Alternative Fuel Switching Technologies For The Glass Sector - Phase - 2 - 2019 - Final - Report

BEIS Industrial Fuel Switching Phase 2
Alternative Fuel Switching Technologies

for the Glass Sector
November 2019
Final Report
Date: November 2019
Authors:
R Ireson (GTS), A Fuller (Glass Futures Ltd.), J Woods (Inperpetuum), R Simon (Element
Energy), G Andrews (University of Leeds), P Bingham (Sheffield Hallam University), S Hakes
(FIC), M Davies (Tecoglas)
Alternative Fuel Switching Technologies for the Glass Sector
Table of Contents
1 Executive Summary 3
1.1 Introduction: Why the UK needs the glass industry to switch to low carbon fuels 3
1.2 Why more than one solution is required 4
1.3 Technical opportunities 4
1.4 Timescales and findings 5
2 Introduction 6
2.1 Remit of Glass Futures 6
2.2 Background 6
2.3 BEIS Industrial Fuel Switching competition remit 6
2.4 Scope of Study 7
2.5 Approach 7
2.5.1 Literature reviews 8
2.5.2 Industry and supply chain engagement workshops and SWOT analysis 8
2.5.3 Direct Industry Engagement 8
2.5.4 Energy networks and suppliers 8
2.5.5 Engagement with wider glass community and other research groups 8
2.5.6 Economic modelling 8
3 Current state of art 9

3.1 Overview of glass manufacturing 9
3.2 Energy consumption and emissions 10
3.3 Diesel and fuel oils 13
3.4 Oxyfuel furnaces 13
3.5 Praxair, Optimelt 13
4 Overview of most attractive Low carbon Fuel Switching technologies 14

4.1 Biofuels 14
4.1.1 Biodiesel 14
4.1.2 Lower grade bio-based fuels 15
4.1.3 Availability of biofuels 15
4.1.4 Potential for reducing CO2 emissions 18
4.2 Hydrogen 19
4.2.1 Applicability of hydrogen fuel for glass melting 19
4.2.2 Availability of Hydrogen fuels 20
4.3 100% Large-scale electric melting 21
4.4 Flexible hybrid fuelled furnace 21
5 Key Challenges and Opportunities 22

5.1 Biofuels 22
5.2 Hydrogen 25
5.3 100% Electric melting 28
5.4 Flexible-Hybrid fuel scenarios 29
November 2019 1
6 Economic assessment 31
6.1 Methodology 31
6.2 Results 31
6.3 Carbon Pricing 32
6.4 Lifetime costs 32
6.5 Proposed Further Development to economic model 34
7 Timeframes and implementation 35

7.1 Biofuels 35
7.2 Hydrogen 35
7.3 Large-scale electric melting 36
7.4 Flexible Hybrid fuel scenarios 37
7.5 Compatibility with CCUS 37
7.6 Summary of possible implementation timeframe with sufficient R&D investment 38
8 Conclusions and Recommendations 39

8.1 Biofuels 39
8.2 Hydrogen 39
8.3 Large-scale 100 % Electric melting 39
8.4 Flexible Hybrid-fuel scenarios 40
8.5 Bio-methane 40
8.6 CO2 reduction potential 40
November 2019 2
1 Executive Summary
1.1 Introduction: Why the UK needs the glass industry to switch to low
carbon fuels
The UK glass sector employs 23,200 people, generates £3bn revenues and contributes £1.6bn
GVA to the UK economy (Ekosgen, 2019). The sector also makes a significant contribution to
many other sectors and to addressing the challenges facing society. Innovations in the glass
sector have the potential to benefit everyone - through improved energy efficiency in
construction, through improved ways of generating green energy, through demonstrating the
circular economy in action through the use of recycled materials, and across many sectors
through the development of novel applications benefiting medicine, agriculture, transport and
advanced manufacturing.
The core glass manufacturing industry produces around 3 million tonnes (Mt) of glass per
annum, generating more than 2 million tonnes of CO2. Of these emissions, 58% are emitted
directly from combustion of fuel, 24% come from primary generation of electricity used on site
and 18% are released from the decomposition of carbonate raw materials. (British Glass 2014).
Whilst the sector has made progress by halving emissions in the last 50 years, there is a need
to urgently accelerate efforts to increase energy efficiency and reduce CO2 emissions to meet
the UK’s 2050 carbon commitments.
As many furnaces due to be installed in the coming years will be expected to run for up to 20
years, new low carbon fuel technologies need to be proven technically and economically within
the next 10 years if the glass sector is to fully decarbonise by 2050.
November 2019 3
To address this need, Glass Futures Ltd. (GFL) successfully applied for and secured a £300k
grant under the BEIS Industrial Fuel Switching Competition Phase 2 to run the feasibility study
entitled ‘Alternative Fuel Switching Technologies for the Glass Sector‘ (which ran from April
2019 to October 2019).
A significant amount of data and feedback has been gathered during this study; the following
report provides a summary of the key findings, conclusions and recommendations.
1.2 Why more than one solution is required

There are significant differences in infrastructure across the UK glass sector in relation to
furnace design, age and specific application. All these factors will influence the most suitable
route to decarbonise a given site. There is also uncertainty over availability and economics of
fuels across the UK and this is likely to vary from region to region. As such, it is unlikely that
one fuel scenario will be able to address the decarbonisation needs of the entire UK glass
sector.
Therefore, to effectively decarbonise the entire sector as fast as possible, it is recommended

that the following four fuel-scenarios need to be investigated and developed by the Glass
Sector in order to maximise the chances of successfully decarbonising manufacturing
process by 2050:
• Biofuels
• Hydrogen
• 100% electric melting
• Hybrid-fuel scenarios
There is a strong argument to add biomethane to this list; although this fuel is considered out of
scope for the current competition and has not been investigated in this current study,
biomethane does offer potential to decarbonise glass making and so it is recommended that it
be included in future studies.
1.3 Technical opportunities

Given the glass sector’s commitment to decarbonise, the UK’s industrial strategy to support that
decarbonisation ambition and the existing research expertise within the UK in combustion
technologies, there is a significant opportunity for the UK glass sector supply chain to bring new
technical concepts to market and become a world leader in the decarbonisation of a heavy
industry. These new supply chains and processes could also provide knock-on benefits to other
sectors, such as Steel, Cement, Ceramics, Waste management and Energy generation.
Through industry engagement activities and literature reviews, this Phase 2 study has identified
that there is a great deal of interest in fuel switching within the glass sector but that are also
significant gaps in knowledge and technical barriers that need to be addressed for this to be
realised. The study also identified that the UK has both the industrial appetite, the necessary
research excellence as well as government backing to address these challenges. The technical
developments and capabilities required to decarbonise the UK glass sector therefore represent
an area of opportunity for UK based businesses and research organisations to become global
leaders in this field.
November 2019 4
1.4 Timescales and findings

The study highlighted the significant impact that the economics of fuel switching will have upon
uptake timescales. All of the four proposed solutions investigated have the potential to enable
full decarbonisation of heat required for glass melting across the UK glass industry before 2050.
However, to fulfil this timescale large scale demonstrations must occur within a relatively short
time frame (<10 years) to allow industry to make the business cases and engage new supply
chains to bring these decarbonised solutions can be brought on-line within 10-20 years. Given
the 15-20 year life expectancy for glass furnaces, these timescales are essential if the industry
is to decarbonise by 2050.
Although an economic study was undertaken, the high level of uncertainty in fuel costs resulted
in a conclusion that any of the options could be the most economically feasible option in the
future. It is important this this is reflected in the decarbonisation roadmap for the glass sector
which currently shows a heavy reliance on electrification.
There is significant concern across the industry that without significant investment now, the
ability for the industry to carry out the research it needs will be difficult. This is due to the
capitally intensive nature of the glass sector and its requirement to run uninterrupted 24 hours a
day 365 days a year which as a process, does not lend itself well to disruptive demonstrations.
The glass industry has limited R&D funds available, much of which is already committed to
product development rather than process development. Given the magnitude of research and
investment required, there is a need for a united approach across all sectors of the glass
industry with significant government backing.
November 2019 5
2 Introduction
2.1 Remit of Glass Futures
Glass Futures is a not-for-profit company, created as a core entity to develop two UK-based
“catapult-like” centres of excellence in glass comprising R&D, innovation, technology incubation
and implementation, training and up-skilling. It brings together the global glass industry and
academia. Led by some of the world’s largest glass manufacturers, supply chain partners and
leading UK university research groups, its aim is to create two centres of excellence:
• A unique multi-fuelled ‘Hot’ glass pilot facility in St Helen’s, Merseyside

• A high-tech ‘cold’ glass research centre based at the University of Leeds
These centres will be supported by a series of smaller research hubs across UK academic and
industry research groups with the aim of strengthening and aligning existing industrial and
academic expertise within the “Northern Powerhouse” region. The ultimate ambition is to create
a globally recognised UK-based hub in glass technology and manufacturing with the capability
to drive significant improvements in productivity and sustainability within the UK glass industry,
providing the platform to drive the sector towards net-zero CO2 emissions by 2050.
2.2 Background
The UK Government has committed to reduce net carbon dioxide emissions to zero by 2050.
National efforts to meet these emissions reduction targets could potentially result in conversion
to a hydrogen gas grid, or alternatively could see localised decommissioning of the gas grid and
a move towards electrification and decentralised energy supply.
It is estimated that 90% of UK industry relies on energy supplied from the gas grid either directly
for their industrial processes or indirectly in the day-to-day business. Whilst the glass sector has
made progress by halving emissions in the last 50 years and its products contribute to energy
savings in other sectors (e.g. glazing and insulation, wind turbines, aerospace), there is a need
to urgently accelerate efforts to increase energy efficiency and reduce CO2 emissions to meet
the UK’s 2050 carbon commitments.
The BEIS Glass Industry Decarbonisation and Energy Efficiency Roadmap to 2050, identified
100% electric melting as the preferred route to decarbonise the industry. However, findings
from subsequent industrial engagement activities have identified other technologies that can
now be considered as real alternative routes to decarbonisation, such as biofuels and hydrogen,
that were not highlighted on the original industry roadmap. This Phase 2 study therefore looked
to build upon the original findings to increase understanding those different options and explore
how to facilitate wide-scale adoption of all low-carbon fuel scenarios across the glass sector.
2.3 BEIS Industrial Fuel Switching competition remit

