin [28], adapted from Griffin et al. [15].

5

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

Fig. 5. Sankey energy flow diagram for the UK steam cracking process. Source: Griffin [28].

assumed to be met by natural gas. Fuel efficiency was taken as 17.4 GJ/ thvc; determined by recalculating the European average of 16.9 GJ/ thvc [28], but then correcting it for the UK feedstock mix. The calculation of emissions via this method provides data within 2% of the verified emissions figure under the European Union’s Emissions Trading Scheme (EU-ETS) [28]. This estimation was based on 2010 data. The average UK fuel SEC for Best Available Technologies (BAT) and Best Practice Technologies (BPT) was also calculated for the UK feedstock mix. The European BPT fuel SEC was 13.1 GJ/thvc [27], and so the UK BPT SEC becomes 13.5 GJ/thvc. According to Worrell et al. [33]), the BAT fuel SEC is 11 GJ/thvc for naphtha steam cracking and 12.5 GJ/ thvc for ethane cracking. Assuming the average of these SECs for other feedstock flows, BAT SEC for the UK was found to be 11.8 GJ/thvc. This

with the proportion of raw materials consumed in 2010 is shown in Table 1 [28,31]. Total feedstock consumption was 4070 kt; equating to a production of about 2800 kthvc. A Sankey diagram depicting the calculated baseline energy flows for steam cracking is shown in Fig. 5. This diagram is based on endothermicity; the theoretical amount of energy required in an endothermic reaction [32]. In this case, it was derived from the energy released from the combustion of process fuel. Estimates of efficiency, energy and carbon dioxide emissions associated with steam crackers are shown in Table 2 [27,28,31]. Combustion emissions were calculated from an estimation of process fuel demand. For simplicity, it was assumed that production of by-product fuel is equal to the approximated share as indicated by Neelis et al. [31] for process energy covered by feedstock. The remaining process energy is

Table 2 Energy and emissions analysis for UK steam cracking. Source: adapted from Griffin [28]. European data for SEC, by-product share and by-product emissions factors from Neelis et al. [31]; readjusted SEC from the IEA [27].

SC fuel SEC (GJ/thvc) SC fuel SEC readjusted (GJ/thvc) SC fuel SEC readjusted (GJ/t) ethylene SC fuel (PJ) Share of SC fuel SEC from feedstock By-product fuel CO2 emissions factor (tCO2/t) Supplementary fuel CO2 emissions factor (tCO2/t) Specific CO2 emission (tCO2/thvc) Specific CO2 emission tCO2/t ethylene) CO2 emission (MtCO2)

Naphtha

Gas oil

Ethane

Propane

Butane

Othera

Total

17.8 16.5 32.9 11.3 100% 48.7 – 0.80 1.60 0.55

20.7 19.2 43.7 1.6 95% 48.7 50.5 0.94 2.13 0.08

21 19.5 20.4 16.5 80% 43.3 50.5 0.40 0.42 0.34

17.6 16.3 22.4 9.4 100% 43.3 – 0.99 1.36 0.57

17.1 15.9 22.9 6.3 100% 43.3 – 1.02 1.47 0.41

17.8 16.5 32.9 3.5 100% 48.7 – 1.29 2.57 0.27

18.7 17.4 24.4 48.6 – – – 0.79 1.11 2.22

Abbreviations: SC-steam cracker; SEC-specific energy consumption. a Assumed identical with naphtha based steam cracking.

6

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

rudimentary assessment implies that the technical improvement potential was 22–38%.

comparison, technical improvement potential for UK ammonia production is about 10%.

2.4. Reformate fractionation of aromatics

2.6. Ostwald processing of nitric acid

The expression ‘aromatics’ (that mainly constitutes benzene, toluene, and xylenes – BTX), which are characterised by double-bonded carbon molecules [5] that can be modified easily. Their name arises from their pungent smell that stems from their extraction via the pyrolysis of ‘gasoline’/petroleum (a by-product of steam cracking) or from reformate (a product of the catalytic reforming of naphtha at refineries). These processes are estimated to use an average of 2 GJ final energy per tonne of BTX aromatics extracted [1]. Ultimate yield of such aromatics from steam crackers with various chemical feedstocks are listed in Table 1 [28,31]. BTX aromatics are critical to petroleum refining and the petrochemical industries [5], and demand for all three products has risen rapidly in recent years. Half of all toluene consumption is utilised as raw material in other aromatics production, i.e., for xylene production (using ‘hydroproportionation’) and for benzene production (employing ‘dealkylation’) [1]. Thus, the full calorific value (45 GJ/t) must be allocated between these chemicals in order to avoid double counting, and the feedstock value of toluene must therefore be corrected by the share of its consumption (by 50%) that is processed into the other aromatics. Benzene (C6H6) is one of the largest-volume petrochemicals and the largest of the BTX aromatics [1,7]. It has been used as a solvent and as a component of motor vehicle fuel for improving gasoline quality. However, 70–75% of benzene is consumed globally for the production of ethyl benzene; primarily for the production of polystyrene and cumene for phenol and acetone [7]. Its use has decreased drastically in recent years, because of the post-2008 economic downturn and its high toxicity. BTX aromatics are also important in the production of polymers, other chemicals and several consumer products (such as solvents, paints, polishes, and pharmaceuticals) [7]. Likewise, they are used in health and hygiene, food production and processing, transportation, information technology and other sectors. Like most petrochemicals, the consumption of BTX aromatics is strongly linked with consumer demand for plastics. Benzene is expected to grow at a rate of 3% annually until 2018, while the estimated growth in overall toluene consumption was less than 3% per year [7]. Xylene (C8H10) is actually the name of three isomeric forms; that depend on their relative place in the methyl groups. According to the JRC report [7] the consumption of xylenes are expected to grow at 4.5% annually until 2020.

Nitric acid (HNO3) is a colourless liquid that is used in the processing of inorganic and organic nitrates and nitro compounds for dye intermediates, explosives (TNT), fertilizers, pharmaceuticals, rocket fuel, and many organic chemicals. It is most commonly produced from ammonia (NH3) via the Ostwald process. The Latvian-born, German chemist Wilhelm Ostwald developed the process, which he patented in 1902. (He subsequently received the Nobel Prize for Chemistry in 1909 for his research on catalysis, chemical equilibria and reaction velocities.) In this process, ammonia is converted to nitric acid in two steps. In the first step ammonia is oxidized by heating with oxygen in the presence of a catalyst, such as a precious metal gauze [36] (e.g., platinum with 10% rhodiumto form), nitric oxide (N2O) and nitrogen dioxide (NO), together with water/steam. This is a strongly exothermic reaction that provides a useful heat source following its initiation. Then, in the second step, the N2O is absorbed in water, which then forms nitric acid (albeit in a dilute form with concentrations of up to 69% [36]). It also reduces a fraction back to NO. This NO is then recycled, and the acid is concentrated to the required strength by distillation. Nitric acid production with state-of-the-art technology [36] induces a significant amount of energy recovery. A modern dual pressure HNO3 plant has a net energy output of 11 GJ/t HNO3-N as highpressure steam, corresponding to about 2.4 GJ/t HNO3 (100%) [36]. The process evidently results in emissions of a potent greenhouse gas - nitrogen dioxide (N2O) [28]. Unfortunately, BAT in this area mainly has the effect of reducing CO2e emissions [7], rather than N2O. The NOX concentration after the absorption section of the Oswald process mainly depends on the pressure applied at that section. Low NOX emission levels can be achieved by high absorption efficiencies (e.g., by the application of high absorption pressures) and/or by end-of-pipe technologies. Among the end-of-pipe measures is the selective catalytic reduction (SCR) process for reducing NOX emissions. Emission reduction rates of up to 95% [34] can be achieved via the state-of-the-art SCR process for new and existing plants. In new plants a combination of the SCR process and high-pressure absorption (< 8 bar) are likely to result in NOX emissions at the level of 100–200 mg NOX/Nm3 (as NO2) [36]. According to Wiesenberger [36], Austrian nitric acid plants currently have emission levels between 1200 and 2750 mg N2O/Nm3, depending on process technology (particularly the pressure level during catalytic ammonia oxidation). The adoption of a high efficiency catalytic ammonia oxidation process will reduce this N2O formation. A commercialised and patented process for the homogeneous decomposition of N2O in the NH3 combustion unit is available [36] for new nitric acid plants. Several test projects for developing N2O abatement processes are underway in industry and at research institutes [36]. Some promising results have been reported [36] on catalytic decomposition processes of N2O in the ammonia combustion unit. This process could also be retrofitted onto existing nitric acid plants. Wiesenberger [36] suggests that catalysts for the catalytic N2O decomposition in the ammonia conversion unit are already being tested on a prototype scale, and that they might be available for industrial use in the near future.

