Deep decarbonisation pathways for Scottish industries: research report

4 Decarbonisation options

The three main options for emissions abatement that are generally applicable across multiple industrial subsectors are energy efficiency measures, fuel switching, and carbon capture, utilisation, and storage (CCUS). Each of the three is discussed in detail within the next three sections, which cover the specific options considered within each high-level category, their abatement potential, and the key enablers and barriers relevant to their implementation. Specific uptake assumptions for each decarbonisation option are presented in Section 5.3.

4.1 Energy efficiency measures

4.1.1 Options considered

The first category of decarbonisation options considered here is that of 'energy efficiency measures', an umbrella term which refers to all improvements that enable an industrial site to reduce emissions by lowering the amount of energy and fuel used per unit of output. A comprehensive list of energy efficiency measures was determined through a literature review for each industrial sector. Following feedback from industry stakeholders, it was deemed appropriate to group these according to the level of disruption that their implementation implies:^[51]

• Incremental improvements, including measures such as waste-heat recovery,^[52] energy-use optimisation, improvement in the generation of steam, and other (relatively) non-disruptive measures listed in Table 3 that can generally be implemented without having to interrupt operations for prolonged periods.
• Major overhauls, which includes the implementation of state-of-the-art processes or Best Available Techniques.^[53] Implementation of these projects implies a significant rebuild of the industrial sites affected and/or the overhaul of large portions of the site setup, and extended downtime can be expected for their implementation.

It is also noted that the distinction between the categories above can be blurry, since even the implementation of measures here classified as incremental improvements can sometimes be highly disruptive. Nonetheless, the distinction proposed here holds approximately and serves as a useful proxy for the likelihood that a given efficiency measure is implemented (all else being equal). Reflecting this distinction, only the first of the decarbonisation pathways assessed in this study – i.e. the Efficiency pathway, presented in Section 5.2.1, which assesses the maximum abatement that can be attained via full implementation of all energy efficiency measures – includes the implementation of major overhauls to improve energy efficiency. By contrast, the other pathways only see deployment of incremental improvements in energy efficiency (next to fuel switching and CCUS, as outlined in Sections 5.2.2 and 5.2.3).

Table 3 – Energy efficiency measures and their abatement potential

Sector (key sources)[54]	Abatement potential		Key decarbonisation options[55]
	Incremental improvements only	All measures considered
Chemicals and pharmaceuticals (1, 2, 3, 4, 5)	5%	5%	Incremental measures: improved heat recovery and reuse. Note that the high level of heat integration already present and the substantial use of internal fuels limit the improvement potential.
Oil and gas (2, 5)	10%	20%	Incremental measures: waste heat and energy recovery; advanced control and improved monitoring. High level of heat integration already present limits improvement potential. Major overhauls: crude unit upgrades; design improvements; integration of crude and vacuum units.
Cement (2, 5, 6, 7)	<5%	<5%	The dry kiln with pre-heaters and pre-calciner installed already constitutes state-of-the-art technology, hence the limited residual improvement potential.
Food and drink (5, 8)	31%	31%	Incremental measures: energy management; good maintenance practice; improvements to steam production, distribution, and end-use; waste heat recovery. Note that several of the large distilleries included in the NAEI already meet a high proportion of their energy needs via CHP plants fuelled by biogas produced on-site via anaerobic digestion.
Iron, steel and aluminium (2, 5)	5%	5%	Incremental measures: waste heat recovery. Already known to be used in some cases. Residual heat tends to be of low temperature and limited value.
Paper and pulp (2, 5)	15%	15%	Incremental measures: improved energy management; focus on maintenance; improved process control; heat recovery. A high proportion of heat requirement for this sector is already met with CHP plants or biomass combustion.
Glass (2, 9, 10)	15%	35%	Incremental measures: waste heat recovery. Major overhauls: Improvements to furnace construction; oxy-fuel combustion.
Other EII (stakeholder interviews)	10%	10%	Incremental measures: Optimising heat use; waste heat recovery. Note that a high proportion of heat requirements already met via biomass, largely in the wood panels industry, which represents 70% of the emissions from this sector.

In the case of the petrochemical industries, one way to increase efficiency would be to use of naphtha as feedstock to the steam cracking process instead of ethane and other gases. Feedback from industry stakeholders however highlighted that such a process change can hardly be classified as an energy efficiency measure due to its far-reaching implications: such feedstock switch would require a different supply chain, would result in different output products (e.g. a wider range of high-value chemicals, the exact composition of which depends on the specific feedstock), and could only be implemented through complete overhaul of the production plant, since the cracking process is fully integrated. For this reason, conversion to naphtha steam cracking would likely only be plausible in the context of a greenfield project, whereas conversion of existing plants could only realistically be expected if a strong strategic motivation exist beyond the desire to improve energy efficiency.

4.1.2 Abatement potential

To accurately estimate the abatement potential of the implementation of energy efficiency measures one would need to possess site-specific information on:

• The initial energy efficiency as measured by the amount of energy used per unit of finished product (e.g. MWh of input energy per tonne of output product).
• The emissions intensity of the energy used, which in turn depends on the source of energy (generally electricity or a fossil fuel).
• The potential reduction in energy demand (for a fixed quantity of finished product) that can be attained through the implementation of energy efficiency measures.