BEIS have stated that the aim of the Industrial Fuel Switching Competition is to identify and
demonstrate solutions which will enable fuel switching in industry from fossil fuels to less carbon
intensive fuels. Fuels in scope include electrification, hydrogen and biomass (whilst biomethane
is a lower carbon fuel, it is not in scope for this competition). The Competition was split into
three phases: Phase 1 was a market engagement and assessment study into fuel switching in
the UK, Phase 2 is feasibility study into a fuel switching solutions (of which this report relates),
and Phase 3 will fund demonstration of these solutions.
November 2019 6
In Phase 1, the market engagement and assessment study conducted by Element Energy
considered the viable energy sources for industrial fuel switching, the industrial processes
compatible with fuel switching, and the potential solutions to achieve these switches; the
complete report will be available at: www.gov.uk/guidance/funding-for-low-carbon-industry
BEIS’ stated aim of Phase 2 was to identify and test the processes and technologies required
for industries in the UK to switch to low carbon fuels, providing funding for the consortium to
demonstrate the feasibility of their proposed technology or approach.
2.4 Scope of Study

In preparation for the Phase 2 bid, Glass Futures undertook a detailed review of the original
glass industry decarbonisation roadmap, alongside discussions with a number of industrial and
academic partners.
These discussions highlighted that, due to differences between manufacturing requirements of
subsectors (float, container and glass fibre), capital refurbishment time tables and predicted
future variations in availability and affordability of different fuels across the UK (e.g. hydrogen
supply may be localised; local grid capacity for electricity supply limited) no single low-carbon
fuel scenario is likely to be suitable for all glass manufacturing processes. It was therefore
agreed that the scope of the Phase 2 study should cover the following four low-carbon fuel
scenarios:
• Biofuels (with potential for subsequent addition of carbon capture utilisation and storage
(CCUS))
• Hydrogen
• 100% large-scale electric melting
• Hybrid fuel scenarios (i.e. combinations of the above with/without natural gas)
There is a strong argument to add biomethane to this list, however this fuel is considered out of
scope and has not been investigated in this current study.
2.5 Approach
Led by Glass Futures and project managed by Glass Technology Services Ltd. (GTS), the
Phase 2 study consisted of five work packages (see Figure 1 below) and was supported by
project partners representing glass manufacturers (Encirc, NSG Pilkington), furnace designers
(F.I.C, Tecoglas), control systems supplier (Siemens), research groups (Sheffield Hallam
University, University of Leeds) and The Society for Glass Technology. Further support was
provided by Element Energy, In Perpetuum, CelSian and the University of Sheffield.
November 2019 7
Figure 1 Inter-relationship of work packages within Phase 2 study
The Phase 2 study aimed to determine technical and economic feasibility for each of the above
four fuel scenarios along with the potential to decarbonise the glass furnace heating process
whilst meeting regulatory requirements.
2.5.1 Literature reviews

GTS, Element Energy, UoL and SHU undertook detailed literature reviews into each fuel
scenario to build a foundational understanding of the current state-of-art in each field. Due to
the substantial nature of this study, these will be published as a separate report, hopefully in the
form of a published academic paper.
2.5.2 Industry and supply chain engagement workshops and SWOT analysis
Three workshops were held (April, July, September), to which academic and industrial partners
were invited. In addition to this, two further workshops were held to explore the strengths,
weaknesses, opportunities and threats to each of the four fuel scenarios. The findings are
presented in Section 5.
2.5.3 Direct Industry Engagement

In addition to the workshops, numerous one-to-one meetings were held with industry
organisations, including meetings and/or e-mail correspondence with all major UK glass
manufacturers and several supply chain partners, many of whom were invited to join the Phase
3 project. These meetings provided valuable insights into the current views of the glass sector
and the scope to implement new low carbon fuel technologies.
2.5.4 Energy networks and suppliers

In order to build a more detailed understanding of current supply network capabilities,
discussions were held with various suppliers of energy and gases, including BOC, Cadent and
Northern Power Grid.
2.5.5 Engagement with wider glass community and other research groups
The Project team also engaged the research groups working in the field of glass technology and
alternative combustion technologies. This provided insights into related research programmes
being undertaken across the wider glass sector, ensuring Glass Futures activities will
complement these research programmes (and not duplicate work). It also provided valuable
opportunities to build partnerships that will lay the foundations for future collaborative research
programmes. The following groups were directly engaged within the Phase 2 study:
• University of Sheffield, PACT • British Glass

• DNV-GL • IPGR
• CelSian • Supergen
• Glass Trend • Progressive Energy
2.5.6 Economic modelling
A high-level economic review of each fuel switch scenario was modelled by Element Energy to
build a more thorough understanding of the potential life-time costs for each fuel scenario. The
findings are presented in Section 6.
November 2019 8
3 Current state of art

3.1 Overview of glass manufacturing
Glass is produced from sands and other minerals that are melted at very high temperatures to
form a material that has found application in a range of sectors e.g. construction, packaging,
pharmaceuticals, automotive, fibre optics and other specialist applications in both nuclear and
oil and gas to name a few.
Around 60% of all UK glass production is classed as ‘hollow glass’ that is glass packaging
containers, used within the food and drink sector. A further 30% of glass output is flat glass,
largely used by the construction and automotive sectors. The final 10% of glass manufacture
consists of fibreglass and speciality glasses (lighting, oven hobs, optical, medical and scientific
uses). The total value of all glass sectors to the UK economy is estimated to be over £1.3
billion 1
The vast majority of glass produced (>95%) can be recycled; the recycled glass (also known as
cullet) is added back into the furnace as a raw material to help decrease the energy
consumption of the process and reduce the need for raw materials. It has been estimated that
every tonne of recycled glass used in glass manufacture saves approximately 320kWh 2 of
natural gas. In addition to the CO2 savings from the reduction in fuel use, there is a CO2 saving
associated with the reduction in use of carbonate raw materials. For every tonne of cullet used
it is estimated that 250kg less CO2 is emitted compared with using virgin raw materials 3
A range of processes can be used to produce glass articles from molten glass into its final form
and shape, including drawing, blowing, pressing and floating. The physical and chemical
properties of glass vary depending on the formulation of the material; however, the methods
used to shape and form the glass as it cools will also have a significant impact on the final
physical properties of the glass. The glass manufacturing process starts with the batch
preparation. Sand, limestone/dolomite, soda ash and minor additives are weighed and mixed
according to the glass formulation; the resulting mix of raw materials is referred to as the ‘batch’.
This glass batch usually includes a percentage of cullet and this mix is then conveyed to a batch
storage bin where the blend is held before being fed to the melting furnace.
The batch blend is charged to the melting furnace that operates at temperatures ranging from
1550 – 1600 °C. In this stage the materials should go through melting, refining, homogenising
and thermal conditioning before leaving the furnace. Melting starts when the batch is charged
into the furnace and should optimally be completed in the first half of the melting chamber.
Several factors affect the rate at which the material is melted such as the temperature in the
chamber, the grain size of the batch materials, the amount of cullet in the mixture and the
homogeneity of the batch. As the molten glass goes through the furnace it reaches the second
half of the chamber where the refining (or fining) process removes gas bubbles formed during
melting of the raw materials. The homogenising of the molten glass occurs throughout the
furnace and is intended to eliminate variations in its properties such as refractive index, density
or coefficient of expansion. It typically takes 24 hours for the glass to go through a container
furnace, however it can take up to 2-3 times longer for the glass to travel through a float
furnace, due to the more stringent quality requirements on appearance. Figure 2 provides an
example of the various steps used in the manufacture of float glass.
1 Prodcom sales figures, 2015, Office of National Statistics

2 Case Number 2003-03-082, Glass Technology Services, 2004
3 Carbon Trust, 2005
November 2019 9
Figure 2 Manufacturing process of float glass 4
There are several glass melting furnace designs most of which are distant relatives of the
1860’s Siemens regenerative furnace designs with natural gas as the primary fuel source. Some
historic context is important within the glassmaking industry as the early designs all used coal
as the primary fuel source. Through developments the primary fuel source across the UK
switched to heavy fuel oil which in turn switched to natural gas due primarily to economics.
Throughout the course of these changes the core design principles of the glass making furnace
have changed very little due to the relatively high efficiency offered by the regenerative furnace
design. Alternative recuperative designs are less efficient at a large-scale and as such do not
see widespread use across the large-scale commercial glass making sector.
In the UK most furnaces are of the Siemens regenerative type, which can be either side-port or
end-port and operate with natural gas as the primary fuel source, however, many current UK
furnaces have the capability to fire diesel, primarily to provide energy security for the large
capital assets. The combustion of natural gas is attained with air and in some cases enriched air
or oxy-combustion are utilised. The fuel is received in the facilities from the natural gas grid and
its pressure is regulated to ca. 1 Bar relative pressure. Combustion air is preheated on the hot
side of the thermal regenerator of the furnace to around 1200°C, which our study has
highlighted is almost unique to the glass sector. Combustion temperatures are in excess of
2000°C and the flue gases leave the furnace chamber at circa 1400°C where the waste heat is
recovered on the cold side of the thermal regenerator, this process reverses every 20-30
minutes and as such the heat from the exhaust gases is mechanically stored and recovered in
highly specialised refractories.
After the melting stage the glass is cooled to around 1100°C and sent to the forming stage; this
step is specific and individual to each type of glass product. For float glass as an example, the
molten glass flows over a tin bath on which the glass sheets are formed and drawn away from
the furnace. For glass containers, the glass is sheared into ‘gobs’ of glass which are fed into an
Individual Section forming machine. For glass fibres, the glass is fed into a platinum bushing
from which the fine fibres are pulled.
3.2 Energy consumption and emissions