2.5. Steam reforming of hydrogen and ammonia Steam reforming is a method of producing hydrogen from natural gas (methane) feedstock. The steam methane reformer (SMR) separates methane into hydrogen (H2) and carbon dioxide (CO2) as a by-product. Most ammonia (NH3) in the UK is also manufactured by steam reforming with methane from natural gas. In that case, the H2 is put through a secondary reformer, the Haber process, in which it is combined with nitrogen from air to form ammonia. The hydrogen for NH3 production in the UK is manufactured at two sites: the Ince plants at Billingham and Saltend. Ince use conventional routes at Billingham, while at Saltend natural gas is used as feedstock for acetic acid and acetic anhydride production. Hydrogen by-product from this process is subsequently fed through a small adjacent ammonia plant [34]. Natural gas for feedstock and combustion in conventional steam reforming was estimated from the UK GHG Inventory [34]. For the 2010 baseline, these were 17.1 PJ and 9.1 PJ respectively [28,30,21]. Production related to this natural gas feedstock was 857 kt, giving an overall fuel SEC of 30.5 GJ/tNH3. The SEC of BAT ammonia production is as low as 27.6 GJ/tNH3 and could be met by the conventional process route, together with heat exchange ‘autothermal reforming’ technology (according to the European Commission [35]). Based on this

2.7. Mercury cell process (chlor-alkali) This is a process of producing chlorine by electrolysis of a salt solution, a co-product of which is hydrogen (H2) [15,16,28,37]. It is alternatively known as the chloralkali (chlor-alkali or chlor alkali) process. In the mercury cell process that has been favoured in Europe [37], sodium (Na) forms an amalgam with the mercury (Hg) at the cathode. The two metals react with the water in a separate reactor (called a ‘decomposer’), where an H2 gas and a ‘caustic soda’, or sodium hydroxide (NaOH), solution at 50% are produced {two moles of NaOH per 7

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

mole of chlorine (Cl)}. Solid salt is required to maintain the saturation of the salt water, and it is usually re-circulated. The brine is first dechlorinated, and then purified by a precipitation-filtration process. It yields chemical products that are extremely pure, so that the chlorine (along with a little oxygen) can normally be used without further purification. The mercury process uses the most electricity and also requires measures to prevent serious environmental contamination. However, no steam is required to concentrate the caustic solution. Na must finally be removed from the hydrogen gas and caustic soda solution. Mercury losses have been considerably reduced over the years, as chlorine producers have increasingly moved towards membrane technology, which has much less impact on the environment [38]. Emissions for all mercury cells across Western Europe reached 0.68 g/t of chlorine capacity in 2015. Some 20 mercury-based chlorine plants currently remain to be phased out or converted to non-mercury technology at a cost of more than 3000 million €. These plants account for an ever decreasing part (less than 20% in 2015) of European chlorine capacity. Ayres and his co-workers [39,40] argued that “industrial secrecy” constrained the availability of the data needed to conduct proper LCA studies of the mercury cell chlor-alkali process in Europe. They therefore constructed a plausible emissions profile from process data and top-down mass balances [40]. There are two other basic processes for the production of chlorine and caustic soda from brine: the diaphragm cell (which dominated US chlor-alkali production in the 20th Century [37]) and the membrane cell (favoured in Japan [37]). The performance of all such devices is governed by the requirements of electrochemical technology [37]. Membrane cells constitute the most modern process, which yields both economic and environmental advantages. Chlorine (at the anode) and the H2 (at the cathode) are kept apart by a selective polymer membrane [38] that allows the sodium ions to pass into a cathodic compartment and react with the hydroxyl (OH) ions to form caustic soda. The depleted brine is dechlorinated and recycled back to the input stage. Overall, the energy consumption in a membrane cell process is of the order of 2200–2500 kWh/t, as against 2400–2700 kWh/t of chlorine for a diaphragm cell process [38]. Moussallem et al. [41], following a review of the development of chlor-alkali electrolysis, suggested that the replacement of the H2 evolving cathodes in classical membrane cells by improved cathodes would require novel electrolysis cell designs, but might lead to the reduction of the cell voltage, and correspondingly the energy consumption, by up to 30%.

Fig. 6. Schematic representation of an integrated top-down and bottom-up modelling approach for the UK industrial sector. Source: Griffin et al. [13], adapted from by Dyer et al. [42].

Energy Database (UED) [15,16], produced by the present authors for the UK industrial sector as part of the research programme of the UK Energy Research Centre (UKERC). Aspects of both top-down and bottom-up models were adopted, with detailed bottom-up studies set within a topdown framework. Using this approach would normally entail focusing on a number of sub-sectors for the bottom-up study [13], with the remainder of the sector being treated in a generic manner. Sub-sectors that use a large amount of energy are obviously prioritised for bottomup studies. In additional, sub-sectors that use energy in a relatively homogeneous manner are easier to analyse and this may also be considered when selecting appropriate sub-sectors. Sub-sectors that are not the subject of detailed bottom-up modelling require a focus on the potential reduction in emissions through widely used, ‘cross-cutting’ technologies [13,15,16].

3. Methods and materials 3.2. Modelling chemical processes 3.1. A hybrid top-down/bottom-up approach The full range of 27 chemical processes modelled and their contribution to the sector’s overall energy demand are depicted in Fig. 4 [28]. The method and assumptions for building the baseline structure are set out in greater detail in the UED [15] and associated spreadsheets [16], but they are summarised here. Chemical process plant capacity data was sourced from chemical profile articles of the industry intelligence provider (ICIS – see < http://www.icis.com/ > and again Griffin et al. [15,16]). Plant production levels were calculated by applying estimated load factors calculated, for the petrochemical industry, from information provided by the Association of Petrochemicals Producers in Europe (APPE) [43]. This body publishes production and capacity figures of key products and derivatives in Western Europe. The average capacity of the products listed was used for other organic chemicals. Inorganic chemical plant load factors were informed in part by further articles sourced from ICIS (Griffin et al. [15]). In cases where information was not available, a default value of 70% was adopted. The exceptions to these methods are: (i) olefins, which were calculated by analysing petrochemical feedstocks (extracted from the annual edition of DUKES [26]) with data on their typical product yields from steam cracking by Meyers [44]; and (ii) NH3, for which an estimate made by

There are two broad ways to modelling the industrial sector [13]: top-down and bottom-up approaches; as illustrated in Fig. 6 (adapted from Dyer et al. [42] and Griffin et al. [13]). A top-down approach splits industry into sub-sectors, usually based on available statistical data, and uses this data to determine energy use, output, energy intensity and other measures for which data is available. This approach has the advantage of covering a large proportion of energy demand, but it is limited by the level of disaggregation available from industry-wide statistical sources. Thus, the conclusions that can be drawn from such top-down studies are often only indicative in nature. In contrast, a bottom-up approach would typically focus on a single industrial subsector. Energy use can then be separated into lower order sub-sectors, processes or manufacturing plants. The data used for this type of bottom-up study typically comes from more specific information sources, such as trade associations, company reports, and case studies. Such a bottom-up study can therefore be useful in terms of presenting more accurate findings [42], although it will be limited in the breadth of its application. A hybrid approach was employed to develop the industrial Usable 8

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

Fig. 7. Sankey energy flow diagram of the UK Chemicals sector as modelled here; baseline data in 2010. Source: Griffin et al. [15].