Unfortunately, this information is generally commercially sensitive and therefore not available in the public domain. Hence, a simplified approach was followed here whereby the average abatement potential from each category of efficiency measure was estimated for each industrial sector through literature review and later validated with stakeholders. This approach relies on estimations that abstract from the (unknown) efficiency levels present today within Scottish sites and will therefore not be precise in the quantification of the potential benefits. However, in the absence of site-specific data this approach provides a good indication of the overall decarbonisation potential of energy efficiency measures, especially considering that the picture emerging from this study is that they have a limited – though not negligible – role to play in the transition to net zero for Scotland's industrial sites.

4.1.3 Barriers and enablers

Obstacles to commercialisation

The primary obstacles to the uptake of energy efficiency measures are centred on the fact that many of these measures are not commercially viable, with the ones that are commercially viable having already been implemented, in many cases. Hence, economic drivers and incentives need to change to make any additional ones viable. In fact, whenever site operators increase energy efficiency they do not just reduce combustion emissions but also save on energy costs. Since energy generally represents a sizeable share of the manufacturing cost in energy-intensive industries, it is perhaps unsurprising that most of the industry stakeholders engaged for this study indicated that they have already implemented all commercially viable measures to increase energy efficiency. Some also mentioned that their companies had already signed Climate Change Agreements (CCAs) that compel them to achieve previously agreed targets in order to pay a reduced rate of the Climate Change Levy.^[56] This explains why the residual decarbonisation potential of energy efficiency measures may be limited unless the economic drivers substantially change, for instance if the cost of energy (or carbon) increases.

It is worth noting that 'commercial viability' – defined as the ability of a project to be backed by a business case that meets the relevant investment requirements (e.g. on the payback period) – is a relatively strict requirement compared to 'economic viability', which simply implies that the project's annualised costs are lower than the annualised savings. A previous study by Element Energy, Ecofys, and Imperial College London investigated the differences in the context of industrial waste-heat recovery, determining that just under half of the technically viable heat-recovery options were also commercially viable without further incentives.^[57] In the decarbonisation pathways assessed in this study it is assumed that sufficient incentives will be made available by government to justify investment in all decarbonisation measures, including energy efficiency.^[58]

Technical challenges

Certain energy efficiency measures might not be fully adopted due to their potential impact on the final product quality. In the Scotch Whisky industry, for instance, the character and flavour of the spirit produced may be affected by the implementation of measures such as thermal vapour recompression, which could reduce steam demand and hence energy consumption in the distillation process. Further work is required to assess the extent to which adoption might be affected.

Potential synergies with fuel switching

Fuel switching could strengthen the business rationale for investing in energy efficiency, sincethe cost of low-carbon hydrogen and electricity is expected to be higher than that of fossil fuels (see Section 4.2.3). This would increase the economic value and commercial viability of any measure that can reduce energy use, including some that may not be considered viable with current energy prices. For this reason, all decarbonisation pathways assessed below assume that all of the incremental improvements in energy efficiency which are considered technically viable today are implemented by 2045. Major overhauls are instead assumed to only be implemented in the first pathway, as noted above. This is because it is assumed that CCUS would be implemented instead of major overhauls in the other pathways, hence their additional implementation would not lead to further emissions abatement though it would still cause major disruption. Lastly, it is noted that, since it is possible that at least some of the efficiency measures will not be implemented, these assumptions may be considered optimistic. However, new ways to improve energy efficiency may also be discovered before 2045, which would increase the technical potential of energy efficiency.

4.2 Fuel switching

4.2.1 Options considered

This assessment primarily focuses on electrification and hydrogen fuel switching as potential substitutes to fossil fuels. Both 'green' and 'blue' hydrogen, collectively referred to as 'low-carbon hydrogen', are considered to be part of the energy mix.^[59] Hydrogen produced via biomass gasification coupled with CCUS is also considered an option to produce low-carbon (possibly carbon-negative) hydrogen. However, with a limited supply of sustainable biomass available, it may be the case that hydrogen from biomass would be unlikely to constitute a large share of the overall hydrogen supply in the long term, therefore this option is not explored further.

Switching to bioenergy is assessed for the cement industry, which already burns some biomass contained within waste-derived fuelsand isassumed to increase its use (while also deploying CCUS, which results in negative emissions – see Section 4.3.2).^[60] Other options for switching to bioenergy which may be considered by some Scottish industries but are not evaluated here are discussed in Box 2.

Key technologies that will need to undergo fuel switching include boilers, furnaces, driers and the other appliances indicated in Table 4. In a few selected cases the characterisation of the fuel-switching counterparts to existing fossil-fuelled appliances warrants further explanation:

• Following stakeholder consultation, the cement kiln is assumed to be converted to co-firing biomass (up to 70% of the heat demand), with the remaining heat generated by a mix of hydrogen combustion (20%) and electric plasma gas (10%).^[61] Due to the characteristics of the mixed-fuel kiln, fuel switching could take place in phases: biomass use could be maximised right away (provided suitable economic incentives existed), whereas the residual energy use could be switched later on when the relevant technologies become available. For simplicity it is conservatively assumed here that fuel switching occurs at the same time once all the required technologies become available.
• In the electrification pathway, the heat demand currently met via CHP plants is assumed to be replaced by a mix of electric steam boilers and heat pumps powered via the grid, whereas electricity produced with CHP plants is substituted by electricity from the grid.^[62]

Technology costs and efficiencies for each appliance were derived from publicly available sources and are also reported in Table 4. When considering these it is important to acknowledge the uncertainty that surrounds the technical and economic characteristics of all fuel-switching technologies, most of which have not yet been demonstrated in an operational environment.

