Deep decarbonisation pathways for Scottish industries: research report

6 Results

6.1 Decarbonisation potential

The three pathways introduced in the previous chapter were evaluated until 2050, and emission reductions from the reference year (2018, when emissions from industries in scope amounted to 6.7 MtCO₂e) were calculated for 2030, 2040, and 2045, i.e. the years for which economy-wide emission targets apply (or 75%, 90%, and 100% reduction against the 1990 baseline, respectively). It should be noted in this context that no industry specific targets have been set, and that the pathways studied here are not constrained to meet any economy-wide or other emission reduction targets, but rather build on the bottom-up uptake assumptions presented in the previous chapter.

6.1.1 Deep decarbonisation pathways

The first finding of this study is that emissions from the industries in scope decrease by over 80% below 2018 levels by 2045 in both deep decarbonisation pathways, reaching1.2 MtCO₂e in the Electrification pathway and 1.3 MtCO₂e in the Hydrogen pathway, as summarised in Table 5. The similar decarbonisation potential of the two pathways is first of all explained by the fact that no fossil-fuelled appliance was found to be replaceable by only hydrogen or only electrical appliances and, second, by the comparable carbon intensities of electricity and hydrogen fuel when compared to fossil fuels.

These considerations also imply that a hybrid pathway, where electrification occurs at certain sites and hydrogen fuel-switching at others, would be able to deliver similar emission reductions as the pathways assessed here, possibly more rapidly and cost-effectively. Section 6.5 discusses what such a pathway may look like.

The main difference in the emission trajectories of the two deep decarbonisation pathways is in how rapidly they decarbonise, which determines their performance against the economy-wide interim target years as well as the cumulative abatement they deliver (8 MtCO₂e higher for the Electrification pathway, by 2045), also reported in Table 6. This can be explained by two factors:

• The greater maturity of electrification technologies means that they can be deployed sooner.
• Only a handful of sites are expected to have switched to hydrogen by 2030, since no low-carbon hydrogen is assumed to be available before 2028.

Despite this, the two pathways deliver comparable yearly emissions reductions by 2045 owing to the assumption that hydrogen technologies can be deployed more rapidly when they become available (see Section 5.3.3). It is also noted that a slightly lower carbon intensity is assumed for hydrogen in 2045 in comparison to electricity, which partly counteracts the slightly greater uptake of electrification options by that same year.

Table 5 – Emissions from industries in scope in 2030, 2040, and 2045

Pathway emissions (MtCO2e)	2030	2040	2045
Baseline	6.5	6.2	6.1
Efficiency	6.3	6.0	5.9
Electrification	5.3	1.7	1.2
Hydrogen	5.9	1.9	1.3

Table 6 – Emissions abatement for industries in scope in 2030, 2040, and 2045

	Electrification			Hydrogen
	2030	2040	2045	2030	2040	2045
Net abatement vs 2018	21%	75%	82%	12%	72%	81%
Cumulative abatement since 2018 (MtCO2e)	6	48	74	3	40	66

Emissions reductions from 1990 levels

To assess what these results would mean for the contribution from all Scottish industries to the achievement of net zero targets, an assumption must be made around the emission reductions achieved by industries not in scope. Two alternative assumptions are proposed here as the likely range of emissions reductions for all Scottish industries by 2045 (note that baseline emissions from all Scottish industries are 21.0 MtCO₂e, and that they were already 45% lower than this by 2018, as noted in Section 3.1):

• If it is assumed that industries out of scope decarbonise at the same rate as industries within scope, overall emissions from all Scottish industries would amount to 2.0-2.2 MtCO₂e in 2045, i.e. 90% lower than in 1990.
• Conversely, if industries out of scope do not decarbonise at all (i.e. the worst-case scenario), overall emissions would be 6.0-6.1 MtCO₂e in 2045, or about 70% lower than 1990 levels.

Table 7 – Residual emissions and estimated abatement from all Scottish industries

Residual emissions in MtCO2e and % reduction from 1990 levels	Electrification			Hydrogen
	2030	2040	2045	2030	2040	2045
If industries not in scope decarbonise at the same rate as industries in scope	9.1 (-57%)	2.9 (-86%)	2.0 (-90%)	10.1 (-52%)	3.2 (-85%)	2.2 (-90%)
If industries not in scope do not decarbonise	10.1 (-52%)	6.5 (-69%)	6.0 (-72%)	10.7 (-49%)	6.7 (-68%)	6.1 (-71%)

6.1.2 Efficiency pathway

While emissions are seen to substantially reduce in both deep decarbonisation pathways, they only reduce by 12% below 2018 levels in the Efficiency pathway. This means that the implementation of all efficiency improvements is estimated to reduce emissions by only 4% more thanwhat can be achieved with the sole implementation of incremental efficiency improvements in the BAU scenario. Despite the relatively small role that efficiency improvements play on average within the context of the transition to net zero, the analysis presented in the next subsection shows that sectors like food and drink can cut their emissions far more substantially by improving efficiency.

There are also broader benefits of increasing energy efficiency – the reduced energy demand reduces fuel costs and also reduces the need for deploying additional energy infrastructure, for instance – it can therefore be understood why improving energy efficiency retains an important role within industrial decarbonisation plans. Nevertheless, since it clearly appears that increasing energy efficiency alone is insufficient to deliver substantial emissions reductions, the Efficiency pathway will not be analysed further.