4. Improvement potential

the US Geological Survey (USGS) [45] was employed. Baseline process efficiencies were taken from a report by the IEA [27], which relates to Western European production. The exception to this was ammonia, for which the data for combustion and feedstock use of natural gas was sourced from UK Office of National Statistics (ONS) [46], oxygen was taken from the International Iron and Steel Institute (IISI) [47], and hydrogen from the European Commission [48]. The sub-sector final demand fuel split [32,49], which is primarily natural gas, was applied to the processes as an average. The sub-sector was split for the purposes of determining steam and electricity demand between self-produced and imported products applied across the various processes. This necessitated that an average emissions factor was employed at the process level. The energy flows through the Chemicals sector are illustrated in Sankey diagram displayed as Fig. 7 [15]; for a 2010 baseline. The data underlying this sector model was primarily based on DUKES [26]. Special attention was given to combined heat and power (CHP) within the Chemicals sector. Some CHP plants were classified as major power producers (MPP), and so would normally be outside the scope of the final energy demand adopted by the UK Government’s statistical service for the purposes of compiling DUKES [26]. Here the fuel use for the Chemicals sector CHP and autogeneration were specified following discussions with industrial representatives and personal correspondence with the analysts at the UK Government’s former Department of Energy and Climate Change [DECC] (see again Griffin et al. [15]). Power consumption was split between grid and autogenerated electricity. Imported heat, exported heat and exported electricity were all accounted for. Approximately 60% of the electricity produced by CHP in the Chemicals sector is exported; 50% to the grid and 10% to other industrial sub-sectors. The remaining fuel demand not used in CHP or autogeneration is used by processes, in addition to auxiliary boilers and other miscellaneous equipment. The processes modelled in a bottom-up manner account for 92% of the non-CHP fuel demand. The steam and electricity demand represented by a summation of the bottom-up modelled processes accounted for 50% and 25% of the total demand respectively. These may appear low as the energy carriers have higher demand for energy overheads (such as the lighting and heating of buildings) and auxiliary processes associated with, but not included in, the chemical process demands listed by the IEA [27].

4.1. Process improvement Identified technologies and measures generally fit into three broad categories. Fuel switching was not identified separately, but occurs via many of the process substitution options. Natural gas, the least CO2 -intensive fossil fuel, is by far the most widely burnt fuel in the subsector; much of it being used in CHP plants. In a report for the IEA [27], a broad range of potential technological improvements for the global chemicals and petrochemicals sector was evaluated. This study aimed to determine which improvement option constituted the Best Practice Technology (BPT) out of a portfolio of 66 of the most common processes. BPTs represent the ‘best’ technology that is currently in use, and therefore economically viable. This can be distinguished from a Best Available Technology (BAT), which includes proven technologies that may not yet be economically viable. Although the IEA report [27] had a global remit, most of the BPTs and the average SEC data was European-sourced, and thus provides a reasonable representation for the UK setting. By taking the SEC data and applying it to UK process outputs, it was possible to build a simple bottom-up estimation of process energy demand and GHG emissions for the sector. Production from most processes was estimated by collecting plant capacity data from published sources (such as the ICIS [15,47]), and then applying estimated load factors. The weighted average load factor for the sector in 2010 was estimated to be 75% [28]. Exceptions to this approach were for the production of lower olefins and ammonia. The energy demand accumulated from the 27 identified processes applicable to the UK chemicals sector that have been examined in some detail in the present study are again represented in Fig. 4 [28]. It can be observed that production of high value chemicals from steam cracking demands by far the greatest amount of energy. The curve then rises steeply for titanium dioxide, chlorine, soda-ash and ammonia, before reducing in gradient. The curve peaks before the last process as some processes are exothermic, and therefore net energy producers. In aggregate, BPTs have the potential to reduce direct energy demand from 98 PJ to 71 PJ, or save 28 PJ (28%) [28]. Primary energy reduction may be simply estimated on the basis of using the sector efficiency of steam generation (95%) and weighted average efficiency of owned and imported electricity. This yields figures of from 121 PJ to 88 PJ, or a saving of 33 PJ (∼27%) [28]. Given the dominance of lower olefin production, applying the BAT standard for steam cracking provides a 9

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

specifically depicted. It can be seen that these options generally perform well compared with the alternative process routes. Abatement from CCS is more advantageous with the biomass-based process routes, because it gives rise to so-called ‘negative emissions’ (provided that carbon in the product is not released at a later stage). On the other hand, these options are more resource efficient and rely on established supply systems. The prospect of applying CCS is reasonable compared with other industrial sectors, such as iron and steel [28], or cement [13,28]. All steam crackers and ammonia plants are situated within CCS cluster regions (see Fig. 9 [13]). Here the distribution of CO2 point sources and potential UK CCS cluster regions are illustrated. The cost of capture from ammonia is also very low because of the purity of its process CO2 emissions. Retrofitting post-combustion capture at steam cracker furnaces has been estimated to cost in the range of £28–52/tCO2 [52]. This range covers the options identified in this study with the lower limit being comparable to oxyfuel capture technology and attained through the utilisation of process waste heat. Unlike oxyfuel capture, however, post-combustion treatment would not require major process modification. Irrespective of capture costs, the cost and availability of CO2 transport and storage presents an even greater challenge [53,54]. Cluster regions of industrial activities have been identified for storage under both the North Sea [54] and the Irish Sea [57] are shown in Fig. 9 [13]. It is interesting to note that the main industrial regions of the UK remain very much as they were at the time of the Industrial Revolution (see Section 2.1 above). Thus, UK chemical processing is still largely centred on Merseyside (within a triangle bordered by St Helens, Warrington and Widnes), Clydeside in central belt of Scotland, and Tyneside (in the North East of England between Newcastle and North Shields) [17].

reasonable approximation for sector BAT potential. Improvement potential of direct and primary energy becomes 41 PJ and 36 PJ respectively, or just over a third of demand. The associated reduction in GHG emissions from BPT and BAT options is consequently 1.8 MtCO2e (22%) and 2.3 MtCO2e (27%) respectively [28]. Due to the uncertainties associated with using European efficiencies and in the estimation of load factors, these improvement potentials should be treated as indicative. 4.2. Process substitution These include retrofit technologies, such as dividing wall columns and membranes for chemical separation. The options assessed in the present study embraced all those reviewed by Ren and his co-workers [32,49] in their thorough account of long-term opportunities for the petrochemicals sector. They range from the replacement of steam cracker plants with the existing state-of-the-art equipment to radically different biotechnologies presently in the R & D stage, such as bioethanol-to-ethylene. Many of the technologies would entail a large structural shift in the way petrochemicals are produced in the UK. Most of the replacements entail a substitution in feedstock as well as fuel, and therefore have significant energy and GHG emissions ramifications for upstream fuel processing industries. These processes were also analysed by Ren and his co-workers [32,49], but were not included in this assessment as they are outside the scope of the Chemicals baseline. Improvement potential of these replacements should not be considered in isolation, and to properly understand their potential would require an extension of the definition of industry to include the refineries sector. 4.3. Fuel switching