Glass manufacture is an energy–intensive process, primarily due to the large amount of energy
required to melt and refine the glass, with an annual consumption of 9 TWh 5. The most common
4 Figure credit: University of Leeds
November 2019 10
furnace types across the UK are fuelled with natural gas with some additional electrical boost.
Table 1 shows the statistics on energy consumption in the manufacture of glass in the UK per
type of energy source. It can be seen that the highest share of energy consumption corresponds
to natural gas, totalising 517 ktoe, which represents 76.5% of the total energy consumption by
this industry. No less important is the consumption of electricity totalling 150 ktoe and is mainly
used in the process for handling of raw materials and products and to provide the energy to fans
and blowers for glass cooling.
Many glass plants in the UK use an electrical boost system, where electrical energy is delivered
through a molybdenum electrode fully submerged in the glass melt. This accounts for a
relatively small percentage of energy delivered to the melts as opposed to natural gas. Electric
boost is used to:
• Provide additional pull giving flexibility of operation and on-the-run expansion. This
includes melting both dark glasses like amber and green (where the majority of the
radiative heat from the natural gas flame is absorbed close to the surface) but also
Flint/Clear.
• Improve glass quality, depending on specific quality issues, not all quality issues can be
fixed.
• Reduce emissions from furnace (through reducing the natural gas requirements)
• Energy substitution (e.g. if cheap electricity is available at certain times of the day/year)
• Provide a back-up heating system, for example adding boost to enable maintenance of
pull whilst organising a regenerator repair to assist furnace breathing.
Gas oil Fuel oil Natural gas Electricity Total

Manufacture of flat glass - 58.2 1709.6 186.1 1965.
5
Shaping and processing of flat 11.6 0.0 69.8 383.8 465.2
glass
Manufacture of hollow glass 11.6 23.3 3140.1 802.5 3965.
8
Manufacture of glass fibres 0.0 0.0 721.1 267.5 988.6
Manufacture and processing of - 0.0 372.2 104.7 476.8
other glass, including technical
glassware
Table 1: Energy consumption in the UK Glass Industry in 2018 in thousands of tonnes of oil equivalent (GWh) 6.
5 British Glass. A Clear Future: UK Glass Manufacturing Sector Decarbonisation Roadmap to 2050. [Online]. 2014.
[Accessed 10 June 2019]. Available from: https://www.britglass.org.uk

6 UK Department of Business Energy & Industrial Strategy. Energy Consumption in the UK: 2018 update. [Online].
2019. [Accessed 10 July 2019]. Available from: https://www.gov.uk/government/statistics/energy-consumption-in-the-

uk
November 2019 11
British Glass estimated the total emissions in 2012 to be 2 million of tonne of CO2 7, most of
which come from the melting process, representing approximately 70%. Other sources of
emissions are the degradation of raw materials (CO2 is released from soda ash, dolomite and
limestone) and the electricity used for e.g. compressors, cooling fans, drives and downstream
processing equipment. Figure 3 provides a high level summary of the main sources of CO2
emissions.
Figure 3 Overview of glass process and original of CO2 emissions (British Glass, 2014)
7 British Glass. A Clear Future: UK Glass Manufacturing Sector Decarbonisation Roadmap to 2050. [Online]. 014.
[Accessed 10 June 2019]. Available from: https://www.britglass.org.uk
November 2019 12
3.3 Diesel and fuel oils

While the delivery method for fuel oils into the combustion chamber is very similar to natural
gas, there are a few distinct differences between fuels, and these typically come in the form of
on-site storage and a higher cost per kWh.
While no UK manufacturer regularly fires using fuel-oil anymore, the use of stored diesel is still
prevalent across a number of UK Glass manufacturing sites to ensure that operators can
maintain the value of their primary capital asset (furnace) in the event of an emergency with the
gas delivery network.
3.4 Oxyfuel furnaces

An oxyfuel furnace fires pure oxygen into the furnace along with the natural gas (i.e. with no
ambient air) and so are noticeably more thermally efficient than furnaces that use ambient air.
Although the flame is hotter, no nitrogen is fed into the furnace and so NOx levels will be very
low. Most oxyfuel furnaces are used within specialist and specific applications which are
generally under the following situations:
• Melting of speciality or high temperature glasses (Borosilicate for instance)

• Where an economic source of oxygen exists
• To aid an old furnace to meet its campaign requirements
The technical aspects of oxyfuel firing are well understood and has some distinct benefits but
comes at a reasonable cost for either on site oxygen manufacture or regular deliveries of liquid
oxygen. It is also notable that due to the energy demand of producing oxygen there is generally
low CO2 reductions from the use of oxy-fuel furnaces.
3.5 Praxair, Optimelt

Across Europe there are a number of designs utilising waste heat as a method to produce
syngas. While these are relatively new to the market, there are only two commercial furnaces in
the EU operating with this technology. The Optimelt system has a number of benefits, including
significant improvements in fuel efficiency, however it is still seen as a technology that requires
a significant capital investment. It should also be noted that although the Praxair system is
more efficient than standard furnaces, it is reliant on a natural gas fuel supply and so is not a
low carbon option.
November 2019 13
4 Overview of most attractive Low carbon Fuel Switching

technologies
The following section provides a brief overview of each of the fuel scenario investigated within
the Phase 2 study.
4.1 Biofuels
Biofuels are fuels derived from biomass. If combined with carbon capture technologies, biofuels
offer a route to net-negative CO2 emissions (BECCS) from glass manufacturing processes.
Biofuels typically burn with a more radiant flame, have lower CV content per kg and can contain
higher moisture content than natural gas or hydrogen fuels, and therefore are expected to have
a higher heat transfer from the flame into the glass melt. Burning at a lower flame temperature,
biofuels are also likely to emit lower NOx levels.
4.1.1 Biodiesel
The most common back-up fuel used by UK glass production sites is fossil gas oil, also known
as diesel. With an equivalent specification, biodiesel from 100% renewable sources (i.e. no
blending with conventional diesel) would provide a low-risk switch. Although most conventional
diesel used in transport and industrial heating applications contains approximately 7% biodiesel
blended into the fossil-fuel derived diesel oil (and so the industry has used fuels with low-levels
of biodiesel), the glass industry has not yet explored the performance of pure, 100% biodiesel
fuel, and how it compares to conventional diesel, and so such a transition is currently deemed
as high risk. Figure 5 provides an outline of a typical process for producing biodiesel.
Figure 5 Manufacturing process of a typical base-catalysed process for producing biodiesel; in this case the
pre-processing of Waste Oil is shown in the dotted outline at the top left of the diagram 8
8 Image courtesy Inperpetuum Partners LLP
November 2019 14
4.1.2 Lower grade bio-based fuels

Lower-grade bio-derived fuels (e.g. pyrolysis of carbon-based bio-wastes) may provide a lower-
cost fuel solution than biodiesel, which may even be economically competitive with natural gas.
As such, the UK glass industry may have some real options for utilising demonstrably
sustainable liquid fuels in the short, medium and longer term.
There is some conflict between UK policies for heat, power and transport with regards to use of
crops from land and this shows some room for consideration. There are various sustainability
schemes that can be adopted to demonstrate compliance with suitable sustainability standards,
although sustainability is something that would need investigating further, both in terms of
available volumes of such carbon-based wastes and the potential knock-on impact on other
sectors which might have to switch to non-sustainable fuels and feedstocks if the glass sector
puts pressure on the supply of sustainable sources of biomass or carbon-based waste streams.
4.1.3 Availability of biofuels