6.1.3 Sectoral and technology contributions

It was noted above that the two deep decarbonisation pathways hold a similar potential to reduce overall emissions from the industries in scope by 2045, and since this similarity also applies at the level of each individual sector only the numerical results for the Electrification pathway are reported in the remainder of this section (results for the Hydrogen pathway are shown in Appendix 8.8). By analysing sector-level emission reductions and the corresponding contribution from efficiency improvements, fuel switching and CCUS, shown in Figure 11, it can be noted that:

• Nearly 60% of the overall abatement occurs within oil & gas and the chemical industries, which are the largest-emitting sectors today.
• CCUS is expected to be the main decarbonisation technology for the oil and gas, chemicals and the cement industries, delivering about 60% of the emissions abatement within these sectors.^[125]
• BECCS can deliver nearly 0.3 MtCO₂e of negative emissions within the cement industry. Without it, residual emissions from the industries in scope would be 22% higher in 2045.
• Fuel switching is essential to all other sectors where CCUS is not assumed to be deployed, accounting for about two thirds of their emission reductions.
• Incremental efficiency improvements offer a more moderate overall contribution (11% on average) but are far more important for the decarbonisation of certain industries, specifically food and drink. It is also worth noting that similar emission reductions could also be achieved without improving energy efficiency, though this would increase the amount of low-carbon energy needed for fuel switching and therefore increase the cost of decarbonisation.

When looking at the cross-sectoral contributions from individual decarbonisation technologies, shown above for the Electrification pathway (results are broadly similar for the Hydrogen pathway), it can be observed that:

• CCUS delivers 49% of the emissions reductions, abating 2.7 MtCO₂e.
• All fuel-switching technologies combined are responsible for 41% of the overall reduction (2.1 MtCO₂e).
• As noted above, energy efficiency improvements account for the remaining share (0.6 MtCO₂e).

Furthermore, the fact that switching fuels in boilers and CHP plants accounts for 81% of the emission reductions related to fuel switching highlights the critical need to focus on low-carbon steam as a key target for fuel switching.

6.1.4 Tackling residual emissions

To address those emissions sources that are not decarbonised within the pathways considered before 2045, additional measures must be considered. Leaving aside the negative emissions from the cement sector for this analysis, it can be seen from Figure 13 that the nearly 2 MtCO₂e of residual emissions in the Electrification pathway are attributable to the following sources (results for the Hydrogen pathway are shown in Appendix 8.8 and are not repeated here due to their substantial similarity with the below):

• CO₂ that escapes capture because of the assumed 90% capture rates (171 ktCO₂e), which could partly be avoided if plants with higher capture rates were to be installed – a decision which ultimately depends on cost.
• Flaring-related emissions from the oil and gas and chemical industries (178 ktCO₂e), which may be tackled by improving process reliability and hence reducing the need to flare. Such improvements also bring other economic benefits and are actively being pursued by the interviewed stakeholders. There is a possibility that at least some flaring emissions may also be addressed via CCUS, but further work would be required to assess this possibility and its cost.^[126]
• Other process emissions from aluminium and glass manufacturing and from the Kinneil gas terminal (167 ktCO₂e distributed in similar proportions across the three industries). CCUS or process changes could be considered to lower these, as discussed in Box 7.
• Emissions embedded within low-carbon energy sources (147 MtCO₂e), which could in theory be eliminated if only renewable energy sources were used to produce it.
• The remainder (809 ktCO₂e) results from residual combustion of fossil fuels used in appliances which have not been replaced, or in appliances which could not be classified and are conservatively assumed to be unsuitable for fuel switching (see Section 3.3). If the uptake of all fuel switching technologies were to be completed by 2045, emissions could be reduced by a further 0.6 MtCO₂e.

Two further possibilities exist to abate industrial emissions further than is achieved by the deep decarbonisation pathways:

• Higher levels of negative emissions can be achieved from combining bioenergy with CCUS (i.e. BECCS) in other industries, and possibly also in the context of negative-emissions hydrogen production, which was discussed in Box 5.
• Demand-side measures such as the substitution of carbon-intensive products with others that are less carbon-intensive can also reduce industrial emissions, though this may also affect industrial activity. A common example of product substitution is to use electricity instead of petrol in the transport sector.

Product substitution could occur due to changes in consumer preference or because of the impact of demand-side policies. This possibility is not addressed here as it violates the assumption that underpins this study, i.e. that industrial products and processes do not change. The potential implications from product substitution and demand-side policies specifically are discussed in the concluding chapter.

Box 7 – Abating residual processes emissions

The deep decarbonisation pathways considered here are unable to abate all process emissions occurring in the glass, aluminium, and oil and gas industries; however, options to reduce – if not eliminate – such emissions exist.

CCUS could be applied to these processes, although it was assumed that this would not happen due to the relatively small size of the corresponding sites, as judged by emissions level. The case of the Kinneil gas terminal may present an exception in this regard, especially if CCUS is implemented at some of the Grangemouth plants: about 18 ktCO₂e of high-purity CO₂ is separated from the feedstock gases already today and could be captured with relative ease.

Alternatively, process changes could be considered. Process emissions from glass can be fully abated if recycled glass (cullet) is used instead of virgin materials in the glass melting process. For this to happen, current supply constraints – especially affecting white flint glass – would need to be overcome. Glass manufacturers across the UK already use 35-40% cullet in their feedstock, but further increases are limited by feedstock availability (especially for flint glass),^[127] which is why this option is considered of limited applicability.