5. Petrochemicals production

The current modelling approach [15,16,28] separates a proportion of generating plant allocated to petrochemical production. This was estimated to account for a quarter of sector heat and power generation, as well as 30% of associated GHG emissions, based on three major sites [28]. The high proportion of emissions was caused by the necessary use of refinery by-product gases. Thus, a limit is set on the degree to which fuel may be switched to biomass. Steam crackers are likely to improve in efficiency going forward, and so it was assumed that additional surplus cracker fuel gas will be made available to CHP plant, thereby lowering the limit or restriction on fuel switching [28]. This particular conflict between improvement options may be circumvented by switching to alternative petrochemical processes. However, upstream refinery gases were assumed to remain unchanged as they will continue to produce such gases regardless [28]. Fuel switching is also an option for conventional generating plant, and is far less restricted in that application. Biomass fuel switching generally applies to coal and natural gas only. In the case of coal, the present hybrid model automatically prioritises biomass substitution before switching away from natural gas [28].

5.1. The petrochemical context The petrochemical technologies assessed here represent options for either substituting existing facilities with radically different process routes, or maintaining existing facilities and retrofitting carbon capture equipment. The representation of alternative processes in the present study (see Figs. 4 and 7) was largely informed from an assessment by Ren and his co-workers [32,49], whilst the application of CCS was mainly underpinned by analysis of Johansson et al. [52]. Both of these studies provide detailed techno-economic comparisons within a wide portfolio of improvement potential options. Many of these possibilities were modelled here with ancillary heat and power systems to meet the envisaged new demand [13,15,16]. The systems were treated as additional to existing onsite heat and power capacity in the sector, which are kept separate to avoid risk of double counting opportunities in the energy sector. However, the present hybrid model (see Fig. 6) allowed for downsizing of existing facilities to compensate for new generating capacity or, in the case of alternative processes, the option to exclude the additional generation capacity. Moreover, for each option examined, there was a choice as to whether the ancillary system should be targeted for retrofitting capture facilities to the petrochemical process. Although these features add flexibility to the model, it is not possible to accurately integrate new and existing systems without more detailed and reliable data on the energy balances of individual petrochemicals sites [28]. This assessment also excludes the processes for manufacturing alternative plastics, such as starch plastics and polylactic acid (PLA), that could partially replace conventional plastics applications (and therefore petrochemical steam cracking); again due principally to a lack of available technical and economic information sources [28].

4.4. Carbon capture and storage Opportunities for retrofitting post-combustion carbon capture and storage (CCS) facilities [50,51] onto existing steam crackers were assessed by Johansson et al. [52]. There is also significant opportunity to apply CCS to steam reformers [53,54]. This option is relatively cheap as the process already produces a pure stream of CO2. It was assumed that the application of CCS would have a 20% energy penalty and a direct capture efficiency of 85% [50]. The energy penalty and capture efficiency values were informed by the literature [50,51,55,56], and applied conservatively owing to the mix of fuels considered in the present UK study [28]. CCS was assumed to be applied at a generation capacity that represents petrochemical production; reflecting the application of CCS to steam crackers. A comparison of retrofit CCS options for existing steam crackers is illustrated in Fig. 8 [28]. Here energy and GHG emissions from petrochemical steam cracker retrofitted CCS are

5.2. Waste plastics utilisation It is possible to convert waste plastics, such as the polypropylene in plastic carrier bags, into naphtha and other oils. The method uses 10

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

Fig. 8. Energy and GHG emissions from petrochemical steam cracker (SC) retrofit CCS. [A –specific energy consumption (SEC) by energy type; B – corresponding GHG emissions, less those avoided by auto generation.] Source: Griffin [28].

route (which has been deployed in South Africa by Mossgas and Malaysia by Shell [49]). Nevertheless, the biomass and biomass/coal blend options via FT processes were modelled in the present study [15,16,28] as they exhibit the greatest reduction in GHG emissions.

hydrogen, steam and other catalysts to produce the petrochemical feedstocks through a stepped process of liquefaction, pyrolysis and separation. Originally developed by BASF in Germany, the technology was abandoned before reaching commercial scale [49]. According to an estimation by WRAP [58] - a registered UK charity and company limited by guarantee delivering waste and recycling assessments for governments and international bodies - plastic waste arisings in the UK were about 3 Mt in 2010. Of this, around 2 Mt was packaging waste that is not collected for recycling. However, not all of this waste would potentially be available as feedstock. The economics of the processed waste and feedstock together, relative to oil prices, would dictate the likely level of substitution. The present hybrid model [15,16,28] enables any substitution rate, though a diversion of 10% of the available waste stream (200 kt) was assumed. Naphtha consumption for steam cracking in 2010 was about 1 Mt, which equates to a substitution rate of 20% [28].

5.4. Methanol routes Methanol provides an alternative route for converting methane, coal or biomass into olefins. The methanol-to-olefins (MTO) route comprises of three steps: methanol production, methanol conversion to olefins and gasoline, and product recovery and separation [28,49]. Methanol is produced from methane, coal and biomass by gasification at elevated temperatures into a syngas, and subsequently converted to methanol via synthesis processes. Heat from methanol synthesis may be utilised to convert some of the methanol into dimethyl-ether (DME) and water; the former being used in the synthesis of olefins using a fluidised or fixed-bed reactor. Recovery, separation and cooling processes are essentially unchanged from those used for steam cracking with the exact yield depending on the severity, catalysts used, and reactor configuration [49]. Research into MTO techniques began 20–30 years ago and two pilot plants currently operate in Norway [49]. The MTO process was modelled in the present study [15,16,28] via biomass gasification, plus coal gasification installed with CCS capture equipment.

5.3. Naphtha routes Naphtha can be produced from methane (natural gas), coal or biomass in various processes, including the Fischer-Tropsch (FT) process. FT naphtha is converted from methane via natural ‘gas-to-liquids’ processes or from coal via ‘indirect liquefaction’ [28,49]. Biomass may similarly be converted to naphtha via Fischer-Tropsch processes, although its efficiency is influenced by the relatively high moisture content. Steam cracking with FT naphtha typically produces a 40% higher ethylene yield than conventional naphtha. But, owing to an absence of aromatics, this produces only a 5% greater high value chemical (hvc) yield. In addition, coal and lignocellulosic biomass may be processed together for added flexibility. Alternatively, coal may instead be used in the process of ‘direct liquefaction’, although this produces aromatic-rich naphtha with a 15% lower hvc yield [28,49]. All these routes have yet to reach full commercialisation, except the methane

5.5. Oxidative coupling Oxidative coupling is a process by which ethylene can be converted directly from methane without the intermediate steps of methanol, naphtha, or ethane production. Oxidative coupling of methane (OCM) is also referred to as ‘catalytic oxidative dimerization’ of methane or partial oxidation of methane to ethylene. A fluidised bed reactor (FBR) is a common reactor design used for this process. Oxygen reacts with methane inside the reactor in the presence of a catalyst to form a methyl 11

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

Fig. 9. Distribution of CO2 point sources and CCS cluster regions in the UK. Source: adapted from Griffin et al. [13].