Grid connection requirements

A further assumption employed in this study is that 1 MW of new grid connections is required for each megawatt of electrical appliance installed, at a cost of £350/kW.^[63] Since it is likely that existing connections would suffice to cover part of the additional power demand, the extent to which new grid connections are required is therefore overestimated through this approach, and future work could refine this assumption by assessing the local grid constraints on a site-by-site basis. It is also worth noting that no corresponding cost is computed for hydrogen fuel switching (e.g. for connection to a future hydrogen grid); rather, it is assumed that the cost of hydrogen fuel presented in Section 4.2.3 already includes all infrastructure-related costs.

Replacements vs retrofits

It should also be noted that, although it is here assumed that all fossil-fuelled appliances must be replaced, it is expected that, when switching to hydrogen, a portion of these would in practice be suitable for retrofitting, which represents a potential advantage of the Hydrogen pathway over the Electrification alternative (discussed in Section 6.5). An extensive discussion of retrofitting industrial natural gas appliances to hydrogen can be found in previous work by Element Energy, Advisian, and Cardiff University.^[64]

Box 2 – Fuel switching to bioenergy

Fuel switching to bioenergy is particularly relevant to industries that generate organic process residues, likethe food and drink, paper and pulp, and wood processing sectors. Organic residues are either combusted in their solid form or can be fed to anaerobic digesters to produce biogas, which can replace natural gas and hence directly reduce fossil CO₂ emissions. To evaluate the overall level of emissions abatement, however, one must also consider the relative carbon intensity of the fuel supply chains as well as the alternative potential uses of the biomass feedstock.

In the case of Scotch Whisky distilleries, for instance, several of them already produce biogas through the anaerobic digestion of draff – an organic residue from the mashing process – and/or pot ale – a viscous liquor which is left in the pot still after the first distillation stage in malt distilling.^[65] Analysis by Ricardo showed that using draff and pot ale in such systems can indeed deliver substantial climate benefits, especially when heavy fuel oils commonly used in remote distilleries are replaced (instead of natural gas).^[66] However, the same study also showed that if these biomass feedstocks are diverted from animal feed uses, the additional climate benefit of converting them to bioenergy would reduce significantly.

The approach followed in this study with respect to fuel switching to bioenergy is that recommended by the Committee on Climate Change,^[67] who only consider bioenergy in combination with CCUS (i.e. bioenergy CCS, or BECCS) or for sites where it is already in use.

Technology maturity and date of first deployment

The maturity of electrification technologies is generally higher than that of hydrogen technologies, as can be deduced by considering their technology readiness level (TRL, see Box 3), also reported in Table 4. Accordingly, the estimated date of first deployment of technologies in the former group generally occurs earlier than that of technologies in the latter group. To estimate this date it was assumed that the first deployment of each technology among the industries in scope would occur 2 years after a technology reaches TRL 9,^[68] which was in turn calculated starting from the technology's current TRL, assuming that it will take 1-3 years to progress from one TRL to the next.^[69] However, it is also assumed that the uptake of hydrogen technologies can proceed faster, since the process of substituting natural gas with hydrogen can be less disruptive than that of electrifying heat.^[70]

Box 3 – Technology readiness level: the TRL scale

The maturity of any technology can be approximately defined by the technology readiness level (TRL), defined by the European Commission as:^[71]

• TRL 1 – basic principles observed
• TRL 2 – technology concept formulated
• TRL 3 – experimental proof of concept
• TRL 4 – technology validated in lab
• TRL 5 – technology validated in industrially relevant environment
• TRL 6 – technology demonstrated in industrially relevant environment
• TRL 7 – system prototype demonstrated in operational environment
• TRL 8 – system complete and qualified
• TRL 9 – actual system proven in operational environment

Two additional constraints were set on the date of first deployment:

• No hydrogen technology is deployed before low-carbon hydrogen becomes available.^[72]
• Technologies that are already at TRL 9 today are not deployed until 2023 at the earliest since earlier deployment would likely be limited by infrastructure constraints and/or by the lack of economic incentives.

Given the uncertainty surrounding the timeline of development of each technology, the first deployment dates quoted in Table 4 represent an informed estimate based on current TRLs. Regardless of the exact year in which each technology is first deployed, all fuel-switching technologies are estimated to become available at the required scale by 2030 – although this expectation is conditional on the implementation of suitable policies and economic incentives to support the development and commercialisation of the required technologies.