In the case of aluminium, use of carbon anodes in the smelting process results in process emissions. It is acknowledged that the Elysis project, originating from the joint venture of aluminium manufacturers Rio Tinto and Alcoa, claims to have devised a solution for carbon-free smelting,^[128] but not enough detail is available in the public domain to assess the possible relevance of their technology to tackle Scotland's aluminium manufacturing emissions. Substitution of virgin aluminium with recycled aluminium would also prevent process emissions since it altogether removes the need to use carbon anodes, however this is not considered to be a probable route for the Lochaber smelter since it would require a completely new facility. Further work would also be required to assess what proportion of the existing demand for aluminium could be met with recycled aluminium.^[129]

Finally, emissions occurring when purging the flare heads at the Kinneil gas terminal (~35 ktCO₂e) could potentially be averted by replacing the current hydrocarbon-based purged gas, for instance with nitrogen.

6.2 Enablers

6.2.1 Essential conditions

The achievement of the emission trajectories shown in Figure 10 is underpinned by the assumption that four essential conditions are met:

• Substantial economic incentives must be put in place via suitable policies. Without these, no significant investment in deep decarbonisation is to be expected.
• All decarbonisation options must be adopted promptly when they become sufficiently mature from a technical and commercial point of view. This is a process which may also be brought forward with appropriate policy interventions.
• Enabling energy assets and the relevant infrastructure must be deployed in advance, otherwise individual decarbonisation efforts might be delayed.
• Site managers and investors need to have sufficient confidence in, and knowledge of, the relevant technologies andin the timescales for their commercialisation, which will likely originate via the successful deployment of relevant demonstration projects within each industrial sector.

Failure to meet any of the above conditions would likely result in the delayed uptake of the key decarbonisation technologies, which willmakeit even more challenging to achieve Scotland's accelerated net zero targets. Considering that a certain proportion of the industrial sites will likely need to replace their current appliances within the next decade – potentially before suitable fuel-switching options are sufficiently mature to be considered as a viable option – a risk exists that these sites will be forced to choose fossil-fuelled appliances. This possibility, known as the risk of 'technology lock-in', is the most likely outcome unless the abovementioned conditions are met. Should this happen, the long investment cycles dictated by useful appliance lives in excess of 20 years mean that, for many sites, there might be only one chance to fuel-switch before 2045; and it also means that, unless the four conditions above are met, the opportunity of a low-carbon replacement could be missed (unless appliances are scrapped before the end of their useful life, which could potentially be mandated by policy). It is worth noting two potential advantages that hydrogen fuel switching may present over electrification in mitigating the risk of technology lock-in:

• Sites requiring to replace their appliances before low-carbon hydrogen is available could install 'hydrogen-ready' appliances, i.e. appliances that are optimally designed to run on pure hydrogen but are initially configured to run on natural gas, with minimal work required at the point of switchover. This would enable operators to switch to hydrogen as soon as it becomes available.
• Retrofits could convert appliances fuelled by natural gas to working with hydrogen, though this may not be applicable to all processes and further technology developments may be necessary.^[130]

Conversely, it should be noted that electrification may be the only fuel switching option available to many sites within the next 5-10 years, hence the risk of technology lock-in could also be prevented by enabling site operators to invest in electrification as soon as they are ready to do so, which would also require suitable economic incentives.

6.2.2 Key deployments

Whenever fuel switching or CCUS are implemented on the largest emission sources, clear downward steps in the emissions trajectories in Figure 10 can be seen.^[131] The key deployments from which the most evident steps arise – and which deliver approximately two thirds of the overall abatement expected by 2045^[132] – are included in the timeline shown in Figure 14 (detailed information is summarised in Table 8 and Table 9).

The fact that fuel switching starts sooner with Electrification explains why the downward steps are seen to occur earlier for this pathway. As a reminder, the delayed start of the Hydrogen pathway is due to the assumption that low-carbon hydrogen only becomes available in 2028, when:

• The first phase of the Grangemouth CCUS project becomes operational and carbon capture starts at the petrochemical plant and at a new natural gas reformer dedicated to the production of blue hydrogen.
• The captured CO₂ is assumed to be transported via the Feeder 10 pipeline to the Acorn CO₂ Storage site, currently under evaluation.^[133]
• In the same year, smaller-scale production of green hydrogen is also assumed to begin.^[134]

It should however be noted that hydrogen could contribute to decarbonising industry sooner than is modelled here since, as it was noted above, gas blends with up to 20% hydrogen by volume should be compatible with the current gas infrastructure and equipment.^[135] Thus, even assuming blue hydrogen production cannot start sooner, some green hydrogen could in theory be injected into the gas grid to help Scottish industries decarbonise sooner, though it is unlikely that green hydrogen production could meet more than a small portion of the energy demand supplied via the gas grid. Alternatively, some of the hydrogen produced at St Fergus within the Acorn project or produced abroad and imported to Scotland may also be available to a few of the industries in scope. Finally, it can also be seen from the tables below that the largest emission sources are assumed to decarbonise by 2045, which explains why the pathways converge by then.

Table 8 – Carbon capture deployments

Site and process	Deployment year	CO2 captured[136] (MtCO2/year)
Grangemouth olefins plant: furnaces[137]	2028	0.4
Grangemouth refinery: furnaces and SMR	2031	1.1
Fife ethylene plant: furnaces	2034	0.5
Dunbar cement plant: kiln and calcination process	2037	0.5
Natural gas reformer for blue hydrogen production[138]	2028	2.1[139]
Total from industry		2.6
Total inc. blue hydrogen production in the Hydrogen pathway		4.7

Table 9 – Main fuel switching deployments

Site and processes	Deployment year		Estimated energy demand (TWh)[140]
	Electrification	Hydrogen
Grangemouth refinery & olefins plant: CHP plants & other boilers[141]	2025, 2031, 2032, 2039, 2043	2033, 2037, 2043, 2044	4.3
Fife ethylene plant: boilers	2025	2030	0.9
Alloa glass plant: furnaces	2030	2033	0.5
Dunbar cement plant: kiln[142]	2032	2032	0.5
All main deployments			6.2

6.2.3 Energy assets and infrastructure requirements

The results presented in the previous section are underpinned by the assumption that substantial growth in the supply of low-carbon energy can occur, which in turn requires the deployment of numerous assets broadly categorised here as infrastructure.Threeassets categories are further defined:

• Assets for the generation of electricity from renewable energy sources.
• Energy conversion assets for producing hydrogen from natural gas or renewable electricity.
• Other infrastructure, which includes all other assets necessary to access low-carbon energy as well as those related to CCUS for blue hydrogen production and at industrial sites.