waste, and subsequent fermentation to ethanol); and lignocellulosic biomass gasification with microbial fermentation or chemical conversion with a catalyst. Detailed analysis on bioethanol production processes as part of a portfolio of ‘white’ biotechnology opportunities was undertaken by Patel et al. [62]. Bioethanol production via lignocellulosic biomass is still at the development stage, and has yet to have been demonstrated at an industrial scale. The production of bioethanol from sugar and maize already exists in production facilities worldwide, and a 0.2 Mt/yr ethanol-to-ethylene (ETE) production plant based on sugarcane began operation in Brazil in 2010 [28]. Installed and planned annual bioethanol capacity in the UK is nearly 1 Mt with 90% derived from wheat, in contrast to 7% from sugar beet, and 3% from municipal solid waste [28]. The UK does not presently have ETE capacity, and predominantly manufactures bioethanol from wheat (together with some sugar beet). The feedstocks reported in the literature consulted for the present work were principally maize and sugarcane [49,62]. As the choice of feedstock can influence energy and emissions intensity [28], these representations should be treated as only indicative for the case of the UK. ETE based on bioethanol from starch, sugar, and lignocellulosic biomass was modelled in the present study [15,16,28]. In the case of lignocellulosic biomass, the option of utilising more of the resource in CHP was also considered.

radical (CH3) and water [49]. By combining together, the methyl radicals form ethane, which subsequently dehydrogenates into ethylene. The catalysts originally used were oxides of alkali, alkaline earth materials, or other precious earth metals, which are needed in part to control the oxygen-ions and to maintain the reaction [49]. However, novel catalysts based on research into genetically modified (GM) bacteriophages developed at MIT are presently being advanced by a San Francisco based start-up [59]. One concern about this process is the difficulty in preventing the reaction from progressing and converting ethylene into CO2 and water. Nevertheless, a recent breakthrough in catalyst design [28] has yielded promising results in two pilot plants built in 2012, and a larger scale demonstration plant is likely to begin operation in the near future. The company has also partnered with German cracker equipment supplier in order to scale up the plant for commercialisation within the next few years [60]. The reaction takes place at two thirds the temperature of steam cracking and is exothermic. However, detailed recent performance data was not publically available for the present study, and therefore the technology modelled was based on the ‘OCM I’ design developed by Ren et al. [49] from the work of Swanenberg [61]. 5.6. Bioethanol routes

6. Future prospects: towards a bio-economy

It is possible to convert biomass into ethanol for subsequent dehydration to form ethylene. There are three well-known methods for producing ethanol from renewable sources [49]: direct fermentation of starch/sugar rich biomass (e.g., maize starch, sugar beet or sugar cane); hydrolysis of lignocellulosic biomass (e.g., wheat, wood or agricultural

Bioenergy can be produced from either biomass (any purpose-grown material, such as crops forestry or algae) or biogenic waste (including household, food and commercial waste, agricultural or forestry waste, 12

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

technological development.

and sewage sludge). Sustainable bioenergy is a renewable resource, used together with flexible technologies, that is often low-carbon and potentially generates ‘negative emissions’ when coupled to CCS facilities. However, virtually no bioenergy is currently used in the chemicals sector, except for the production of bio-hydrogen. The UK Government’s UK and Global Bioenergy Resource Model (an updated feedstock availability model; awaiting open publication at the time of writing) suggests that there is substantial quantities of indigenous biomass and biogenic waste available even accounting for the application of more stringent sustainability and land use criteria; perhaps equivalent to some 1200 PJ. But many industrial sectors will be competing for this resource alongside, for example, power generation. Indeed, in their stakeholder engagement with the UK Government, representatives of the chemicals industry identified what they regarded as a high level of uncertainty over the supply of low-carbon bioenergy. A number of the chemical processes (e.g., ammonia production) presently utilise very large quantities of natural gas feedstock. In order to replace it, this would need the replacement by major gasification capacity, together with associated bioenergy supply. The sector believes that this is unrealistic in view of the cost and possibly restricted availability of feedstock. The focus of biomass and biogenic waste gasification is therefore likely to be on burning biogas, hydrogen (replacing steam reforming of natural gas) and syngas going forward. However, burners designed for natural gas would need to be adjusted as gaseous bioenergy has a lower calorific value. The large number of smaller fossil-fuelled heaters scattered around a typical chemical factory would also make it difficult to distribute the correct fuel, or fuel mixture, to given burners. Pulverised biomass or biogenic waste could be adopted for high temperature process heat, but that would give rise to dust deposition on heater tubes and an increase in air pollution (possible beyond legal limits). Lower olefins primarily serve as the building blocks to polymers for the production of plastics, rubbers and fibres. For example, more than half of ethylene production is polymerised into polyethylene with most of the remainder forming precursors of other plastics, such as polyvinyl chloride (PVC), polystyrene, polyester, and so [28]. A lot of research attention has been given to the potential for substituting these products with bio-based polymers. There are three main approaches to producing bio-based polymers [28]: (i) modifying natural polymers (e.g., starch plastics); (ii) the manufacture biomass-derived monomers by fermentation or conventional methods and polymerise), or directly synthesise from micro-organisms; and (iii) via genetically modified crops. The UK appears to have the technical expertise and feedstock capabilities [63], as well as the downstream industrial demand, to develop a bio-based polymers manufacturing industry. The UK produces some 2.5 Mt of primary plastics, or resins, per year and nearly 5 Mt of converted plastics, such as packaging, downstream [28]. It is an importer of primary plastics and consumption to the tune of 5 Mt [28]. Domestic production of bio-based plastics is presently very low with no large-scale production facilities, although processors have indicated a willingness to expand their use of bio-based plastics and already import them [29]. A feasibility study commissioned by the National Non‐Food Crops Centre (NNFCC) suggests that a wheat-based 200 kt/yr polylactic acid plant could be commercially viable in the UK market. Britain also produces the feedstock for ethanol-to-ethylene (ETE), as indicated in Section 5.6 above, or bio-based polypropylene via its growing bioethanol sector, and is well placed to produce wheat starch for starch-based plastics [28,64]. However, the industry is presently in its infancy, although wheat supply could be increased through the utilisation of set-aside land that might yield an additional 2.1 Mt/yr [64]. In the long-term, a further 0.5 Mt/yr could be made available from the conversion of temporary grassland, whilst another 2 Mt/yr could yield higher intensity crop rotation [64]. Global production capacity for bio-based polymers is expected to grow substantially from 800 kt in 2012 and to over 5 Mt in 2017 [65]. Appropriate public support and incentives would be required in order to push forward