UK volumes of biodiesel are largely focused on the use of wastes and residues from various
sources:
• Food supply chains: specifically cooking oil turned into used cooking oil methyl ester
(UCOME). There is also the potential use of animal fat or tallow that can be turned into a
biodiesel known as tallow methyl ester (TME). Finally, there are a range of greases and
fats collected from domestic sources via the sewers through fat traps in the sewers and
from waste-water cleaning operations. These are also known as UCOME biodiesel fuels.
Similar feedstocks can also be used to make HVO fuels.
• Oil bearing crops like oil seed rape or palm oil as well as starch crops such as wheat or
maize can be converted into traditional biofuels and hydro-treated vegetable oil (HVO).
However, due to fuel versus food debates and concerns over sustainability (e.g. over
links to deforestation), legislators do not consider these types of feedstocks as
sustainable fossil fuel replacements at this time.
• Waste and residue plastics and tyres from various supply chains could be converted into
fuels. Such conversions may be appropriate if the fuel supply route turned out to be
economically competitive in the long run compared to comparative fuels. Although,
unless the wastes were derived from renewable sources, such fuels would need to be
used in conjunction with CCUS technology to be classed as a low-carbon source.
November 2019 15
Figure 6.1 sources of Biofuels used in the UK 9
Figure 6.2 sources of Biofuels used in the UK by country 10
9 Source: DfT (covering period 15th April 2016 – 14th April 2017)
10 Source: DfT (covering period 15th April 2016 – 14th April 2017)
November 2019 16
Figure 6.3 Types of Biofuel used in the UK 11
Our study estimated that the current UK glass sector would require annual volumes of 400
million litres of biodiesel. Current biodiesel supply in the UK is at around 804 million litres 12,
however the production of biodiesel has been stagnating in previous years mainly because of
the level of supply being limited by the Renewable Transport Fuel Obligation (RTFO); the lack of
policy or market incentives has led to loss of confidence from investors which has resulted in
facilities operating below installed capacity. Investigations by Inperpetuum identified that many
biodiesel facilities are running below maximum capacity due to the decline in numbers of diesel
vehicles and that demand is falling by approximately 300,000 litres year on year.
Therefore, the consensus between producers and studies is that there is scope to create
sufficient feedstock accessibility to meet a foreseeable uplift in supplying 100% of the fuel needs
for the glass sector, however, this may be affected if competing industries emerge. E4tech
reported the availability of feedstock will become more constrained in the period to 2030, and
consequently, the long-term deployment of biofuels plants in the UK would need to rely more on
feedstock imports or switching feedstock use from power to biofuel applications. However, in
the short term, through conversations with the CEO for Global Biofuels at Greenergy, the
Director of Corporate Affairs at Argent Energy and the Commercial Director at Mabanaft Ltd, it
was identified that if the UK Glass Industry market needed volumes of vegetable oil or waste
derived biodiesel, then they would all want to be supplying this finished product and would be
able to source the needed feedstocks.
There is a very low fuel duty payable on biodiesel used in industrial heating. Due to legislators
and policy influencers pushing markets away from food / feed crops for use in biodiesel, it is our
view that waste feedstock biodiesel, such as tallow and used cooking oil (UCO), should be
preferred to reduce the CO2 impact of the fossil derived product. The use of tallow and UCO in
the production of biodiesel is now commonplace within the UK with the two largest producers
being Argent Energy, using tallow / UCO and Greenergy, using UCO.
Due to increases in demand in recent years, within the UK and EU, the feedstock and the
finished Biodiesel are starting to be imported on a more regular basis and the volumes growing
12 DfT: Renewable Fuel Statistics 2019 First Provisional Report
November 2019 17
dramatically, showing that if the market for this type of product exists the global market will
supply. Biodiesel producers, traders and blenders have seen growth in demand, but expect
biodiesel demand to fall in line with Department for Transport forecasts as diesel demand falls in
the UK and across Europe. This is in line with forecasted increase in adoption of electric
vehicles within the passenger car fleet and potential shift to bio methane use for heavy goods
vehicles. Below is the view of the UK Department for Transport (the blue box is the biodiesel
that we suggest using in UK Glass Industry):
Figure 7 potential decline in future biodiesel use
Biodiesel trades as a commodity and has shown over many years how it can be affected by
policy levers in different geographies. Biodiesel price has however fallen relative to crude oil
over recent years despite strong growth in demand. This is understood to be due to increase in
availability of feedstocks from wastes that has expanded to include large levels of imports to the
European Union.
4.1.4 Potential for reducing CO2 emissions

The potential CO2 emissions reductions for fuel switching within the Glass industry are shown in
Figure 8 below, which gives an estimation of the relative CO2 emissions savings associated with
switching from natural gas to heating oil and biodiesel, as well as from heating oil to biodiesel.
The chart indicates that a move from main natural gas to fossil derived heating oil would result
in an increase in CO2 emissions by 42%. However, a move from natural gas to biodiesel would
see a 91% reduction in CO2 emissions.
November 2019 18
Figure 8 Comparison of net CO2 content of natural gas, heating oil and biodiesel fuels
4.2 Hydrogen
A number of groups in the UK are exploring the possibility of converting the natural gas network
to 100% hydrogen as a route to decarbonise both industrial and domestic energy applications.
If a ‘decarbonised’ hydrogen supply could be delivered directly to site through existing natural
gas pipelines, it could provide an ideal opportunity to decarbonise the glass manufacturing
process.
4.2.1 Applicability of hydrogen fuel for glass melting

The substitution of natural gas with hydrogen to fire the glass furnace is subjected to the
feasibility of hydrogen to provide the radiative heat transfer required by the melting process. At
atmospheric conditions hydrogen has a gross calorific value of 12.1 MJ/m3 and a relative
density of 0.0696, which is equivalent to a Wobbe Number of 45.88 MJ/m3. Natural gas used in
the UK must currently have a Wobbe Number in the range 47.20 – 51.41 MJ/m3; therefore, to
achieve the same heat release a slightly higher hydrogen mass flowrate is required.
The substitution of methane by hydrogen has barely been investigated by the glass
manufacturers due to the poor radiative properties of the hydrogen flame and combustion
products. The literature reviews undertaken by academic partners UoL, SHU and UoS only
identified one reference to investigations on glass melting with hydrogen.
Although there is minimal evidence of the application of hydrogen in glass furnaces, there are a
number of research groups developing hydrogen burners that are suitable for glass furnaces
(e.g. the Flamatec division of Glass Service). Andrews et al. 13 addressed the use of pure
hydrogen using a rapidly mixed jet burner design and demonstrated the feasibility of operating a
combustor with low NOx emissions. They showed that to achieve the desired level of NOx in the
13
Andrews, G.E., Altaher, M.A. and Li, H. Hydrogen Combustion at High Combustor Airflow Using an Impinging Jet
Flame Stabiliser with No Flashback and Low NOx. In: ASME Turbo Expo 2012: Turbine Technical Conference and
Exposition, 2013, pp.1479-1489.
November 2019 19
exhaust, very lean mixtures are required. In addition, the flame stability attained with hydrogen
was shown to be higher than that of both, direct and premixed propane. They showed that NOx
emissions lower than 25 ppm are achievable with hydrogen up to temperatures of 1600 K.
However, these temperature levels are lower than the typical melting temperatures in glass
furnaces.
It has also been speculated that alternative solutions to enhance the radiative properties of
hydrogen in the glass melting furnace might be the use of hydrogen/natural gas or
hydrogen/biogas co-firing schemes or the injection of other additives that increase the
luminosity of the flame, without damaging the glass quality or the furnace integrity.
4.2.2 Availability of Hydrogen fuels

If a fuel switching scenario with hydrogen as fuel is technically feasible from the process
standpoint, the question remains on how to produce it. According to the report of the Energy
Research Partnership (ERP) on the potential role of hydrogen, the UK production is about 26.9
TWh/yr, with half coming as a by-product from the industry, and being used onsite or sold as
chemical feedstock 14.
The two main routes within the UK to produce low carbon hydrogen on a large scale are steam
methane reformation (SMR) and electrolysis of water using renewable electricity.
The demand of hydrogen as a fuel can be met at the early stages of the transition using SMR,
typically referred to as ‘blue’ hydrogen. Currently, 48% of the production of hydrogen is attained
via SMR as it is a mature and reliable technology, with a lower cost in comparison to
electrolysis 15. In order to contribute to glass decarbonisation, hydrogen production needs to be
low or zero carbon; most of the modern steam reformers can achieve high efficiency levels that
reduce the CO2 emissions, however, further reduction will only be achieved by the integration of
CO2 capture and storage/utilisation in the process scheme.
The addition of a CCS unit to the SMR process increases the CAPEX and OPEX of the facility,
the International Energy Agency (IEA) 16, investigated the economics of deploying CCS in a
SMR based hydrogen plant and concluded that the addition of CO2 capture would increase the
plant cost by 18% to 79%.
Hydrogen produced using renewable sources is referred to as ‘green’ hydrogen. The obvious
choices are to produce hydrogen either centrally or distributed by means of wind or solar energy
and electrolysis, or to produce it through steam reforming of bio-sources.
The production of hydrogen via wind energy represents a more likely scenario than the use of
solar energy. In 2018, the solar energy installed capacity was 13 GW, which represented a
share of 3.9% of the total electricity generation in the UK, compared to 7.9GW installed wind
power. However, the support from the UK government to solar power projects has waned since
2016 when it announced the suspension to subsidies for renewable energy projects and wind
capacity is increasing at more than twice the rate of solar. As such the case for wind energy is a
better one.
14 Energy Research Partnership. Potential Role of Hydrogen in the UK Energy System. [Online]. 2016. [Accessed 10
June 2019].
June 2019].
16 International Energy Agency. Techno-Economic evaluation of SMR based standalone (Merchant) Hydrogen Plant
with CSS. [Online]. 2017. [Accessed 10 October 2019]. Available from: https://ieaghg.org/exco_docs/2017-02.pdf
November 2019 20
Several questions remain with respect to the use of solar or wind energy in the production of
hydrogen. The applicability of this option depends on the availability of a surplus of low-price
electricity to reduce the OPEX of electrolysers. This surplus, however, depends on certain
parameters like the seasonal weather or the design characteristics of the wind turbines (e.g.
height of the hub). Although it is expected that the CAPEX of electrolysers will decrease with
time, the uncertainty of the availability of electricity surplus would lead to scenarios where
electrolysers operate with low load factors or at reduced efficiencies which would create capital
burden and, possibly heat management or safety issues 17.
If projects such as HyNet and H21 North of England are successful, the UK glass industry would
have access to 100% supply of low-carbon hydrogen (i.e. hydrogen with no or very low net CO2
emissions associated with its production) to site through existing natural gas pipelines.
Therefore, the UK glass industry needs to be ready to respond to this scenario.
4.3 100% Large-scale electric melting