Table 10 – New low-carbon energy assets required ^[143]

Energy end use form	Primary energy source	Generation assets	Conversion assets	Other infrastructure	Focus of quantitative assessment
Hydrogen (blue)	Natural gas[144]	-	Natural gas reformers	CO2 transport, CO2 storage[145]	Amount of CO2 processed in 2045 [MtCO2/year] Natural reforming capacity in 2045 [TWh/year]
Hydrogen (green)	Renewable energy sources[146]	Wind farms, solar farms etc.	Water electrolysers	Hydrogen distribution	Water electrolysis capacity in 2045 [TWh/year] Renewable generation capacity in 2045 [TWh/year]
Electricity	Renewable energy sources	Wind farms, solar farms etc.	-	Grid upgrades	Renewable generation capacity in 2045 [TWh/year] Grid upgrade needs by 2045 [GW]

Depending on the end use and primary forms of the low-carbon energy used, different sets of assets are needed. 'End use energy' refers to what end users see, i.e. hydrogen and electricity, which can in turn be produced via a multitude of routes and from different (primary) energy sources. For instance, hydrogen can either be produced via natural gas reforming or from water electrolysis powered by renewable energy sources, and it is easy to see why the infrastructure requirements strongly depend on the choice of primary energy form, as summarised in Table 10. At the same time, certain infrastructure needs only depend on the end use form: hydrogen fuel switching for instance hinges on the development of new hydrogen distribution channels regardless of whether blue or green hydrogen is used.

An additional infrastructure-related assumption which was implicitly made when developing the decarbonisation pathways is that sufficient distribution infrastructure is developed in due time so as not to constrain uptake of hydrogen technologies from industry. Finally, no assumption is made around which technology is used to generate renewable electricity or to produce green hydrogen, or as to where these assets are located.

End use energy demand

To assess the infrastructure requirements within each decarbonisation pathway it is first necessary to quantify how much low-carbon energy is needed in both its end use form and in its primary form. End use energy demand totalled just over 27 TWh before 2020, as estimatedby breaking down emissions by fuel type (see Figure 4) and accounting for the average emission factors for each fuel type. The demand for hydrogen and electricity^[147] was then calculated after accounting for the assumed energy efficiency improvements (see Table 3) and for the different efficiencies of fossil-fuelled and fuel-switching appliances (see Table 4). Energy requirements for CCUS deployments at industrial sites were also evaluated (see page 50), since these significantly affect the total demand for low-carbon energy as is easily seen from Figure 15.^[148]

Following the approach outlined above it was estimated that the overall amount of end use energy required within the Electrification pathway will be 3.1 TWh lower in 2045 than it is today, whereas it will only be 0.3 TWh lower in the Hydrogen pathway, as shown in Table 11. Since it is assumed that industrial output remains steady, the fact that energy demand reduces is mostly due to the incremental improvements in energy efficiency, whichreduce the demand by 2.7 TWh in both pathways. The higher energy efficiency of electrification appliances (specifically heat pumps) compared to both fossil-fuelled and hydrogen appliances however contributes to reducing energy demand by a further 2.4 TWh in the Electrification pathway. The combined energy demand from all industrial CCUS deployments (over 90% of which is heat for carbon capture), which require 2.1 TWh of hydrogen and 0.2 TWh of electricity in both pathways, partly offsets the effect of all efficiency improvements. Analysis of this table also helps quantify the impact of the later deployment of hydrogen technologies, which implies a residual use of fossil fuels in 2045 that is 0.2 TWh greater in the Hydrogen pathway. Conversely, internal fuel use is assumed to remain constant in both pathways for the reasons explained in Section 3.2.

Table 11 – Breakdown of end use energy demand in 2045

Energy vector	Initial demand (TWh)	Electrification (TWh)	Hydrogen (TWh)
Internal fossil fuels	9.5	9.5	9.5
Purchased fossil fuels	17.7	3.9	4.1
Electricity	-	8.1	0.3
of which for industry	-	7.9	0.1[149]
of which for CCUS	-	0.2	0.2
Hydrogen	-	2.3	12.6
of which for industry	-	0.1	10.5
of which for CCUS	-	2.1	2.1
Biomass	-	0.4	0.4
Total	27.2	24.1	26.9

Primary energy demand

The demand for primary energy plotted in Figure 16 was calculated from that for end use energy after accounting for the energy losses along the supply chain (specifically those in the hydrogen conversion step and in electricity transport and distribution) according to the net efficiency values reported in Table 12.

Table 12 – Efficiencies and load factors of generation and conversion assets

Generation asset	Load factor	Net efficiency
Renewable electricity	27%[150]	92% (generation-to-use)[151]
Water electrolyser for green hydrogen	27%[152]	60% (power-to-H2)[153]
Natural gas reformer for blue hydrogen	85%	77% (gas-to-H2)[154]

The following assumptions were also made:

• Dedicated renewable electricity generation is assumed to be used for green hydrogen production, which implies that that green hydrogen is only produced when the generation assets are operating (which explains why the respective load factors match) and that 1 GW of generation capacity is needed for each GW of installed electrolyser capacity.
• Only green hydrogen is used in the Electrification pathway, whereas a 50:50 mix of blue and green hydrogen is used in the Hydrogen pathway by 2045 (the cases where only green or only blue hydrogen are used are also discussed below).