Table 4 – Modelling assumptions on fuel-switching technologies

Fuel	Technology (source)[73]	Suitable for	Lifetime (years)	TRL	Date of first deployment	Capex (£/kW)	Opex (£/kW/y)	Efficiency
Electricity	Electric boiler (1)	Steam-driven processes	15	9	2023	120	4.0	95%
	Electric oven (1)[74]	Direct low-temperature heating	15	9	2023	120	2.4	95%
	Electric process heater (1)	Indirect low-temperature heating	15	9	2023	120	2.4	95%
	Electric plasma gas furnace (1)	High-temperature heating	15	5	2028	262	3.0	90%
	Heat pump (1)	Low-temperature heating (inc. steam)	20	8	2023	450	9.0	150-350%
	Grid connection (2)	See assumption on page 36.	N/A	9	2023	350	N/A	N/A
Hydrogen	Hydrogen boiler (1)	Steam-driven processes	25	7	2028	199	4.0	92%
	Hydrogen oven (1)	Direct low-temperature heating	15	5-6	2028	232	4.6	92%
	Hydrogen heater (1)	Direct low-temperature heating	25	5	2028	232	4.6	92%
	Hydrogen furnace (1)	High-temperature heating	25	4-5	2028-2030[75]	232	4.6	92%
	Hydrogen CHP (3)[76]	Replacing gas-fired CHP	25	7	2028	489	35.0	80%
Mixed fuel	Kiln co-firing 70% biomass (1)	Cement kiln	25	7-8	2025	83	1.3	84%
	Kiln co-firing 70% biomass, hydrogen, & plasma gas (1)[77]		25	4	2030	156	2.5	84%
Fossil fuels	Steam boiler (1)	Steam-driven processes	25	9	<2020	166	3.3	92%
	Oven (1)	Direct low-temperature heating	15	9	<2020	193	3.9	92%
	Dryer (1)	Direct low-temperature heating	25	9	<2020	193	3.9	92%
	Furnace (1)	High-temperature heating	25	9	<2020	193	3.9	92%
	CHP (3)	Steam- & electricity-driven processes	25	9	<2020	406	24.0	80%

4.2.2 Abatement potential

Hydrogen and electricity can be far less carbon intensive than fossil fuels and their use does not yield any scope 1 emissions.^[78] However, neither option is necessarily carbon neutral. Fuel use along other segments of the respective value chains (e.g. in electricity generation or hydrogen production), may reduce the net decarbonisation benefit of fuel switching. To avoid overstating the emissions abatement that can be achieved through electrification and hydrogen fuel switching, this study also accounts for the emissions associated with hydrogen production and electricity generation.^[79]

Grid carbon intensity

The carbon intensity of grid electricity is assumed to decrease linearly from 173 gCO₂e/kWh in 2017 to 41 gCO₂e/kWh in 2035,^[80] later reducing at a slower rate of 10% year-on-year and reaching an assumed 14 gCO₂e/kWh in 2045, as shown in the figure below. Notably, the average carbon intensity of the electricity generated in Scotland in 2018 was already 48gCO₂e/kWh,^[81] which is similar to the level expected for Great Britain as a whole in 2035. Hence, it could be argued that Scottish industries that use electricity generated in Scotland could decarbonise even more rapidly than is estimated here.

Hydrogen carbon intensity

Residual emissions from blue hydrogen depend on the rate of carbon capture and net energy efficiency of the hydrogen production process. Due to the high CO₂ purity of the flue gases from the steam methane reforming (SMR) and auto-thermal reforming (ATR) processes, it is expected that high carbon capture rates can be achieved regardless of which method is selected for hydrogen production. In this study we acknowledge that there are relative merits to the SMR or ATR routes but do not make any assumptions around which is preferable or more likely to be implemented. Regardless, blue hydrogen's embedded emissions are assumed to be 10% of those from natural gas in 2025, decreasing to 6% by 2050 following more efficient hydrogen production and increased capture rates.^[82]

Green hydrogen is assumed to be carbon neutral, which implies that hydrogen must be produced exclusively from renewable energy sources (either from excess renewable energy that would otherwise be curtailed, for instance due to network constraints, or from dedicated off-grid renewable generation).Further information on embedded, scope 3, emissions relating to the manufacturing of equipment for hydrogen production and renewable electricity generation can be found in previous work by E4Tech.^[83]

Although the carbon intensity of natural gas is assumed to remain constant (at 205 gCO₂e/kWh), it is possible that some hydrogen (or biomethane) could in the future be blended into the gas grid, which is expected to be able handle blends of up to 20% hydrogen by volume (or 8% by energy content) without significant challenges.^[84] Hydrogen injections into the gas grid could thus offer an additional way for reducing emissions from gas use.

Green and blue hydrogen mix

Considering that neither blue nor green hydrogen is commercially available today, the following assumptions were made around the sources of low-carbon hydrogen for industry:

• Within the Hydrogen pathway, the hydrogen used by the Scottish industries is assumed to be 90% blue and 10% green in 2028.^[85] Over time, the penetration of green hydrogen is assumed to increase, reaching a 45% share of the hydrogen market by 2045.
• Only green hydrogen is used in the Electrification pathway, where it is exclusively used within the cement industry or for CCUS.

Based on the assumptions discussed above, emissions from hydrogen use are estimated to be 17 gCO₂e/kWh in 2028 and 7 gCO₂e/kWh in 2045 in the Hydrogen pathway and 0 gCO₂e/kWh in the Electrification pathway. Given the fact that the carbon intensities of both hydrogen and electricity are expected to be below 10% of those of natural gas in the long term, it can be deduced that switching to either hydrogen or electricity can result in comparable emissions abatement, a feature which is evident in the results concerning emissions trajectories within each pathway presented in Section 6.1.