Electric melting is well established and significantly more efficient than equivalent heating
technologies that rely on combustion and would offer an ideal route to decarbonise glass
melting at the point where 100% ‘green’ electricity is available. Existing all-electric melting is
considerably more energy efficient than comparable sized fossil fuel fired furnaces so
developing larger units is very desirable for fuel efficiency and decarbonisation.
The largest commercial electric furnaces available have a capacity of upwards of 300t/day;
significant further developments in furnace design are required if this is to provide a low carbon
replacement for new furnaces that can have a capacity of up to 900t/day. CFD modelling has
demonstrated that large all-electric furnaces are possible but not with conventional vertical
melting. Horizontal electrical melting looks promising but demonstrating it is essential before it
will be adopted. The additional advantage of the all-electric horizontal melter is that it utilises an
almost identical footprint as existing furnaces.
The technical feasibility of 100% electric furnaces is promising, however there are significant
economic barriers to its uptake. The greatest concerns relate to the future economic viability of
using electricity, primarily due to the higher cost of electricity compared to other fuels in the UK,
but also due to the CAPEX costs associated with upgrading site infrastructure. The challenges
and costs associated with upgrading the electricity supply to site are also significant and
represent a major challenge for the UK glass sector.
4.4 Flexible hybrid fuelled furnace

As discussed in Section 3, many glass furnaces have the ability to use electric boost, to
complement the natural gas firing. It was identified that this concept could be extended to
include furnace designs powered by a range of different fuels (e.g. the ability to switch from
100% natural gas to 100% electric to 100% hydrogen or biofuels, and any combination in
between). This could offer greater protection against fluctuations in fuel prices or supply
disruptions, whilst facilitating the transition from natural gas towards new low-carbon fuels.
A flexible-hybrid design also offers the advantage in future to allow for more dynamic fuel
switching; for example, a furnace could dynamically switch from electric to biofuels in order to
support future smart load balancing networks (effectively acting as a dynamic battery come
power plant).
June 2019].
November 2019 21
5 Key Challenges and Opportunities

5.1 Biofuels
Our study did not identify any references to trials of biofuels within the glass sector beyond
small lab-scale experiments.
Adhering to the same standard as conventional diesel-fuel, biodiesel should be compatible with
existing glass furnace infrastructure with minimum additional CAPEX investment required (with
the caveat that port design will need to be changed). Although changes to port design and
burner assemblies would be required to convert from natural gas to biodiesel in order to achieve
optimum fuel efficiency, the time and costs associated with this are minimal compared to
hydrogen.
As discussed in Section 4, it should be possible for the UK biofuel sector to up-scale production
in order to supply the entire glass sector, if needed. Biodiesel therefore offers a low-risk route to
decarbonise UK glass manufacturing with minimal changes to current infrastructure and
relatively low risk to production. The emissivity of the combustion flame from biodiesel (and
also other bio-oils) is also well aligned to that required for good transfer of heat to the glass melt
(diesel is up to 5% more efficient than natural gas18 and also likely to be more efficient than
hydrogen (although this is yet to be confirmed)). A further benefit is that a fuel-oil burns at a
lower temperature than natural gas (similar is expected for biodiesel and bio-oils), so should
theoretically reduce NOx emissions however, again, a lack of any data in this area highlights the
need to carry out further testing. The main barriers to implementation are a lack of data to
demonstrate that 100% biodiesel will not adversely impact furnace infrastructure, nor glass
melting.
Historically, glass furnaces have been operated on much lower grade fuels than diesel (such as
heavy fuel oils and even coal dust). Whilst investigating biofuels, the current study identified a
whole series of low-grade oils from sources such as the pyrolysis of carbon-based wastes such
as tyres and plastics, that could be used in a glass furnace (although such products are
currently manufactured from fossil-fuel sources they do at least reduce the demand on
extraction of virgin fossil fuels, and in the future such products may well be manufactured from
renewable sources). However, the performance of these fuels when combusted in a glass
furnace is much less well known and so industrial trials would be much riskier than biodiesel.
It is therefore recommended that a series of lab-studies should be undertaken to understand the
performance of such fuels. In parallel, industrial trials of biodiesel would provide a lower risk
route to build confidence in using biofuels within the glass sector, thus offering a stepping-stone
to the use of these lower-grade bio-oils (e.g. derived from waste) which have potential to be
economically competitive with natural gas.
One of the main challenges with biofuels is that the combustion process will require careful
control so as not to leave unburnt carbon (which could create a reducing atmosphere affecting
the redox state of the glass chemistry and colour as well as emissions), although it should be
noted that the glass industry has a great deal of experience at managing such challenges.
Sustainability is also a potential issue with biofuels and can be difficult to demonstrate when
feedstocks are imported from outside the EU. Even if the chosen biofuel is from a sustainable
source, care must be taken to ensure that the glass industry does not create a deficit within
another sector that then needs to be met through creation of additional non-sustainable biomass
18 [Industry data, British Glass, Glass BREF Section 4.8]
November 2019 22
sources (e.g. the cooking oil used to make biodiesel is often used in animal feeds in parts of
Asia and so removing large volumes of cooking oil from the supply chain may result in a greater
demand for non-sustainable palm oil). This issue will need to be assessed in any future studies
that explore the use of biofuels within the glass sector.
Compared to large-sale electric melting, biofuels offer a lower cost route to decarbonise existing
glass furnaces, with significantly less investment required into new infrastructure (furnace and
supply-side) and therefore should be easier and faster to implement. Switching the glass
industry to biofuels would also reduce competition low carbon electricity capacity to decarbonise
other sectors (e.g. transport). CCUS may be more challenging for biofuels than for electric
melting (where the only emissions are the release of CO2 and other gases from the
decomposition of the carbonate-based raw materials), as the volumes of CO2 and other gases
will be greater, however the greater volumes of gases may offer economies of scale and if
CCUS can be applied, biofuels offer a carbon-negative glass-melting process.
The following table provides a summary of the main considerations that need to be considered
by a glass manufacturer looking to switch to biofuels.
Biofuels
Factor Impact of alternative energy Scalability potential & mitigation

source measures
Impact All biofuels should be safer to handle Biofuels could easily be adopted across
on H&S than natural gas and comparable to industry in a short timeframe given
diesel, with lower risk of explosion current infrastructure, with minimal
compared to natural gas. Current additional H&S training required. All sites
infrastructure and practices across handle diesel for transport so current
the UK glass sector are well systems are well equipped to handle
equipped to handle liquid biofuels large biofuel volumes.
and no changes to existing H&S
processes should be required.
Some investigations into H&S
impacts of contaminants will be
required, which may raise additional
H&S measures (e.g. biohazard risks
if fuels are derived from animals).
Impact on Emissions from biofuels should be Based upon the levels of emissions
regulatory similar to those of a diesel-oil measured the current UK abatement
measures/ powered furnace, possibly containing infrastructure should be able to deal with
Environment additional components such as emissions from biofuels and so minimal
alkalis and chlorine which may additional abatement equipment will be
require more sophisticated required.
abatement systems. Although Most sites have back-up diesel tanks
comparison with diesel suggests it which may require minor modifications to
should meet regulatory measures, cope with biofuels; additional fuel tanks
biodiesel is as-yet unproven at scale may also be required if running 100%
and so there is a risk of unseen biofuels, but this should be relatively low
regulatory issues arising. The cost and quick to implement.
literature review indicates that overall Current infrastructure for managing
SOx and NOx emissions are expected diesel spillages should be more than
to be lower. adequate to eliminate risk of biodiesel
The Phase 3 project will provide escaping into local watercourses.
quantitative analysis of emissions to
November 2019 23
build on this knowledge.