Three significant differences between the two pathways emerge when comparing the detailed breakdown of primary energy requirements provided in Table 13:

• Primary energy demand in 2045 is 5.9 TWh higher in the Hydrogen pathway than it is in the Electrification pathway.
• Even though electricity demand is 8.6 TWh higher in the Electrification pathway, the overall need for new renewable generation is only 2.7 TWh higher than in the Hydrogen pathway. This is because of the large demand for new renewable electricity for green hydrogen production.
• The demand for natural gas as feedstock to produce blue hydrogen (+8.9 TWh) partly offsets the reduction in the demand for purchased fossil fuels (-13.6 TWh), most of which is natural gas.

The first point above warrants further explanation. The higher efficiency of electrical appliances, already discussed in the context of end use energy, partly explains the difference in primary energy demand, but another reason which impacts demand to a similar extent can be found in the lower end-to-end energy efficiency of the hydrogen supply chain, compared to that for electricity. Specifically, over a third of the renewable energy that powers the water electrolyser is lost in the processes, whereas less than 10% of the electricity is lost while transported along the electricity grid. Incidentally, this also explains why the demand for new renewable generation is still substantial in the Hydrogen pathway and would in fact be even higher than in the Electrification pathway if only green hydrogen was used.

Table 13 – Primary energy demand in 2045

Energy source	Initial demand (TWh)	Electrification (TWh)	Hydrogen (TWh)
Internal fossil fuels	9.5	9.5	9.5
Purchased fossil fuels	17.7	3.9	4.1
Bioenergy	-	0.4	0.4
Renewables		12.6	9.9
of which for electricity use	-	8.9	0.3
of which for green hydrogen	-	3.8	9.6
Natural gas for blue hydrogen	-	-	8.9
Total primary energy	27.2	26.4	32.8

New asset and infrastructure requirements

Building on the analysis presented above, the new asset and infrastructure requirements can finally be quantified using when relevant the load factor assumptions listed in Table 12. The results reported in Table 14 (referring to 2045, like the rest of the discussion below) highlight several differences between the new asset and infrastructure requirements of the two pathways, some of which are best highlighted when considering two sensitivities for the hydrogen pathway, i.e. the case where only green or only blue hydrogen (H₂) are used.

Both deep decarbonisation pathways will likely require significantly higher levels of renewable electricity generation,with the exception of a Hydrogen pathway that mostly relies on blue hydrogen. Looking for instance at the Electrification pathway, the 5.3 GW of new renewable capacity required to decarbonise industry would represent a 45% increase over the current level of renewable generation in Scotland.^[155] This increase is on top of the additional renewable generation capacity necessary to meet increasing clean electricity demands from other parts of the energy system (for electric vehicles and domestic heat pumps, for example). The requirement for new renewable generation would not be much lower in the Hydrogen pathway: if more than 60% of the hydrogen used by industry were to be produced from water electrolysis, it would indeed be higher.

Table 14 – New assets and infrastructure required by 2045

Asset category	Description	Electrification pathway	Hydrogen pathway
			Default	100% green H2	100% blue H2
Generation	Renewable generation capacity [GW]	5.3	4.2	9.0	0.1
Conversion	Water electrolyser capacity [GW]	1.6	4.0	8.9	-
	Natural gas reforming capacity [GW]	-	1.2	-	2.2
Other infrastructure	Grid upgrades [GW]	2.3	0.01	0.01	0.01
	Hydrogen distribution [TWh/year]	2.3	12.6	12.6	12.6
	CO2 transport [MtCO2/year]	2.6	4.8	2.6	6.7
	Cumulative CO2 storage [MtCO2]	36	67	36	79

Substantial deployment of water electrolysis and/or natural gas reforming (plus CCUS) will be necessary to decarbonise industry with hydrogen fuel switching: up to 8.9 GW of electrolysers or up to 2.2 GW of natural gas reformers would be needed by 2045 if only green or only blue hydrogen were used respectively. For context, a UK company is currently building an electrolyser manufacturing facility in Sheffield with the aim to manufacture up to 1 GW of electrolysers per year by 2025.^[156]

A minimum of 2.6 MtCO₂/year would also need to be transported and stored in both pathways, with up to a further 4.1 MtCO₂/year related to blue hydrogen production in case no green hydrogen is available. This would in turn imply that 36-79 MtCO₂ would need to be stored (or utilised) by 2045. For reference, saline aquifers in Scotland have an estimated storage capacity of 4,600-46,000 MtCO₂).^[157]

The required grid upgrades were conservatively estimated to be 2.3 GW in the Electrification pathway (assuming that deployment of electrical appliances always requires an upgraded grid connection).^[158] Grid upgrade requirements for the Hydrogen pathway are instead minimal as they only relate to the mixed-fuel kiln used by the cement industry.

Finally, hydrogen distribution infrastructure able to process 2.3-12.6 TWh/year of hydrogen will be needed by 2045. While this does not change depending on whether blue or green hydrogen is used, it may be that the characteristics of such infrastructure will differ depending on whether centralised blue hydrogen production or decentralised green hydrogen production are more prevalent.