7. UK chemical ‘technology roadmaps’ to a low-carbon future by 2050 7.1. Background A set of technology roadmaps have been developed in order to evaluate for the potential deployment of the identified chemical technologies out to 2050. The extent of resource demand and GHG emissions reduction was therefore estimated and projected forward. Such roadmaps represent future projections that match short-term (say out to 2035) and long-term (2050) targets with specific technological solutions to help meet key energy saving and decarbonisation goals. A bottom-up technology roadmap approach has been adopted, based on that were previously used by Griffin et al. [13,66] to examine the impact of UK cement decarbonisation (for further details see Griffin [28]). Thus, their contents were built up on the basis of the improvement potentials associated with various processes employed in the chemicals industry and embedded in the UED [13,15,16]. 7.2. Baseline UK chemical technology projections The projected baseline is affected by sector output, grid decarbonisation, and deployment of BPT/BAT. It is assumed that the grid will decarbonise by around 85% over the period 2010–2050. Owing to the high trade intensity of the sector and high dependency on fuel prices, it is impossible to project forward with high confidence the future trend in petrochemicals production. Investment is going into shale gas imports into Britain, as well as potential local shale gas extraction, that will be sent to the Grangemouth refinery complex in Scotland; in large part to produce high value chemicals (hvc) [4,63,67]. However, the extent to which shale gas will affect the market over the short-mediumterm is uncertain [4]. Based on planned capacity closures, and the 2010 load factor, lower olefin production has been an estimated 2.4 Mthvc in 2015 [28]. For simplicity, and in the absence of more detailed information, hvc production levels were assumed to remain at roughly this level into the future. Similarly, the production from other chemical products modelled in the present study [15,16,28] were assumed to remain at their 2010 levels. It has been presumed that progress towards the take-up of BAT will be gradual, and reach the 2010 BPT level by about 2050 [28]. Reductions in steam and electricity demand are thought not to lead to significantly reduced generation of heat and power in the chemicals sector, but rather to increase exports to the public distribution system. Efficiency improvements via CHP plant were not directly assessed here, due largely to uncertainty about the impact of fuel switching. General energy saving methods, such as improved motors, better boiler efficiency, and so on, were assumed to steadily reduce sector energy demand by 10% in 2050. This reduction was taken as being similar across fuel, steam, and electricity demand. 7.3. Scenario definition The identified improvement technologies for the UK were incorporated into a chemical technology roadmap framework through a series of scenarios. The baseline year for the framework was taken as 2010. Full details of the both the 2010 baseline and the BAT/BPT improvements can be found in the UKERC industrial UED [15,16]. Four future scenarios were devised in order to demonstrate this approach. The chemical industry has been active in the area of technology roadmapping, particularly at the global level via the IEA [1], and ideas from such roadmaps were drawn on in constructing some of the scenarios detailed below [13,28,66]:

• Low Action (LA). This scenario describes a path of only slight improvements. No further investment is made in additional process

13

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

Fig. 10. GHG emission pathways associated with the technology roadmaps for the UK chemicals sector (on a credit basis). Source: Griffin [28].

• • •

alternative lower olefin production routes, as modelled for example by Ren et al. [32], are excluded in order to avoid conflict with existing generation plant. That also provides a simpler basis for comparison. As electricity and heat are exported from the sector, it is useful to analyse the associated emissions by what would otherwise be emitted from public distribution systems. Due to commercial restrictions on data availability it was not possible to accurately determine the extent of sector autogeneration and heat trading over the earlier years. Instead these graphs are used as a way of comparing the future potential of the constructed roadmaps and their relationship with the wider UK system. Indirect emissions presented in Fig. 10 were estimated on a ‘credit basis’ [28]. Graphs B and C in Fig. 10 indicate that heat and electricity is exported from all pathways with emission factors calculated from their generation. Roadmaps with lower GHG emissions intensity of generation therefore have a lower emissions trajectory. The baseline only requires a 10% biomass demand for CHP, and so it benefits by exporting this heat and electricity to other users to whom the higher GHG emissions factor would elseways be attributed. Conversely, the 80% biomass CHP roadmaps show a narrower improvement on their counterparts and all CCS roadmaps, for which emission is captured from auxiliary

technology improvements, and efficiency is only improved incidentally through the replacement of retired plants. Reasonable Action (RA). All identified efficiency technologies are installed by 2025, and retired plants are replaced with best practice ones by 2030. Reasonable Action including CCS (RA-CCS). This scenario is based on RA, but includes CCS. Biomass co-firing with CCS may, of course, mitigate upstream emissions on a full life-cycle basis, due to potential ‘negative emissions’ [68]; something that will need careful study in future studies. Radical Transition (RT). This scenario explores a boosted or radical version of the reasonable action (without CCS) scenario [28].

7.4. Alternative UK chemical technology roadmaps The chemicals sector GHG emission pathways are illustrated in Fig. 10 (on a ‘credit basis’ [28]) for the assessed roadmaps and the projected baseline. Graph A includes direct emissions (scope 1), whereas Graph B includes direct and indirect emissions (scope 1–2/3). Graph C includes GHG emissions from the combustion of lower olefins (on top of Graph B emissions). Additional autogeneration plant for 14

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al.

the emissions avoided through switching to biomass or biogenic waste derived feedstock. Depending on the scope of analysis, carbon fixation from biomass growth may be subtracted, or GHG emissions from endproduct incineration may be assumed zero. In order to reduce emissions via waste recycling, the waste stream must be diverted from degradation in landfill or incineration. The prospect of applying CCS [50–57] is reasonable compared with other industrial sectors, such as iron and steel [28], and cement [13,28,66]. All steam crackers and ammonia plant are situated within CCS cluster regions (see again Fig. 6) [13,28]. Pipeline technology for building a CO2 transport network is ready to be rolled out, and the UK already has preliminary plans for at least two large transport ‘hubs’ that will eventually be centrally located amongst a cluster of CCS power stations and industrial sites. The cost of capture from ammonia is also very low, because of the purity of its process CO2 emission. Retrofitting post-combustion capture at steam cracker furnaces has been estimated to cost in the range of £28–52/tCO2 [52]. This range covers the options identified in the study with the lower limit being comparable to oxyfuel capture technology and attained through the utilisation of process waste heat. Unlike oxyfuel capture, however, post-combustion would not require major process modification [53].

generation plant producing surplus electricity, appear less attractive. An indication of the relative degree of decarbonisation of the grid and district heat systems may also be deduced. Biomass demand highlights the greater levels necessary for radical abatement under the Radical Transition (RT) set of roadmaps as shown in Fig. 10. RT3 gives a similar abatement to RT5 [bio-CHP], but requires much larger amounts of biomass. RT5 would be a far more effective way to utilise this level of consumption. Thus, Graphs A and B show that the RT roadmaps abate less than the Reasonable Action including CCS (RA-CCS) roadmaps; especially so on a credit basis [28]. However in Graph C, which includes GHG emissions from the combustion of hvc, the RT roadmaps which use biomass as a feedstock display significantly improved, relative performance. In particular, RT3 (lingo FT naphtha) and RT5 (bio-ETE) abate more than the CCS roadmaps in Fig. 10. The use of biogenic wastederived naphtha, which is limited to replacing 20% of naphtha feedstock, shows only a marginal net improvement in Graph C. The trajectory of the petrochemicals sector were determined against the higher and lower carbon caps of the EU-ETS (although they are not shown here – the interested reader should consult Griffin [28]). Because the scope of the petrochemicals sector must include feedstock production to enable comparison between lower olefin production routes, caps are proportionally adjusted up from the GHG emissions verified for steam cracking. No constraint specific to this scope of emissions exists under the EU-ETS as the allocations are necessarily site-based. However, this study has provided an indication of the potential for meeting the cap. Direct emissions, as indicated in Fig. 10, do not fully account for the whole picture. It can therefore be deduced that the EU-ETS unfairly incentivises the use of retrofit CCS over bio-based process routes [28]. This is because CCS specifically addresses emissions at the point of installation.