Biodiesel may have greater impact
on local ecosystems if escaped into
watercourses (e.g. higher biological
oxygen demand).
Fuel Delivery There would be increased road-traffic by a small amount to sites that already
Logistics employ large amounts of site traffic e.g.
• Largest plants (1600t/day across 2 furnaces) have up to 54 haulage
movements/day
• Fuel is an additional 5-6 tankers per day
• A 10% increase in site traffic for an overall net decrease of 91% CO2
emissions from fuel(increase in transport CO2 not considered during this
study)
Delivery may be more difficult in plants in urbanised areas, though this is not
seen as a significant issue. The GF-P2 feasibility study showed the availability of
biodiesel for the UK glass industry is good, with additional work to estimate the
availability of suitable low-grade bio-oils to be done before low-grade bio-oil/mass
rollout could occur.
The location of the glass industry relative to areas of biodiesel production are
very well placed to ensure consistent and reliable deliveries. Mains gas as a
backup would also be used to ensure continued production in the event of a fuel
logistics issue.
Production Given data from past firing setups with diesel we expect biofuels to increase
Disruption furnace efficiency by up to 5%, which would give installed UK capital
infrastructure a productivity boost if successful. Switching from natural gas may
cause short-term disruption (e.g. impact on glass colour) whilst operators adjust
to the new fuel technologies, but this is likely to be minimised through a phased
switch over (e.g. switching one burner port at a time).
Relation to The performance of biodiesel is expected to be similar to diesel type fuels already
state of the art in use as a backup and therefore swap-over is expected to be a lower operational
risk than switching to other fuel sources. However, other biogenic materials are
unknown and therefore require further study and development.
Barriers to The main barrier will be cost (mainly fuel cost) and the inherent risk associated
implementation with trialling a new fuel as the first mover. Minor infrastructure and training would
be required at some sites across the UK, so this investment is unlikely to be a
barrier.
The impact of impurities within the fuel on furnace infrastructure are unknown.
For example, much research will be required to see the effect of the biofuel on
the regenerator refractories as depending on source of biofuel we may be putting
new compounds down the regenerator much like Vanadium from heavy fuel oil.
Driving factors Development of low-grade 2nd generation bio-oils could provide a lower cost bio-
for adoption oil in the short term (pre 2030) with only minor capital infrastructure costs for
uptake.
The ability to become carbon negative if applied with CCUS (BECCS) is highly
attractive.
A number of new innovative technologies exist within the use of waste heat from
furnaces to drive new symbiotic processes, especially the pyrolysis of wastes into
biofuels.
November 2019 24
As all large-scale glass furnaces in the UK use natural gas and most have diesel
back-up systems, there is significant potential for rapid replication across all
sectors of the glass industry
Commercial Biodiesel (~6.5p/kWh) is currently more than twice as expensive as natural gas
(2.3p/kWh), so is unlikely to compete without carbon costs considered.
Lower grade bio-oils (e.g. from wastes) could be lower cost than biodiesel.
Lower cost bio-oils might compete with natural gas + carbon over a 30 year
lifetime for a furnace constructed in 2024 or 2025 if available at low cost <5p/kWh
and efficiency is same as diesel.
Cost of delivery will impact economics, unless fuel can be made on-site from a
local source/waste stream.
There is a significant question around the competition within the future bio-oils
sector and the impact this may have on price.
Table 13 Overview of bio-based fuels suitability for the glass industry
5.2 Hydrogen
Low-carbon hydrogen offers a potentially lower fuel cost than electricity, however further
economic modelling is required to understand how the future costs might compare to biofuels.
One of the greatest barriers to the adoption of hydrogen melting are the many technical
unknowns associated with the process and lack of process data even at lab-scale. Although it
is anticipated that it should be possible to convert the furnace design used with current natural
gas furnaces to a pure hydrogen fuel with relatively minimal disruption (at least with minimal
changes to the furnace footprint and geometry), a plant will still need to make significant
investment in infrastructure, from new H&S measures and training to new furnace designs with
new advanced refractories. It is likely that an oxyfuel generation plant or supply pipeline would
also be required (to avoid high NOx), which would further add to the CAPEX and OPEX due to
the electricity required. However, despite this, the lower cost of hydrogen compared to natural
gas + carbon may well off-set this over the life of the furnace.
A further advantage is that hydrogen could be delivered through existing natural gas pipelines,
maintaining the existing methods of fuel delivery (albeit potentially with some pipework
replacement) and reducing cost of delivery to site (significant advantages over biofuels and
electricity). However, the need for ATEX approved zones and stainless-steel pipework will be
challenging and costly to implement on existing glass plants and operators will need to be
educated on the H&S risks associated with using hydrogen fuels.
Whilst a transition to 100% hydrogen is likely to be very technically challenging, our findings
highlighted that, in the short term, it is likely that hydrogen will be blended into the natural gas
grid at reduced levels (e.g. starting at 10%, increasing to 20-30% hydrogen by volume). As
such, furnace modelling indicated that a ‘hydrogen + natural gas’ fuel-mix (up to 50% hydrogen
by volume) would operate under conditions much closer to that of an existing furnace (such that
it might even be possible to use such a mix with minor modifications to existing infrastructure),
which could provide a short-term measure to reduce carbon emissions without significant
disruption to sites or furnace changes.
November 2019 25
Assuming a scenario whereby an oxyfuel configuration is employed, the need for further CCUS
on-site will be minimal, especially if carbon-emissions from raw materials can be eliminated (e.g.
through increased use of recycled glass combined with use of non-carbonate raw materials).
by a glass manufacturer looking to switch to hydrogen melting.
Hydrogen Melting

source measures
Impact Current conventional H&S systems To use pure hydrogen, industry will
on H&S in place should be sufficient to require a comprehensive change in site
handle hydrogen however H&S systems and infrastructure for the
significant training of operators and entire industry. If this is part of a
site-engineers will be required to nationwide push, this will reduce the
ensure safe practice is adhered to impact as change adoption strategies can
and so that clear practices are in be shared across industries. Our
place in the event of problems. feasibility study has shown a very small
New H&S equipment will be proportion of the industry is used to
required for use of hydrogen e.g. ATEX/DSEAR regs which will put extra
ATEX rated and enclosed fuel skids; workload in a rapid adoption scenario.
this is expected to require significant
changes to site infrastructure
depending on the fuel delivery
methods employed.
Hotter flames and higher volumes of

gas throughput may also require
additional H&S measures. Pipework
may also have to be upgraded to
stainless steel and more frequent
checks employed, given the greater
liability of hydrogen to leak.
Impact on Hydrogen burns at a higher Existing NOx handling technologies

regulatory temperature; if N2 is present then should be able to cope with any
measures/ greater volumes of NOx will be additional NOx produced, although may
Environment present [GWI, Glass Trend, 2018]; need to be upgraded/expanded.
even if an oxyfuel process is used The GF-P2 feasibility study identified that
some NOx will be generated (as N2 work to explore how concentrations of
still leaks into the furnace). volatiles vary with combustion conditions
Enhanced evaporation of volatile in a lab-environment is essential before
species as a result of higher furnace any large-scale trials can be carried out.
operating temperatures may add to There are many unknown factors
emissions handling costs. surrounding the emissions expected
when pure hydrogen combustion is used
None of the above factors lead to to melt glass. As such small-scale
any regulatory problems or industrial demonstration is necessary to
challenges that the industry is not inform the industry of the effects of
already addressing. As such the hydrogen combustion in large scale
risk of hydrogen combustion leading industrial settings.
to issues in meeting regulatory
compliance is very low.
Fuel Delivery Any hydrogen delivered via new or old gas grid infrastructure would work well for
Logistics the industry given its use of natural gas. On-site production of hydrogen through
November 2019 26
electrolysers is not seen as viable for the Glass industry as all-electric would be a
more efficient fuel delivery method to the glass melt. For tanker based deliveries
(~8 required/day for a 10 MW furnace) Large scale hydrogen storage on site is
seen as a risk for long term deployment, and the energy density of hydrogen
means it is less suitable for tanker-based deliveries, so it would be preferable to
be delivered through gas grid network infrastructure.
Production Switch over to hydrogen would require a major change in capital infrastructure
Disruption above an 80/20% natural gas/ hydrogen blend. The effect of pure hydrogen
combustion has not been carried out in relation to glass and is highlighted by our
feasibility study as a major area of research.
Pure hydrogen combustion chemistry needs to be studied across a number of

glass types due to the unknown effects of a very high moisture content in the
combustion gases. Work with Dartington glass identified issues when melting
lead oxide glasses due to the formation of lead crystals, some other effects may
exist that are as yet unknown for more mainstream soda lime silica/boron
glasses.
If combustion is steady and able to fulfil environmental obligations then

theoretically furnace throughput could be maintained, the longer-term impacts on
capital equipment, especially refractories needs to be studied further.
Relation to Implementation of full-scale pure hydrogen firing is at a low TRL with many
state of the art hurdles to overcome across the board before it sees major uptake in industry.
Rapid widespread adoption is not expected, as deployment of hydrogen
technology is likely to be a phased approach by region, meaning some glass
manufacturers may not have access to hydrogen until significantly later than
other sites.
Barriers to Lots of work needs to be carried out to assess the full range of implications for
implementation large scale adoption of hydrogen in the glass sector and reassure a risk-averse
industry. Large changes to site safety infrastructure could impose a large CAPEX
cost to change fuel sources, and if hydrogen is not cost competitive with other
fuel sources the industry will move to the most economical option.
Driving factors Hydrogen is seen as a good option to utilise grid assets to remove carbon
for adoption emissions from the melting process, if there is not major CAPEX investment
required then uptake could be over a medium timescale and part of wider
(industrial cluster) decarbonisation plans for appropriate regions.
Commercial hydrogen could become lower cost than natural gas, if carbon costs included) by
2042-45. In this scenario the process could be lower cost. However, CAPEX
costs could be higher due to requirement of new furnace designs and H&S
infrastructure. Due to the large number of unknowns and lack of suitable furnace
designs it is not possible to quantify this gap currently.
Table 14 Overview of hydrogen fuel suitability for the glass industry
November 2019 27
5.3 100% Electric melting

Electric furnaces typically operate with a ‘cold-top’ whereby raw materials (or batch) are fed into
the furnace from the top (where the temperature can be as low as 50-100˚C), melting as they
pass down through the furnace. This ‘cold top’ forms an insulating layer which reduces thermal
losses from the glass melt. Electric furnaces are more efficient than air/fuel furnaces because
they do not produce large volumes of hot waste gases. The ‘cold-top’ traps many of the
pollutants produced by the melting process (e.g. SOx, generated by refining agents) and has no
associated NOx emissions as long as nitrates are not used in the batch. Although the furnace
has a shorter working life (6-8 years) compared to gas-powered furnaces (12-15 years), it is
significantly cheaper to build and more efficient. It could be argued that, if economically
comparable, a shorter life is an advantage, enabling manufacturers to exploit new furnace
technology developments.
Larger-scale electric furnaces are expected to need a ‘semi-cold-top’, with efficiencies lower
than a ‘cold top’ but still more efficient than combustion equivalents (natural gas, biofuels,
hydrogen), and with minimal emissions. It has been determined that large all-electric furnaces
will probably need a small amount of top heat in addition to in glass heating in order to melt
some coloured glasses and especially to achieve comparable glass quality due to the
mechanics and chemistry of glass melting.
by a glass manufacturer looking to switch to a 100% electric furnace.
Electric Melting
Factor Impact of alternative energy source Scalability potential & mitigation

measures
Impact on H&S Electric melting should be safer than Current use of some electrical boosting
natural gas as it has a cold top, with no in industry gives all-electric melting a
hot waste gases and uses no low risk in application across the UK.
combustible fuels. All sites currently have good H&S
systems in place to properly deploy all-
Even if a semi-hot-top design is used, electric melting across the UK
reduced volumes of hot gases will
mean that H&S risks are lower than for
conventional natural gas-powered
furnaces.
Impact on Electric furnaces are well established There is near zero risk of unseen
regulatory within the glass sector internationally at regulatory issues arising upon
measures/ small scale. upscaling this technology.
Environment
Cold or semi-cold top has near zero
emissions of NOx or SOx and so is
capable of meeting all regulatory
requirements. A semi-hot top furnace
may lead to additional emissions (e.g.
SOx not be captured in cold-top) if
natural gas is used to provide the heat,
however much lower than a gas-
powered equivalent and this heat could
be provided by electrical heaters.
November 2019 28
There are currently limitations on