It should be noted that although it is often thought that the existing gas network might at least in part be repurposed to handling hydrogen, this is neither completely certain nor sufficient to handle the hydrogen volumes (if all gas is replaced with hydrogen). Speirs et al. indeed note that "[t]here is limited real-world evidence on the capability of low-pressure gas networks to transport 100% hydrogen gas streams effectively",^[159] hence there is a possibility that the hydrogen distribution infrastructure will need to substantially rely on new build. Even if the existing gas grid can be converted to carry hydrogen, the significantly lower energy density of hydrogen^[160] implies that substantial grid expansions would be nonetheless required to enable hydrogen to replace natural gas in full. For remote sites not connected to the gas grid it may be more likely that hydrogen distribution will occur via land, e.g. by lorry or by train, a solution which may also be temporarily implemented for sites wishing to fuel switch before the hydrogen infrastructure is ready.

6.3 The cost of decarbonisation

The cost of decarbonisation was calculated as the additional cost incurred in the decarbonisation pathways compared to that incurred in the BAU scenario. Four cost factors were quantified by utilising the cost assumptions presented in Chapter 4:^[161]

• Capital expenditure (CAPEX), also known as 'upfront cost' or 'investment cost'. This was broken down into an annualised CAPEX plus a financing cost.^[162]
• Operational expenditure (OPEX), which includes fixed and variable operations and maintenance costs but excludes the cost of energy.
• Energy costs, calculated as a variable cost per unit of fuel or electricity used.
• CO₂ transport and storage fees (T&S fees) for sites that implement CCUS.

A further cost factor which does not affect the cost of decarbonisation but was calculated because of its relevance to the business case for investing in deep decarbonisation is thecarbon cost, i.e. the policy cost induced by the carbon price, assumed to increase over time in line with what is specified in the guidelines by BEIS (see Appendix 8.6). More specifically, it was the carbon cost avoidance resulting from deployment of the different decarbonisation measures which was calculated, and this is plotted in Figure 17 alongside the other cost factors, assessed separately for industry and for CCUS so to provide a clearer view of how and why the two deep decarbonisation pathways differ (note that CCUS is deployed to the same extent and at the same time in both pathways). As for the incremental improvements in energy efficiency, they have no impact on the additional cost of decarbonisation. This is because the efficiency measures implemented in the deep decarbonisation pathways are also assumed to be implemented in the BAU scenario, hence their cost contributions cancel out.

6.3.1 Additional cost compared to business as usual

By analysing Figure 17, which shows how the cost of decarbonisation evolves over time as the uptake of fuel switching and CCUS progresses, two important results become apparent. The first is that the industries in scope can be expected to incur additional costs of around £0.8-1 billion per year by 2045 in the Electrification and Hydrogen pathway, respectively (neglecting carbon cost avoidance).For context, the Scottish manufacturing industry turned over £35 billion in 2017.^[163]

The second key result is that carbon cost avoidance can significantly offset the cost of decarbonisation. If this is counted as a saving, annual decarbonisation costs reach a peak of around £200 million per year in the 2030s and slowly decrease afterwards due to the assumed increase in carbon price. (The next chapter however discusses why carbon cost avoidance is not a reliable basis for investment). In both pathways there comes a point when the cost of decarbonisation becomes negative, i.e. decarbonisation becomes cheaper than paying for unabated emissions.This inflection point occurs earlier in the Electrification pathway (2043, vs 2048 in the Hydrogen pathway) because of the considerably lower cost of energy in this pathway.

The cumulative cost of decarbonisation, i.e. the sum of all costs incurred up to and including 2045, was estimated to be around £11 billion (discounted net-present values are shown in Table 15).^[164] The analysis of the contribution from each cost factor highlights that:

• The additional cost of low-carbon energy compared to fossil fuels represents the greatest cost factor for both pathways. The low-carbon energy used in industrial appliances in fact bears an overall additional cost of £4.7 billion (Electrification) and £4.9 billion (Hydrogen), a value which grows by a further £1.5 billion when the energy used for CCUS is included. This underlines the importance of reducing the cost of low-carbon energy.
• Capital expenditures related to industrial appliance replacement account for £1.4 billion in the Electrification pathway but only £0.9 billion in the Hydrogen case. The cost of grid upgrades (~£0.4 billion, only in the Electrification pathway) largely explains the difference between the pathways. Also, a slightly larger share of fossil-fuelled appliances is electrified by 2045 (see Table 11).
• A total financing requirement of £3.0 billion (Electrification) and £2.5 billion (Hydrogen) can be estimated when also including the CAPEX on carbon capture.^[165]
• Operating costs (OPEX plus CO₂ transport and storage fees) contribute a further £1.7 billion (Electrification) and £2.4 billion (Hydrogen). It is worth noting that the expenditure relating to operating electrical appliances is expected to be lower than that referring to fossil-fuelled appliances, which explains the negative OPEX shown in Table 15 for the Electrification pathway.

Table 15 – Breakdown of the cumulative cost of decarbonisation

Cost factor	Electrification pathway			Hydrogen pathway
	Undisc. £ billion	Disc. £ billion	% of total	Undis. £ billion	Disc. £ billion	% of total
Capex (industrial appliances)	1.4	0.7	13%	0.9	0.4	8%
Opex (industrial appliances)	-0.3	-0.2	-3%	0.4	0.2	3%
Energy cost (industry)	4.7	2.5	43%	4.9	2.3	43%
Capex (CCUS)	1.6	0.8	14%	1.6	0.8	14%
Opex (CCUS)	1.1	0.6	10%	1.1	0.6	10%
Energy cost (CCUS)	1.5	0.8	14%	1.5	0.8	14%
T&S fees (CCUS)	0.9	0.5	8%	0.9	0.5	8%
Total cumulative cost	11.0	5.6	100%	11.2	5.5	100%

The results above indicate that both pathways face a significant additional cost which will need to be addressed for deep decarbonisation to take place. It is in this context useful to consider the substantial cost savings that industry could benefit from by reducing carbon emissions and the related charges. Indeed, cumulative costs reduce by 83% (Electrification) or 73% (Hydrogen) if carbon cost avoidance is accounted for, as shown in Table 16.