4.2.3 Barriers and enablers

Higher energy costs affecting competitiveness

Even though several of the fuel-switching technologies considered in this study have already been demonstrated, few are used in industry today due to the higher cost of low-carbon energy sources, compared to natural gas. Cost estimates for hydrogen range widely:

• The Hydrogen Council place the production cost^[86] of green and blue hydrogen at 12.7 p/kWh and 4.5 p/kWh, respectively.^[87]
• A review by Spears et al. finds production costs in the range of 4-9 p/kWh for green hydrogen and 2-5 p/kWh for blue hydrogen.^[88]
• A recent report on the first phase of the Gigastack project calculates a cost for green hydrogen in the range of 10.7-17.5 p/kWh, depending on the project configurations.^[89]

Considering that low-carbon hydrogen is still a pre-commercial fuel, it is generally expected that its cost will substantially reduce with time. By 2030, the cost of green and blue hydrogen production in Europe could reduce to 5.3 p/kWh and 3.8 p/kWh, respectively, according to the Hydrogen Council, and Bloomberg NEF predicts that the cost of green hydrogen delivered to industrial users (i.e. not just the production cost) could reduce to 4.2 p/kWh in 2030 and 2.1 p/kWh in 2050.^[90]

In light of the wide spread in the cost estimates reported above, it is assumed in this study that hydrogen will cost 5.0 p/kWh when first deployed in 2028, linearly reducing to 3.0 p/kWh by 2050, whereas no assumption is made regarding its cost pre-2028, hence the gap for the corresponding years in the figure below. While in practice the hydrogen price may be different for blue and green hydrogen, also due to their potentially different characteristics (e.g. purity), the price assumptions employed here refer to the mix of blue and green hydrogen described previously.

As for electricity and gas, the CCC calculates that large industrial users in the UK paid an average of 6.4 p/kWh and 1.6 p/kWh in 2016, respectively– values which are here assumed to apply throughout the timeline of interest.^[91] According to these energy price estimates – and before accounting for the cost of carbon – gas will remain cheaper than hydrogen and electricity through to 2050 at least. Hence, switching from natural gas to electricity or low-carbon hydrogen would cause a substantial increase in the cost of energy that would be hard to justify from a commercial perspective unless the economic incentives change.

It is of course possible that the price of the various energy forms will differ from what is assumed here, also considering that the publicly available sources from which the relevant energy price data were sourced do not clearly state their assumptions (e.g. around network costs, or on taxes). This could affect the business case for fuel switching. However, it is the difference between the price of gas, electricity, and hydrogen, rather than their absolute levels, that affects the cost of decarbonisation (see corresponding results in Section 6.3). Since the price of blue hydrogen is linked to that of natural gas^[92] and that of green hydrogen depends on the cost of renewable electricity generation, it can be argued that the different price levels are going to be tightly coupled in the long term, with any difference between them imputable to energy conversion technologies and taxation regimes. Hence, rather than developing complex energy price projections that would necessarily be subject to high uncertainty it was preferred to employ simple assumptions that enable a transparent comparison of energy costs in the different pathways.^[93]

If the cost of carbon is included, fuel switching could be justifiable on economic grounds alone. With a projected carbon price of £193/tCO₂ in 2045,^[94] the total cost of unabated natural gas (inclusive of the carbon cost) would approach 6 p/kWh – a value that is higher than that assumed for hydrogen and comparable to that assumed for electricity, as evidenced by the converging lines in Figure 8. Should the carbon price in the UK increase at the rate modelled in this study, a different concern would however arise; local industries facing higher energy costs than their counterparts abroad (who have lower carbon costs) might struggle retain competitiveness, an issue discussed in the concluding chapter. Finally, it should also be noted that new technologies for producing low-carbon hydrogen at cost lower than is assumed here may also emerge. Indeed, already today there are companies claiming to be able to offer hydrogen at less than 10% of the costs assumed here.^[95]

Availability of low carbon energy and infrastructure

As already noted, no low-carbon hydrogen is available in Scotland today at the scale required to support the decarbonisation of industry, although roadmaps and feasibility studies have been assessing ways to initiate this as indicated in Section 2.1. This poses a barrier to industrial decarbonisation via hydrogen fuel switching. The limited availability of low-carbon hydrogen is one of the main barriers to the deployment of hydrogen technologies. In this study it is assumed that low-carbon hydrogen becomes commercially available in 2028 (see Section Error! Reference source not found.).

A similar challenge affects industrial electrification efforts since the current capacity of the electricity network may not be sufficient to meet the increased demand for electrical power. Hence, electrification of large heating processes such as those in the petrochemical industry may only be possible after major upgrades to the grid infrastructure (and potentially large-scale deployment of energy storage).

Technical challenges

Several technologies considered in this section are not yet commercially available, and it is possible that some of them never will be. Early-stage technologies that are not yet technically mature are particularly susceptible to this risk, since they may encounter obstacles on their way to market.

A key challengethat may affect the technical – and therefore commercial – viability of certain fuel-switching technologies includes the strict requirement to meet specific heating profiles demanded by some industrial processes. This is especially a challenge when the heating profile directly impacts product quality, as is sometimes the case in the food and drink industry. Unfortunately, it is not possible to rule out the possibility that some of the fuel-switching technologies considered may turn out to be unsuitable because of this. It is however also possible that the increased societal focus on decarbonisation may accelerate the development of new technologies that were not considered by this study. Hence, despite the risk that some of the technologies may never be deployed, it is assumed that others will be used in their stead.^[96] It is worth mentioning that this challenge does not apply to indirect heating processes that make use of steam, which represent the largest single type of heat demand that is suitable for fuel switching (see Section 3.3).