amount of recycled glass that can be
used in electric furnaces (reducing
environmental benefits).
Fuel Delivery When installed, fuel delivery is simple and effective. However, the GF-P2 study
Logistics has shown installation of significantly greater HV electricity distribution is required
and take up across the UK glass sector likely requires major energy infrastructure
upgrades at a large capex cost.
Production Some glass types and colours are technically more difficult to melt with all-electric
Disruption furnaces, however electric melting is currently used on all common glass colours,
only requiring some minor changes in some cases (e.g. Amber glass is typically
handled in semi-hot top melting configuration).
There are a number of concerns around stability of supply and the impact of
potential blackouts if furnaces run all-electric; could be mitigated through use of
on-site battery packs and/or generators.
Relation to Large scale all-electric melting is currently seen as technically unfeasible due to
state of the art the melting mechanics of large (>600) ton per day furnaces. Smaller scale is
possible (up to 300 tpd) so there is some work to do in the field of all-electric
melting for large furnaces (which account for around 6/32 UK furnaces).
More fundamental research followed by demonstration is required to ensure
advancement in this area.
Barriers to The fuel cost, as well as the cost to upgrade supply to site (grid and on-site), are
implementation likely to prevent the industry switching to all-electric melting in the near future.
The requirement for significant infrastructure changes and running costs in the
region of 3x mean this fuel source is not seriously considered in the current UK
market, even where it is clearly technically feasible.
Driving factors The decarbonisation potential, reduction of waste gases and widespread use
for adoption globally as a technically recognised alternative fuel option mean all-electric
melting will always be considered for a large majority of the market but the
decision is driven by fuel economics.
The smaller footprint offered by the lack of abatement systems is seen as a
distinct advantage for plants where space is at a premium. Moreover, the all-
electric horizontal melter utilizes an almost identical footprint to existing natural
gas furnaces.
Commercial The main concern raised across the glass sector was fuel costs and potential
costs associated with upgrading site infrastructure.
Table 15 Overview of All electric fuel suitability for the glass industry
5.4 Flexible-Hybrid fuel scenarios

Beyond a small amount of electric boost to complement natural gas combustion, there is little
experience of hybrid scenarios.
The main benefits of the flexible-hybrid scenarios are to (a) provide ability to increase glass pull-
rate quickly, (b) enable significant demand side response on electricity use, (c) reduce risk of
fuel supply disruption.
November 2019 29
Although it is likely that a hybrid-fuel furnace will have higher CAPEX costs, it is likely that the
above benefits will off-set such costs, particularly if significant costs are required to upgrade fuel
delivery infrastructure (e.g. the local electric grid).
A hybrid furnace scenario would also offer the capability to respond to fluctuations in the price of
electricity. For example a furnace with the potential to use excess electrical energy at night is of
particular interest (as the industry already participates in demand side response schemes in
some areas), although there is uncertainty as to whether this scenario would materialise e.g.
electric vehicles may ‘mop up’ any spare electricity generated overnight. It should be noted that
there are existing predictive control systems currently on the market that can optimise fuel
usage based on price on a real time.
There are also likely to be optimum fuel-mixes whereby the fuels complement one another to
give improved melting behaviour, efficiency, or emissions, or adapt to regulatory changes. A
hybrid furnace should also increase operational flexibility to make it easier to control glass
colour through redox atmosphere in the furnace as/when required.
by a glass manufacturer looking to switch to a flexible hybrid furnace design.
Hybrid Furnace technologies

source measures
Impact on H&S Operational costs, H&S and Potentially easier to scale than other
and environmental regulations will vary fuel-scenarios, as allows for sites to use
Environmental significantly depending upon fuel mix existing infrastructure as much as
issues used (also glass subsector and possible and reduce volumes of any one
region) but the main considerations new fuel.
have already been covered above.
Fuel Delivery The way in which fuel is delivered to site depends on the specific hybrid solution,
Logistics & with different ratios of hybrids also having different implications for production
Production processes but, in general, hybrid systems should reduce fuel/energy supply risks
Disruption due to the flexibility to switch when required.
Relation to Current precedent exists for electrical boosting of natural gas furnaces, additional
state of the art use of more electricity is TRL8 at the moment. New novel hybrids highlighted in
our feasibility study show good potential but are low TRL (4)
Barriers to Economic drivers around existing hybrid models are a key barrier (e.g. increased
implementation CAPEX, fuels need to be close in price to encourage variability). Novel hybrid
concepts will need to show economic parity with current models to be considered.
Driving factors The ability to change fuel sources based on fuel cost, carbon pricing and
for adoption availability is attractive to the glass sector. Significant ability to offer large scale
load balancing for the national grid frequency is of particular interest.
Commercial Depending on the type of hybrid furnace there are a number of scenarios where a
commercial advantage could be gained through the use of intelligent demand
side load balancing for the grid. There is also a number of scenarios available
where fuels could be switched seasonally if this results in a more economically
sustainable process
Table 16 Overview of hybrid fuel suitability for the glass industry
November 2019 30
6 Economic assessment
To provide a provisional indication of how each fuel scenario might compare to natural gas, a
high-level cost-modelling exercise was undertaken by Element Energy, supported by Glass
Technology Services and furnace designers TECOGLAS and FIC.
The model compared the lifetime furnace costs of different fuel technologies – natural gas,
hydrogen, electric, biodiesel and flexible hybrid scenarios – including CAPEX, OPEX, fuel cost,
carbon cost, rebuild and repair cycles and high/low controls. It should be noted that, due to
limited time and budget available within the Phase 2 study, there are several additional costs
which are not currently included in the model and so all figures should be treated as indicative
estimates. It should also be noted that the model only considered biodiesel and not any other
forms of biofuels.
6.1 Methodology
The model uses input parameters such as the capacity of the furnace, the rebuild schedule, and
the type of glass produced (currently limited to container or float). These are then combined with
detailed information on operational parameters of different glass furnaces (furnace efficiency,
repair cycles, etc.) and informed assumptions around costs of components and fuels to output
the CAPEX and OPEX cost of each fuel switching option in annually 19. As each glass making
site operates differently, producing different products and with different operational parameters,
the parameters used in modelling can be tuned. Further controls are available to adjust some of
the inputs and assumptions, and these will be expanded within future studies to allow increased
customisation on a site by site basis.
6.2 Results
The costs of the different fuel switching options were calculated. Illustrative costs over a 30 year
time period are shown below for a 300 t/day container furnace built in 2035 operating on 100%
hydrogen, 100% electric and 100% biofuel, as well as the cost difference to a 100% natural gas
furnace.
Figure 17 Illustrative costs of fuel switching options for a glass furnace deployed in 2035 compared to natural gas
(price of carbon omitted).
19 Energy cost predictions based on BEIS data, 2019
November 2019 31
6.3 Carbon Pricing

The base case cost comparison is without the carbon price applied due to the uncertainty over
future carbon pricing and free allowances granted to the industry (due to the strong international
competition). As such, none of the fuels are cost competitive with natural gas and are very
unlikely to become so in the future. However, if incentives such as carbon pricing were applied
this could make it economically viable to convert to these fuels, and the model has functionality
to include carbon price trajectories in calculations and outputs. Whether these incentives will be
provided through fuel subsidies, allowing the sale of free carbon emissions permits or by
removing free carbon emissions allowances is unknown, however any solution must ensure the
competitiveness of UK industry is maintained and the risk of carbon leakage is mitigated. Figure
18 below illustrates the effect of carbon pricing being applied but omits and reference to CCUS
costs as they are currently uncertain.
Figure 18 Illustrative costs of fuel switching options for a glass furnace deployed in 2035 compared to natural gas
(including price of carbon).
6.4 Lifetime costs

Figure 19 below shows the discounted total lifetime costs of the different options broken down
into CAPEX, fuel costs, carbon cost and other OPEX, and figure 20 shows the undiscounted
annualized cost.
Figure 19 Illustrative lifetime costs (300 t/day container furnace deployed in 2035, discounted at 5%) 20
20When discounted the carbon cost has a lower impact, due to a strong increase with time over the project’s course.
Annualised costs are undiscounted.
November 2019 32
Figure 20 Annualised cost (undiscounted) for 300 tpd container furnace deployed in 2035
Due to the increased fuel costs for alternative fuels, none of the fuel switching scenarios
investigated is cost competitive with the natural gas case (when the cost of carbon or similar
incentives is not included). When carbon pricing is included, the biofuel and hydrogen scenarios
become cost competitive with natural gas for furnaces constructed in approx. 2035. However,
this is sensitive to a number of factors, including discount rate, fuel and carbon price, and
furnace efficiency for the different options, and could range from 2030 to 2045. It should also be
noted that lower-grade bio-oils may well prove more cost-effective and so should be included in
future studies.
The dominance of fuel (and carbon) costs in the lifetime costs means residual uncertainty in the
costs of fuel (and carbon) has a significant impact on the economically preferred fuel switching
option. Due to the high level of uncertainty in fuel costs, any of the options could be the most
economically feasible option, and Table 21 shows lifetime costs in 2025, 2035, and 2045
together with the range between the low and high sensitivities. As well as the technical
unknowns, this economic uncertainty is an important reason why all fuel options need to be
explored in future work.
Estimated Lifetime Costs (£ millions) 21. (Low – High Sensitivities)
Fuel Option 2025 2035 2045
Natural Gas 90 (68 – 103) 91 (68 – 103) 91 (68 – 103)
Natural Gas + CO2 cost 143 (81 – 263) 174 (88 – 355) 201 (93 – 444)
Biodiesel 165 (138 – 291) 163 (136 – 338) 162 (136 – 384)
Hydrogen 169 (155 – 228) 168 (154 – 236) 167 (153 – 245)
Electric 211 (196 – 246) 210 (196 – 248) 210 (196 – 253)
Table 21: Estimated Lifetime Costs including Low and High Sensitivities for Fuel Options
21 30 year lifetime cost for 300 T/day container (end fired) furnace, 5% discount rate.
November 2019 33
6.5 Proposed Further Development to economic model