Table 16 – The cost of decarbonisation net of carbon cost avoidance

Cost factor	Electrification pathway			Hydrogen pathway
	Undisc. £ billion	Disc. £ billion	% of total	Undisc. £ billion	Disc. £ billion	% of total
Total cumulative cost of decarbonisation	11.0	5.5	100%	11.2	5.5	100%
Carbon cost avoidance	-9.1	-4.4	-83%	-8.2	-3.9	-73%
Total cumulative cost net of carbon cost avoidance	1.9	1.2	17%	3.0	1.6	27%

In interpreting these results it is important to note the uncertainty surrounding future energy and technology costs, which implies that the values shown here can only be approximate estimates of the true cost of decarbonisation.

'Optimism bias' could be affecting the technology cost estimates utilised in this study, since as HM Treasury reports "[t]here is a demonstrated, systematic, tendency for project appraisers to be overly optimistic", especially in the absence of robust primary evidence.^[166] There are also reasons to believe that certain costs could turn out to be lower than was assumed here.It is for instance possible that the cost of low-carbon electricity will reduce substantially over time due to the large-scale penetration of renewables with zero marginal cost of generation. Likewise, unpredictable technological advancements and economies of scale could lead to technology cost reductions.

6.3.2 Carbon cost avoidance and abatement cost

To evaluate the relative merits of fuel switching and CCUS it is possible to analyse their individual contributions to the cumulative cost within each pathway as well as the decarbonisation benefit they provide, as measured by the reduced carbon charges. A review of the results in Table 17 shows that fuel switching is responsible for a higher portion of the overall cost, compared to CCUS, but offers lower carbon savings. Indeed, the net cost of CCUS is relatively small, meaning that investment in CCUS could nearly pay off on the basis of carbon cost avoidance alone, provided the carbon price increases as per the assumptions in Appendix 8.6. In contrast, fuel switching is expected to incur a net cost of £1.6 billion (Electrification) and £2.7 billion (Hydrogen). However, it is unlikely that the increasing carbon price alone would trigger the required investment in decarbonisation due to the impact that it would have on industrial competitiveness and the consequent risk of carbon leakage, discussed in the next chapter.

Table 17 – Relative costs and benefits of fuel switching and CCUS

Cost contribution	Electrification pathway			Hydrogen pathway
	Undisc. £ billion	Disc. £ billion	% of total pathway cost	Undisc. £ billion	Disc. £ billion	% of total pathway cost
Total cost of fuel switching	5.9	3.0	53%	6.1	2.9	54%
Carbon cost avoidance from fuel switching	-4.2	-2.1	-39%	-3.4	-1.6	-30%
Net cost of fuel switching	1.6	0.9	15%	2.7	1.3	24%
Total cost of CCUS	5.1	2.6	47%	5.1	2.6	46%
Carbon cost avoidance from CCUS	-4.8	-2.3	-44%	-4.8	-2.3	-43%
Net cost of CCUS	0.3	0.3	3%	0.3	0.3	3%

Another way to think about the cost of decarbonisation is to consider the levelised cost of abatement (LCOA), which represents the carbon price that would make each pathway cost neutral when accounting for the avoided carbon charges. The LCOA is calculated as the sum of the discounted additional costs of a pathway (independently for fuel switching and CCUS) over the sum of the discounted emissions savings arising from the pathway (see detailed methodology in Appendix 8.10). The results in Table 18 indicate that the LCOA is lower in the Electrification pathway (£157/tCO₂e) compared to the Hydrogen pathway (£188/tCO₂e), a result which could have also been inferred by noting that, although the two pathways present comparable cumulative costs, the Electrification pathway offers greater cumulative abatement (see Table 6). The LCOA can also be evaluated separately for fuel switching and CCUS, which confirms that fuel switching would cost more than CCUS, per tonne of CO₂ abated, given the cost assumptions employed in this study. This is especially true for the Hydrogen pathway, where the LCOA of fuel switching (£255/tCO₂e) is 76% higher than that of CCUS (£145/tCO₂e) (whereas it is only 17% higher in the Electrification pathway); once again, this difference between the two pathways is due to the later start of the Hydrogen pathway and correspondingly lower cumulative decarbonisation potential.

Table 18 – Levelised cost of abatement

LCOA in £/tCO2e	Electrification pathway	Hydrogen pathway
Fuel switching	169	255
CCUS	145	145
Overall	157	188

6.4 Feedback from industry stakeholders

The results presented above provide an indication of what financial resources will be needed to make the deep decarbonisation pathways investigated in this study possible. Out of the many challenges affecting these pathways, industry stakeholders highlighted the below as particularly difficult to mitigate.

Technical challenges

• Achieving the required heating profiles with alternative fuels can be challenging for certain processes. If the heat is applied indirectly via steam this would generally not be a problem since the quality of the steam would not change. In direct heating processes that are quality critical, however, the applicability of certain fuel-switching options might be restricted (which makes a hybrid pathway more likely, as discussed in the next section).
• In light of the above, the greater similarities between hydrogen combustion and fossil fuel combustion (both of which yield a flame), might make switching to hydrogen preferable to electrification.
• Although this study found that there is no insurmountable obstacle to the uptake of either hydrogen or electrical appliances in the industries considered, the two are not always equivalent. Certain deployments are potentially more disruptive to site operations – often designed to run uninterrupted – than others. In extreme cases it might be preferable to implement decarbonised processes in a greenfield project (i.e. at a new site), rather than to attempt decarbonisation of current processes.
• Investment in first-of-a-kind technologies was also mentioned as a risk which would likely persist for several years after a novel technology is first deployed and until its operational characteristics are well understood.