Disruptiveness of switching and other operational challenges

Significant changes to the plant configuration may be required to make fuel switching possible, especially in the case of integrated processes (e.g. in petrochemical plants) where changes to the energy source cause other system impacts.

The core processes within each industrial subsector are often designed to operate uninterrupted to avoid costly down-time. Not only can interruptions lead to missed revenue due to reduced output, but they can also adversely impact CO₂ emissions (for instance in the case of event-related flaring, which is caused by unwanted interruptions). This is especially true for high-temperature processes (e.g. those in the petrochemical industry) that require precise control of the heat load and that may need time to be fully operational following an interruption – during this time, large amounts of energy and low-quality product can be wasted, causing unnecessary emissions and process waste. For these reasons it is common practice to have back-ups in place to guarantee continued operations in case of faults with the primary equipment, e.g. back up steam generators in case the primary boilers fail.

Pathways that rely extensively on electrification expose industrial users to the reliability risks facing the grid. To mitigate this risk, clean alternatives to back-up diesel generators could include battery storage or, if hydrogen is already used on site, hydrogen fuel cells. Back-up hydrogen equipment would similarly be required. Considering that back-up equipment generally has low load factors and consequently marginal impact on carbon emissions, investment to replace it with alternative back-up appliances would likely be deferred to a later date, when the cost of fuel-switching technologies may be similar to that of current natural gas appliances. For these reasons, the impact of back-up equipment on the overall additional cost of the deep decarbonisation pathways is considered negligible.^[97]

Space availability is also considered a challenge, for instance in the case of installations of heat pump systems at existing distilleries, where space constraints are likely to limit their uptake as noted by Ricardo.^[98]

Retrofitting opportunity for hydrogen technologies

It was already mentioned above that fossil-fuelled appliances may in some cases be retrofitted to operate with hydrogen (see Section 4.2.1). This represents a potential advantage for the Hydrogen pathway, since the possibility to retrofit appliances that may have a long residual lifetime without having to replace them could enable prompter uptake of hydrogen technologies. In this study it is conservatively assumed that appliances are instead replaced at the end of their useful life.

4.3 Carbon capture, utilisation, and storage

4.3.1 Options considered

Sites implementing carbon capture

A wide range of factors must be considered when developing a CCUS project, a few can help pre-select which sites are most likely to implement this on some or all of their emissions sources:

• The absolute amount of carbon that must be captured, important to reach sufficient economies of scale.
• Geographical factors, and in particular proximity to relevant CO₂ infrastructure for transport and storage, or to users of the captured CO₂.
• Availability of alternative pathways for deep decarbonisation.^[99]

Far from being an exhaustive list, the above provides a first set of criteria for deciding whether CCUS is likely to be relevant for a site. Based on these criteria, it was assumed that the following Scottish sites would be most likely to take part in CCUS projects before 2045:

• The refinery and petrochemical plant in Grangemouth, which make substantial use of internal fuels and also feature a source of process emissions (i.e. the refinery SMR).
• The Dunbar cement plant, which also features substantial process emissions.

The selected sites are the largest emitters among those which cannot be deeply decarbonised by fuel switching, whether because of internal fuel use or due to the presence of process emissions. Other large emitters in the area, like the Grangemouth CHP plants, are instead assumed to decarbonise via fuel switching. Other sites with unavoidable process emissions also exist (i.e. the aluminium smelter and the glass manufacturing sites), but these are considered too small and/or remote to justify CCUS implementation or, in the case of flaring, are deemed unsuitable for capture (see page 26). It is expected that process changes might be the more likely option to decarbonisation these emissions sources, as discussed in Section 6.1.4.

The selected sites are also located in an advantageous geographical area. Except for the Dunbar cement plant, the other sites are clustered around Grangemouth, where it is envisioned that the initial shared CO₂ capture infrastructure will be created. Later on, this infrastructure could also be extended to reach sites that are further away, like the cement plant itself. The timeline for the deployment of CCUS at these sites is discussed in Section 5.3.

Carbon capture and compression: technology and cost assumptions

Various technologies exist to extract CO₂ from a flue gas stream, differing in technology maturity, energy requirements (discussed in the next section), and ultimately cost. The most commonly used technology relies on amine scrubbing, in which CO₂-containing flue gas passes through vats containing amino compounds (i.e. amines), which absorb most of the CO₂. Considering that the accelerated emission targets place Scotland among the most ambitious countries in the pursuit of net zero, it is assumed that the Scottish industrial sites that implement carbon capture will be among the first of their kind to do so in a commercial setting. In light of the limited time available for capture technologies other than (first generation) amines to reach commercial maturity, it is assumed that amine-based technology will be adopted by all sites where CCUS is deployed. This is intended as a conservative assumption since the cost of capture would likely reduce if other technologies like advanced amines, calcium looping, or oxy-fuel combustion with carbon capture were used.^[100]

The complete set of assumptions underpinning the analysis of CCUS is reported in Appendix 8.7, but some of the key factors with a significant impact on the cost of capture include:

• The CO₂ concentration in the flue gases, which depends on the emission source. Specifically, it is easier and cheaper to capture CO₂ when it is not excessively diluted (the limit case is that where atmospheric CO₂ is captured).
• The capture rate, i.e. the proportion of CO₂ contained by the incoming gas stream which is captured (which also affects the abatement potential; a capture rate of 90% is assumed though higher rates are also possible, as discussed in the next section).^[101]
• The fuel used to meet the significant heat demand from the capture process, assumed to be low-carbon hydrogen, rather than natural gas, so as to enable the maximum emissions abatement (see discussion in the next section).
• The absolute emission level, which determines scale economies.
• The pressure to which the captured CO₂ must be compressed before it is transported. It is assumed the CO₂ is always captured at atmospheric pressure (0.11MPa) and must be compressed to 10MPa.