Acknowledgements This is an extended and updated version of a paper originally presented at the 8th International Conference on Applied Energy (ICAE2016) held in Beijing, China over the period 8–11 October 2016 (denoted then as paper ICAE2016-560). The work reported forms part of a programme of research at the University of Bath on the technology assessment of energy systems and transition pathways towards a low-carbon future that has been supported by a series of UK research grants and contracts awarded by various bodies associated with the Research Councils UK (RCUK) Energy Programme for which the second author (GPH) was the holder. That associated with industrial energy demand and carbon emissions reduction originally formed a part of the ‘core’ research programme of the UK Energy Research Centre (UKERC); Phase 2, 2009–2014 [under Grant NE/G007748/1]. The first author (PWG) and third author (JBN) undertook their contributions to the present work as part of a UKERC flexible funding project entitled ‘Industrial Energy Use from a Bottom-up Perspective’ [for which the second author (GPH) was the Principal Investigator]. The second author (GPH) was also a CoInvestigator of the UK Biotechnology and Biological Sciences Research Council’s (BBSRC) Sustainable Bioenergy Centre (BSBEC) during 2009–2013, as part of the ‘Lignocellulosic Conversion to Ethanol’ (LACE) Programme [under Grant Ref: BB/G01616X/1]. During the preparation of this paper, the second (GPH) and third (JBN) authors continued to work in the field of industrial energy use and carbon emissions reduction; supported by the UK Engineering and Physical Sciences Research Council (EPSRC) ‘End Use Energy Demand’ (EUED) Programme, as part of the Centre for Industrial Energy, Materials and Products (CIE-MAP) [under Grant EP/N022645/1], as a Co-Director and Research Fellow respectively. The authors' names are listed alphabetically.

8. Concluding remarks The opportunities and challenges to reducing industrial energy demand and carbon dioxide (CO2) emissions in the Chemicals sector have been evaluated with a focus is on the situation in the United Kingdom (UK), although the lessons learned are applicable across much of the industrialised world. This sector can be characterised as being quite heterogeneous, and as sitting on the boundary between energy-intensive (EI) and non-energy-intensive (NEI) industrial sectors (see again Fig. 2). Currently-available technologies will lead to further, short-term energy and CO2 emissions savings in chemicals processing, but the prospects for the commercial exploitation of innovative technologies by mid-21st century are far more speculative. There are a number of nontechnological barriers to the take-up of such technologies going forward. Consequently, the transition pathways to a low-carbon future in the UK chemicals industry by 2050 will exhibit large uncertainties. The attainment of significant falls in carbon emissions over this period depends critically on the adoption of a limited number of key technologies [e.g., carbon capture and storage (CCS), energy efficiency {including combined heat and power (CHP)} techniques, and bioenergy], alongside a decarbonisation of the electricity supply [13]. Thus, this technology assessment and associated roadmaps help identify the steps needed to be made by industrialists, policy makers and other stakeholders in order to ensure the decarbonisation of the UK chemicals sector. The chemicals sector has long been the largest owner of generating plant in UK industry. Most generation is from CHP plant with significant amounts of excess electricity exported to the grid or other industrial sectors. Special care should be taken not to ‘double count’ autogeneration and grid decarbonisation, and the relative contributions to decarbonisations of each should be accounted for separately. Replacement process technologies (e.g., bio-processing and methanolto-olefins) cannot be properly compared without consideration of upstream feedstock processing. This is because they replace both the stream cracking process and production of petroleum feedstocks, e.g., naphtha and ethane. Special care should also be taken when modelling

References [1] International Energy Agency (IEA), International Council of Chemical Associations, DECHEMA. Technology roadmap: energy and GHG reductions in the chemical industry via catalytic processes. Paris (France): IEA Publications; 2013. [2] Jacquelin L-M, de Bucy J, Caujolle A. Facts & figures: energy and carbon for green chemistry. Paris, France: ENEA Consulting; 2015. [3] International Energy Agency [IEA]. World energy outlook: are we entering a golden age of gas? Special report. Paris (France): IEA; 2011. [4] Hammond GP, O’Grady A. Indicative energy technology assessment of UK shale gas extraction. Appl Energy 2017;185(2):1907–18. [5] Saygin D, Patel MK, Tam C, Gielen DJ. Chemical and petrochemical sector: potential of best practice technology and other measures for improving energy efficiency. Paris (France): OECD/IEA; 2009.

15

Applied Energy xxx (xxxx) xxx–xxx

P.W. Griffin et al. [6] International Council of Chemical Associations (ICCA). Innovating for greenhouse gas emission reductions: a life cycle quantification of carbon abatement solutions enabled by the chemical industry. Brussels (Belgium): ICCA; 2009. [7] Boulamanti A, Moya JA. Energy efficiency and GHG emissions: prospective scenarios for the chemical and petrochemical industry. JRC Science for Policy Report (JRC105767/EUR 28471 EN). Luxembourg: Publications Office of the European Union; 2017. [8] The European Chemical Industry Council [Cefic]. European chemistry for growth: unlocking a competitive, low carbon and energy efficient future. Brussels (Belgium): Cefic; 2013. [9] European Climate Foundation [ECF]. Europe’s Low-carbon Transition: Understanding the challenges and opportunities for the chemical sector. Brussels (Belgium): ECF; 2014. [10] Norman JB. Industrial energy use and improvement potential PhD Thesis Bath (UK): University of Bath; 2013 [11] Dyer CH, Hammond GP. C.I. Jones CI, McKenna RC. Enabling technologies for industrial energy demand management. Energy Policy 2008;36(12):4434–43. [12] Hammond GP. Industrial energy analysis, thermodynamics and sustainability (In memoriam: Willem van Gool). Appl Energy 2007;84(7–8):675–700. [13] Griffin PW, Hammond GP, Norman JB. Industrial energy use and carbon emissions reduction: a UK perspective. WIREs Energy Environ 2016;5(6):684–714. [14] Hammond GP, Norman JB. Decomposition analysis of energy-related carbon emissions from UK manufacturing. Energy 2012;41(1):220–7. [15] Griffin P, Hammond G, Norman J. Industrial energy use from a bottom-up perspective: developing the usable energy database (Beta Version). Report UKERC/ WP/ED/2013/002, London (UK): UK Energy Research Centre; 2013 < http://data. ukedc.rl.ac.uk/cgi-bin/dataset_catalogue//view.cgi.py?id=15 > [accessed 20 October 2014]. [16] Griffin P, Hammond G, Norman J. Industrial energy use from a bottom-up perspective: usable energy database (spreadsheet - beta version). London (UK): UK Energy Research Centre; 2013 < http://data.ukedc.rl.ac.uk/cgi-bin/dataset_ catalogue//view.cgi.py?id=15 > [accessed 20 October 2014]. [17] Buchanan RA. Industrial archaeology of Britain. Harmondsworth, UK: Penguin Books; 1972. [18] Butt J, Donnachie I. Industrial archaeology of the British Isles. London, UK: Paul Elek; 1979. [19] Derry TA, Williams TI. A short history of technology: from earliest times to A.D. 1900. London: Oxford University Press; 1960. [20] Williams TI. The chemical industry: past and present. 2nd ed. Wakefield: Ep Publishing; 1972. [21] Al-Hassani STS, Woodcock E, Saoud R, editors. 1001 inventions: muslim heritage in our world. 2nd ed.Manchester (UK): Foundation for Science Technology and Civilisation; 2005. [22] Al-Khalli J. Pathfinders: the golden age of Arabic science. London, UK: Allen Lane; 2010. [23] Cossons NBP. The book of industrial archaeology. 2nd ed. Newton Abbot (UK): David & Charles; 1987. [24] Heaton CA, editor. An introduction to industrial chemistry. 2nd ed.London (UK): Blackie Academic & Professional; 1991. [25] Royal Society of Chemistry [RSC]. Industrial chemistry case studies. London (UK): RSC; 1998. [26] Department of Business, Energy and Industrial Strategy [BEIS]. Digest of United Kingdom energy statistics. London (UK): BEIS; 2016 [annual]. [27] International Energy Agency [IEA]. Chemical and petrochemical sector: potential of best practice technology and other measures for improving energy efficiency. Paris (France): IEA; 2009. [28] Griffin PW. Radical change in energy intensive UK industry PhD Thesis Bath (UK): University of Bath; 2013 [29] Office for National Statistics [ONS]. Energy use by industry - source and fuel 1990–2012. London (UK): ONS; 2014 [Spreadsheet]. [30] Webb N, Broomfield M, Cardenas L, MacCarthy J, Murrells T, Pang Y, et al. UK greenhouse gas inventory 1990 to 2011: annual report for submission under the framework convention on climate change. Report Ricardo-AEA/R/3355. Didcot (UK): Ricardo-AEA; 2012. [31] Neelis M, Patel M, De Feber M. Improvement of CO2 emission estimates from the non-energy use of fossil fuels in the Netherlands. Utrecht (The Netherlands): Department of Science, Technology and Society; Universiteit Utrecht; 2003. [32] Ren T, Patel M, Blok K. Olefins from conventional and heavy feedstocks: energy use in steam cracking and alternative processes. Energy 2006;31(4):425–51. [33] Worrell E, Price L, Neelis M, Galitsky C, Zhou N. World best practice energy intensity values for selected industrial sectors Report LBNL-62806 Berkeley (CA), UK: Lawrence Berkeley Laboratory; 2008 [34] Department for Energy and Climate Change [DECC]. GHG inventory research: nonenergy use of fuels. London (UK): HMSO; 2007. [35] European Commission [EC]. Integrated pollution prevention and control (IPPC): reference document on best available techniques for the manufacture of large volume inorganic chemicals - ammonia, acids fertilisers. Seville (Spain): EC; 2007. [36] Wiesenberger H. State-of-the-art for the production of nitric acid with regard to the IPPC directive. Monograph 150 (M-150). Vienna (Austria): Umweltbundesamt GmbH (Federal Environment Agency Ltd); 2001. [37] Brinkmann T, Giner Santonja G, Schorcht F, Roudier S, Delgado Sancho L. Best