Infrastructure challenges

• Alternative fuels will be evaluated not just on cost and quality but also on the reliability of their supply chains. This is because operational downtime (which could be triggered by lack of fuel availability) is very costly. This could imply that fuel switching would start at smaller sites with easier-to-meet requirements, rather than at the larger sites as is assumed here.
• New infrastructure is required before alternative fuels can even be considered. Electrification may require costly electricity grid upgrades and hydrogen will require a novel infrastructure altogether. Delays in infrastructure deployment may therefore hold back the pathways considered here. Policy may have a role in ensuring that no such delays happen.
• A related challenge is that it might be hard to match the high reliability guaranteed by CHP plants (often backed-up by redundant steam boilers) when connecting to the grid. This may in turn mean that additional costs must be incurred, e.g. for behind-the-meter energy storage, in addition to the ones computed here.

Investment challenges

• It is expected that private investment in the technologies considered here will be challenging on solely commercial grounds. Quick payback targets (often around 2-3 years and sometimes as low as a 6 months) are considered hard to achieve via investment in decarbonisation. Longer payback periods (e.g. 5-10 years) were generally only considered possible for projects backed by demonstrated or 'bankable' revenue streams.^[167]
• Long investment cycles – which are due to the long lifetime of industrial appliances, often lasting longer than 15 years, and in a few cases longer than 40 years – can cause a 'technology lock-in' situation, where industrial sites are unable to decarbonise if this would mean writing off recent investments with long residual useful lives. The possibility to retrofit gas-fired technologies to work with hydrogen may represent an advantage for the Hydrogen pathway in this regard.
• Competition for capital in international businesses could further complicate the investment process.
• Above all, the fact that both fuel switching and CCUS increase operating costs and hence adversely affect international competitiveness makes it hard to justify investment in these technologies. Without policy support, the increasing carbon cost might cause industries to shut and, potentially, relocate, rather than to decarbonise. Hence, greater policy certainty could be essential in mitigating the risk of carbon leakage.

Clean growth as driver for investment

Among the many challenges facing projects that aim at deep industrial decarbonisation, one potential opportunity was also highlighted by industry stakeholders: if investment in decarbonisation could lead to increased market competitiveness and be associated with growth, this would be a more powerful driver for investment compared to cost-cutting. This is especially so when the avoided costs originate from policy (e.g. the carbon cost) and is even more relevant in the context of mature industries facing limited growth prospects, or perhaps even operating in markets that are already contracting.

6.5 Considerations around a possible hybrid pathway

It was noted at the start of this chapter that a hybrid pathway, where certain sites electrify their processes while others switch to hydrogen, should deliver a similar level of decarbonisation to the Electrification and Hydrogen pathways. What is more, a hybrid has the potential do so more cost effectively. However, it is not possible to determine a priori which pathway would be preferable for each site or sector, since multiple factors that are beyond the scope of this study would need to be assessed (in addition to cost, which was previously discussed). Some of the factors with the greatest impact on whether a site would opt to electrify or switch to hydrogen are:

• Infrastructure availability: the availability of sufficient spare capacity at the local electricity substation could make electrification cheaper and faster than is considered here. The prior development of a suitable hydrogen distribution channel serving the area where a site operates would instead be essential for switching to hydrogen.
• Technology availability: there may not always be a choice for investment in low-carbon technologies, especially for investments made before the hydrogen supply chain ramps up; in this case electrification would be the only viable option, unless site operators and investors have a high confidence that relevant hydrogen appliances are going to be available within a reasonable timeframe.
• Technical characteristics: it was already noted in Section 4.2.3 that certain technologies may be more operationally disruptive than others, and the full impact that emerging technologies have on the quality of industrial products may only become clearer in the next stages of development. If so, this might strongly influence whether a site chooses hydrogen or electrification.
• Retrofitting potential: the possibility of retrofitting natural gas appliances to work with hydrogen might not only reduce the cost of fuel switching^[168] but it might also allow it to happen more rapidly, since site operators need not wait until the end of an appliance's useful life to convert it.

It is clear that analysis of these and other factors on a site-by-site basis would be necessary to judge the relative merits of different decarbonisation pathways. However, the analysis and stakeholder feedback presented above contain several important possibilities around the likely feature of a future hybrid pathway:

• Some sites will use a mix of technologies. This is likely going to be true for the cement industry (see page 35) and is also considered to be the preferred pathway for the Scotch Whisky subsector, where hydrogen could have an essential role in meeting the peaks in energy demand which heat pumps would find hard to meet.^[169]
• Electrification can in many cases start now. For appliances needing replacement in the next 5-8 years, electrification would represent a safe way to decarbonise. For this to happen, policy support would need to be offered ahead of time to make this a cost-effective, competitive option.
• In the context of CHP plants, electrification implies that these plants would be dismantled and replaced by a mix of grid connection, electric boiler, and heat pumps (see Section 4.2.1). Hence, the only way for the CHP plant to continue operating would be to switch to hydrogen or deploy CCUS.
• A preference for hydrogen was expressed by operators of processes with very high heat demands who are sceptical around the ability of the electricity grid to meet their demands. More analysis would be required to assess this.

In conclusion, it is noted that the results presented in the preceding sections provide sufficient information to evaluate the cost and infrastructure requirements for a potential hybrid pathway, which will necessarily lie somewhere in-between those estimated for the Electrification and Hydrogen pathway.