The cost of carbon capture and compression can be calculated by summing up the lifetime expenditures (capital, operational, and energy-related) and dividing the total by the cumulative amount of captured CO₂.^[102] This yields an average levelised cost of capture of around £100/tCO₂, about 60% of which is imputable to the cost of energy,and an average levelised cost of compression of around £5/tCO₂.^[103]It shouldbe noted that the net cost of abatement is slightly higher than that presented here due to the fact that the energy used for CCUS is not fully carbon neutral. In conclusion, Box 4 reviews two alternative ways for CCUS to tackle emissions from the combustion of internal fuel gases.

Box 4 – CCUS to tackle emissions from internal fuel combustion

There are two options for CCUS to abate emissions from internal fuel combustion in the refining and petrochemical subsectors:

• The first option sees post-combustion capture applied to the flue gases resulting from combustion of internal fuels.
• In the second option, internal fuel gases^[104] are diverted to a natural gas reformer (SMR or ATR) to produce blue hydrogen and capture is applied onto the flue gases from the reforming process. This hydrogen in turn replaces internal fuels in the furnaces leading to zero combustion emissions.

There are pros and cons to both options. The second has the advantage that carbon capture is only needed at the natural gas reformer. However, this route incurs a higher energy penalty due to the losses occurring when natural gas is converted to hydrogen, which would reduce the net decarbonisation benefit of CCUS. In this study the first option is assumed to be implemented. It is however acknowledged that, ultimately, the choice depends on the relative cost and on other strategic considerations. It is worth highlighting in this context that the substantial demand for blue hydrogen in hydrogen-centred decarbonisation pathways may make it preferable to accept higher conversion losses in light of the considerable economies of scale that would be achieved by only capturing CO ₂ from very large standardised reformers, which it is assumed would also be installed within the Grangemouth cluster.

Transport and storage

The captured CO₂ can be transported to its destination in multiple ways. Options for onshore transport include pipelines (e.g. the Feeder 10 pipeline which the Acorn project aims to convert to transport CO₂ from Grangemouth), trucks, and trains. Conversely, options for offshore transport include underwater pipelines or shipping. It was already noted that a site's distance from the CO₂ infrastructure (e.g. from CO₂ pipelines or from the storage site) and the characteristics of the storage site itself can have a marked impact on the cost of storage. A recent assessment of the UK's CO₂ storage resource carried out by Pale Blue Dot, Axis Well Technology and Costain indicates that the levelised cost for offshore transport and storage across different UK sites ranges between £11-18/tCO₂ stored,^[105] and the upper end of the range is selected for the modelling here. The additional cost of onshore transport (inclusive of the cost of transport in the Feeder 10 pipeline as well as any connection to it)^[106] is assumed to be:

• £5/tCO₂ for the Grangemouth sites, assumed to be connected via pipeline to the Feeder 10.
• £10/tCO₂ for the petrochemical plant in Fife, also assumed to be connected via pipeline (either to the Feeder 10 directly or to the Grangemouth infrastructure).
• £15/tCO₂ for the Dunbar cement plant, assumed to be connected to the Grangemouth infrastructure via truck or, considering that the site is already rail connected, via train.

The cost of CO₂ transport and storage is therefore assumed to range between £23/tCO₂ and £33/tCO₂, modelled as a fee (T&S fee) paid to third parties assumed to operate the CO₂ transport and storage infrastructure.

It is noted that the only scenario discussed so far is that where CO₂ is permanently stored underground (and specifically offshore, in the case of the Acorn project). However, the captured CO₂ could also be used as feedstock by industries willing to pay for it, which would transform CO₂ from a potentially very large liability (especially for sites far from the storage infrastructure) into an asset, thus significantly improving the business case for CCUS.

4.3.2 Abatement potential

Permanent CO₂ sequestration vs utilisation

Since this study specifically investigates pathways for deep industrial decarbonisation, the only CCUS options of interest here are those able to deliver emissions savings comparable to the case of permanent storage (i.e. CCS). It is highlighted in this context that not all CO₂ utilisation routes satisfy this criterion. Indeed, if the CO₂ is released to the atmosphere again after capture (as might be the case when it is used to create synthetic fuels or when it is used in greenhouses to accelerate plant growth), the climate benefit from CCUS would be significantly lower.