[38] [39] [40] [41]

[42]

[43]

[44] [45] [46] [47] [48]

[49] [50]

[51] [52] [53] [54]

[55]

[56] [57]

[58]

[59]

[60] [61]

[62]

[63] [64] [65] [66] [67]

[68]

16

available techniques (BAT) reference document for the production of chlor-alkali. JRC Science for Policy Report (JRC91156/ EUR 26844 EN). Luxembourg: Publications Office of the European Union; 2014. World Bank Group. Chlor-alkali plants. Pollution prevention abatement handbook. Washington (DC, USA): World Bank Group; 1998. p. 279–328. Ayers R. The life-cycle of chlorine, part I: chlorine production and the chlorinemercury connection. J Indust Ecol 1997;1(1):81–94. Ayres RU, Ayres LW. The life cycle of chlorine, part II: conversion processes and use in the European chemical industry. J Indust Ecol 1997;1(2):65–89. Moussallem I, Jörissen J, Kunz U, Pinnow S, Turek T. Chlor-alkali electrolysis with oxygen depolarized cathodes: history, present status and future prospects. J Appl Electrochem 2008;38(9):1177–94. Dyer CH, Hammond GP, McKenna RC. Engineering sustainability: energy efficiency, thermodynamic analysis and the industrial sector. Environ Eng (J Sustain Environ Eng, Inst Engin, Australia) 2008;9(2):17–22. Association of Petrochemicals Producers in Europe [APPE]. Capacity and production data (online) < http://www.petrochemistry.eu/about-petrochemistry/factsand-figures/capacity-and-production-data.html > [accessed 5 August 2013]. Meyers R. Handbook of petrochemicals production processes. New York (USA): McGraw-Hill Education; 2004. US Geological Survey [USGS]. Mineral commodities summaries: nitrogen (fixed) ammonia. Reston (VA, USA): USGS; 2012. Office of National Statistics [ONS]. Annual business survey: section C – manufacturing. Newport (UK): ONS; 2012. International Iron and Steel Institute [IISI]. Energy use in the steel industry. Brussels (Belgium): IISI; 1998. European Commission [EC]. Integrated pollution prevention and control (IPPC): best available techniques (BAT) reference document for iron and steel production. Seville (Spain): EC; 2012. Ren T, Patel MK. Basic petrochemicals from natural gas, coal and biomass: energy use and CO2 emissions. Resour, Conserv Recycl 2009;53(9):513–28. Hammond GP, Ondo Akwe SS, Williams S. Techno-economic appraisal of fossilfuelled power generation systems with carbon dioxide capture and storage. Energy 2011;36(2):975–84. Hammond GP, Spargo J. The prospects for coal-fired power plants with carbon capture and storage: a UK perspective. Energy Convers Manage 2014;86:476–89. Johansson D, Sjöblom J, Berntsson T. Heat supply alternatives for CO2 capture in the process industry. Int J Greenh Gas Control 2012;8:217–32. Collodi G, Wheeler F. Hydrogen production via steam reforming with CO2 capture. Chem Eng Trans 2010;19:37–42. Element Energy. Potential for the application of CCS to UK industry and natural gas power generation [Final Report for the UK Committee on Climate Change]. Cambridge (UK): Element Energy; 2010. Intergovernmental Panel on Climate Change [IPCC]. IPCC special report on carbon dioxide and storage. Cambridge: Cambridge University Press; 2005 [Prepared by Working Group III of the IPCC]. International Energy Agency [IEA]. Cost and performance of carbon dioxide capture from power generation. Paris (France): IEA; 2011. Department of Energy and Climate Change [DECC]. 2012. CCS roadmap: supporting deployment of carbon capture and storage in the UK. London (UK): DECC; 2012. Waste & Resources Action Programme [WRAP]. UK plastics waste – a review of supplies for recycling, global market demand, future trends and associated risks. Banbury (UK): WRAP; 2006. Bullis K. Chasing the dream of half-price gasoline from natural gas. MIT Technol Rev (online) 15 January 2014 < https://www.technologyreview.com/s/523146/ chasing-the-dream-of-half-price-gasoline-from-natural-gas/ > [accessed 15 January 2017]. Tullo AH. Siluria’s oxidative coupling nears reality. Chem Eng News 2014;92(27):20–1. Swanenberg GMJM. Cogeneration of ethylene and electricity with oxidative methane coupling: study of technical feasibility and economic merit. Delft (The Netherlands): Stan Ackermans Instituut; 1998. Patel M, Crank M, Dornburg V, Hermann B, Roes L, Husing B, et al. Medium and long-term opportunities and risks of the biotechnological production of bulk chemicals from renewable resources - the potential of white biotechnology The BREW Project Utrecht (Netherlands): Utrecht University; 2006 The Knowledge Transfer Network [KTN]. From shale gas to biomass: the future of chemical feedstocks. Horsham (UK): KTN; 2016. Barker M, Safford R. Industrial uses for crops: markets for bioplastics. London (UK): HGCA; 2009. European Bioplastics. Bioplastics: facts and figures. Berlin (Germany): European Bioplastics; 2013. Griffin PW, Hammond GP, Norman JB. Prospects for emissions reduction in the UK cement sector. Proc Inst Civil Eng: Energy 2014;167(3):152–61. Hammond GP, O’Grady A. The life cycle greenhouse gas implications of a UK gas supply transformation on a future low carbon electricity sector. Energy 2017;118:937–49. Kruger T, Darton R. Negative emissions technologies could become the world’s largest industry. Proc Inst Civil Eng: Civil Eng 2013;166(2):51.