Capture rates

The capture rateis one of the factors that most directly affects the abatement potential of CCUS since any CO₂ that remains in the flue gases after the capture process is released to the atmosphere. As mentioned above, capture rates of 90% are assumed to apply in this study,^[107] though higher rates (e.g. 95-99%) are deemed possible and are considered in the context of tackling residual emissions (see Section 6.1.4). Higher costs are to be expected when increasing the capture rate, given that CO₂ would need to be removed from flue gases containing ever lower CO₂ concentrations (since CO₂ has already been stripped from them). A recent report by the IEA Greenhouse Gas R&D Program (IEAGHG) however indicates that the additional cost of increasing capture rates to 95% could be negligible in the case of gas and coal power plants (per unit of CO₂ captured),^[108] though it is not clear from this study whether the increased energy demand and the corresponding emissions intensity were accounted for. It is also noted that the size of the capture plant increases with higher capture rates, hence space constraints would need to be considered when deciding on the capture rate.

Emission intensity of the energy needed for CCUS

The high energy requirements for CO₂ capture are a known limitation to the net abatement achievable through CCUS. If the heat demand is met via natural gas, as would be most likely be the case today, it would therefore be expected that capture technologies with higher energy requirements (e.g. first generation amines) would deliver the lowest level of abatement, strengthening the case for using more advanced technologies instead. To ensure that CCUS delivers the greatest amount of abatement possible for the selected Scottish industries it is assumed that hydrogen will be used to meet the heat requirements for carbon capture. This assumption is considered realistic since low-carbon hydrogen is envisioned to be produced in Grangemouth more or less at the same time as CCUS deployment starts, and it could be possible that natural gas is temporarily be used to bridge any delay in its availability with negligible impact on the overall decarbonisation pathway.

Negative emissions with BECSS

No discussion on the abatement potential of CCUS would be complete without a mention of the potential for negative emissions which could be unlocked by combining CCUS with bioenergy combustion (known as bioenergy CCS, or BECCS). This possibility is assessed in this study only in the context of the cement industry for two reasons. First, this industry already uses a small share of biomass (contained in waste-derived fuels) and it is assumed that the share will increase to 70% of the total energy required by the kiln when implementing fuel switching (see Section 4.2.1) – which would make the cement plant the single largest industrial user of bioenergy in Scotland. Second, CCUS is assumed to be necessary to abate process emissions within the cement industry, hence the additional cost of capturing emissions from the kiln (as well as from the calcination process) would likely be reduced compared to that of a standalone installation.

This does not imply that other industries cannot also contribute negative emissions. For instance, the whisky distilleries and the wood panel manufacturers that already use organic residues could in theory also implement BECCS. Their limited size and often remote location – potentially far from CO₂ transport and storage infrastructure – however suggest that CCUS will be costlier to implement at these sites. Additional possibilities for achieving negative emission with CCUS are discussed in Box 5.

Box 5 – Additional possibilities for negative emissions

Some negative emissions are achieved within the deep decarbonisation pathways by implementing CCUS on the cement kiln, where substantial amounts of biomass can be combusted. More extensive implementation of BECCS within the industries in scope could potentially be proposed, for instance, if the refinery or petrochemical sites were to co-fire biogas within their furnaces.

Alternatively, negative emissions could be attained by mixing biogas in the natural reformer used for blue hydrogen production. According to a previous study by Element Energy,^[109] a gas mix with just under 10% biogas could lead to a negative emission intensity for blue hydrogen of -10 gCO₂/kWh. Speirs et al. instead find that hydrogen produced via biomass gasification^[110] could deliver hydrogen with an emission intensity of -371 gCO₂/kWh.^[111] While there are a number of bioenergy routes to the production of negative-emissions hydrogen, it is important to reflect on the scarcity of sustainable bioenergy sources and on the fact that their optimal use may be found in other sectors.^[112]

Direct air carbon capture and storage (DACCS) – i.e. the capture of the highly diluted CO₂ contained in the atmospheric air – could also become commercially available at a price that is competitive with other options for deep decarbonisation. DACCS might in fact be a cheaper solution, if implemented at scale, compared to the capture of CO₂ from many small emission sources, in spite of their higher CO₂ concentration. Cost savings could in this case arise from economies of scale and from the possibility to locate DACCS close to the relevant CO₂ transport and storage infrastructure.

4.3.3 Barriers and enablers

Key barriers which may hinder CCUS deployment include:

• The high cost and consequent impact on competitiveness.
• The low maturity of capture technologies which have never been commercially deployed within certain sectors, which would entail deployment of first-of-a-kind (FOAK) technologies.
• The commercial complexity of CCUS projects, which must bring together emitters, transport infrastructure developers and storage operators in the context of what is often FOAK projects for the parties and regions involved.
• Exposure of each party to 'counter-party risk', i.e. the risk that another party may at some point withdraw with disastrous consequences for the overall project economics, is also considered an important commercial barrier.
• Distance from the CO₂ infrastructure (CO₂ pipelines and storage sites) increases cost and hence is often an obstacle to the development of CCUS. For the Scottish industries on which CCUS is assumed to be deployed in this study, however, this is not expected to be an important challenge considering the possibility to connect to the Acorn CCS project via pre-existing pipelines.
• Availability of sufficient space on site for the installation of the (large) capture equipment.
• Planning constraints.
• Exposure to uncapped liability in case CO₂ leaks from the storage site.

A detailed review of the main barriers to CCUS deployment can be found in a previous study by Element Energy for IEAGHG,^[113] where one of the main enablers is specifically investigated: i.e. to pool demand for CO₂ by clustering, which could make it more economical and less risky to develop a CCUS project.