It should be noted that the current model has omitted several cost categories. As such it is
recommended that the economic fuel switching model should be further developed and
enhanced to provide a comprehensive tool for the glass industry (including use by individual
glass manufacturing sites) to assess the likely fuel switching solutions. This should include:
• Feedback from technical demonstrations – further technical demonstrations and

technical modelling are required to enhance the understanding of the different fuel
switching options. This will enhance the modelling of the different CAPEX and OPEX
components and fuel costs on an operational furnace.
• Inclusion of lower-grade bio-fuels – Such as oils from pyrolysis of carbon-based

wastes, which may offer a fuel that has a closer life-time cost to that of natural gas.
• Impact of fuel switching on site infrastructure – Site infrastructure audits should be

undertaken to allow enhanced analysis of existing infrastructure on glass sites. This will
help understand requirements for each of the different fuel switching options and will
evaluate cost and necessity of subcomponent repair/replacement at stages in furnace
lifetime, site wide infrastructure changes (e.g. ATEX compliance), and the potential for
and benefits of ancillary infrastructure (e.g. fuel storage, batteries) which could be
installed on site. When incorporated into the model, this will enhance the comparative
assessment process.
• Customisation – once developed, it is anticipated that the model will be used to inform
glass sites about their fuel switching options and scenarios. This will require significant
additional functionality to customize model runs and outputs, accounting for subsector
specific (glass type/colour) requirements, possible impacts on and requirements to
achieve glass quality, and site specific requirements (e.g. size of electricity connection,
space constraints, pipework components). As well as these, it is recommended to
include additional parameters with impact on the lifetime cost assessments (e.g. cullet %
- high % difficult with all electric melting) and additional fuel switching scenarios (a range
of hybrid scenarios and custom scenarios) within the model.
• Completion of Model – to achieve usability for glass sites, the model must be
informative, usable, and updatable. Usability and usefulness of the model will be iterated
upon accounting for feedback from glass sites (the future users). To ensure the model
can be used for as long as possible, it needs to be future proofed with appropriate
customisation and functionality to allow updating as projections change (fuel costs) or as
further technical requirements of the fuel scenarios come to light.
November 2019 34
7 Timeframes and implementation

7.1 Biofuels
Adhering to the same standard as conventional diesel-fuel, biodiesel fuels should be compatible
with existing glass furnace infrastructure with minimum additional CAPEX investment required
to adopt this fuel technology. As operators have experience of diesel firing, there will lower risk
and associated costs in up-skilling the work force compared to 100% electric or hydrogen (both
of which would be new technologies to much of industry and its workforce) and this process
would not take long (in-house training would probably be sufficient).
The only significant technical barrier to implementation is a lack of data to demonstrate that
100% biodiesel will not adversely impact furnace infrastructure, nor glass melting. Therefore,
with a suitable R&D programme of work to investigate and trial biofuels, it is estimated that
within 3-5 years the industry will have a proven low-carbon technology that can be implemented
at all sites and will have had time to up-grade facilities, accordingly, should the market
conditions make the use biodiesel economically viable. Given that the UK has potential to
create biodiesel capacity to supply the entire glass sector, this could provide a route to
decarbonise all glass furnaces within the UK by 2030, although detailed studies into the
sustainability aspects of these fuels would be needed. Although very unlikely to ever be cost
competitive with natural gas without a different approach to carbon pricing, if carbon pricing is
considered then they are estimated to be cost competitive over a 30 year lifetime for a furnace
constructed in the early 2030s.
Lower-grade bio-oils, or carbon-based fuels from waste-streams (e.g. cooking oils, pyrolysis of
plastics or tyres, or other bio-wastes) have been identified that may offer a more cost-effective
source of low-carbon fuels. It is likely that specific glass plants will need to work with the supply
chain to develop reliable, consistent supplies of biofuels from specific sources of wastes, most
likely to be determined by geography. The Phase 2 study also identified that there might be
opportunities to utilise waste heat from the glass furnace to drive the pyrolysis processes that
convert waste-streams into bio-oils, thus further reducing the net CO2 emissions of these fuels.
There are many unknowns surrounding the technical and economic viability of such bio-oils,
such as levels of contaminants, variations in calorific and moisture content. Many of the
methods for processing such wastes are also still unproven at scale. Therefore, although plants
may begin to bleed in such bio-oils into existing furnaces in the short term (e.g. firing through
only 1-2 of the multiple burner ports on the furnace) to give a partial decarbonisation, it is
expected to take 7-10 years (probably with support from further grant-funded projects) to
establish a stable supply chain such that these fuels provide a full decarbonisation solution.
However, if a route could be identified to produce suitable bio-oils such fuels such that the price
is comparable (or lower) to that of natural gas, then favourable economics may drive this
technology forwards more rapidly.
One key risk identified in this Phase 2 study was a reliance of road-transport on fuel and the
seasonal variation/availability of some biofuels; such issues will need to be investigated in future
to provide a clearer picture of such risks and mitigation strategies.
7.2 Hydrogen
Numerous gaps have been identified in the technical understanding of how hydrogen fuels can
be integrated into a glass furnace. One of the greatest risks surrounds uncertainty over
required health and safety measures specific to the glass sector. Significant further work is
November 2019 35
required to identify key technical challenges that need to be addressed if hydrogen is to be

widely adopted as a fuel within the glass sector. A portfolio of evidence will also need to be
developed, along with engineering designs, furnace simulations and a study outlining costs to
integrate an hydrogen supply from the mains into the furnace, with the objective to provide glass
manufacturers with a detailed (technical and economic) business case to make a decision as to
whether to proceed with a trial on their furnace. Training programmes for operators will also
need to be considered, and these are likely to be extensive, requiring significant investment and
a number of years to properly establish across the glass sector.
Discussions with BOC indicated that current hydrogen supply could only meet 5% furnace fuel
requirements for a plant based in NW England (due to production and transport limitations).
However, it is estimated that within 5-10 years there could be sufficient supply to meet the
needs of a full furnace (based upon HyNet and H21 timescales). As such it is likely that the grid
will begin bleeding hydrogen into the existing natural gas supply before then and so urgent R&D
work is required to provide some groundwork as to the impact of this on the glass manufacturing
process so that the industry is well positioned and informed as to actions that may need to be
taken to reduce the risks associated with this.
Despite all of the challenges, this study has identified several research groups equipped with
suitable equipment and expertise to tackle the challenges and unknowns surrounding use of
hydrogen to melt glass. As such, with a suitably intense R&D test programme, the glass
industry could be in a position to technically switch to 100% hydrogen melting within 5-10 years.
Although very unlikely to ever be cost competitive with natural gas without carbon pricing, if
carbon pricing is considered then they are estimated to be cost competitive over a 30 year
lifetime for a furnace constructed in the early 2030s, similar to the biofuels fuel scenario.
7.3 Large-scale electric melting

The 2015 glass industry decarbonisation roadmap highlights this route as the preferred option
for decarbonising the industry. Furnace designers are confident that they already have
technically viable 100% electric furnace designs capable of melting upwards of 600 t/day, and
they just need a manufacturer willing to take the risk of making such an investment. As such,
large-scale 100% electric furnaces could be realised within 2-3 years (although it may require a
number of years of operation before other sites would be willing to follow).
However this study has identified that this will be challenging to implement primarily due to the
high cost of electricity, as well as the significant requirement to upgrade the electricity grid and
supply of electricity to sites, combined with the need for new designs of large-scale furnaces
and trials at scale before industry will adopt.
There are also questions as to whether suitable volumes of economically attractive ‘green’
electricity will be available by 2050. Currently a 100% electric furnace would only offer the
ability to off-set some carbon considering the grid is in the region of 30% decarbonised 22, this
would lead to a net increase in total CO2 emissions. As such this would be an effective method
to move carbon emissions but not to reduce them when looking at the overall decarbonisation of
glass melting in the short term. As such, it is unlikely that 100% electric furnaces will be a
reality before 2040, primarily due to economic factors and availability of supply.
There is a need to address some of the unknowns surrounding the practicalities and costs for
upgrading UK glass sites to all electric, partly so that this can be benchmarked against the other
fuel options, but also to cover the eventuality that there may be local supplies of green electricity
22 Energy and emissions projections 2018, BEIS, 2019
November 2019 36
that make this scenario viable (e.g. if located near a windfarm). Further modelling of large-scale
electric-furnace designs is also required such that there are designs ‘on-the-table’ should a
manufacturer wish to choose this option.
As an example of scale, typical furnaces in the 200tpd range have in the region of 1 MWh of
electrical boosting installed and would consume a total site demand of 3-4MWh for all other site
processes. If this furnace were to be 100% electric it would likely need an additional 6MWh,
which represent a huge sit increase in electrical distribution infrastructure.
Electric melting also has limitations in terms of the ability to use 100% recycled glass (if this
were ever an option), as such other raw materials will need to be used. Although electric
melting does not produce any emissions from the fuel at the point of use, some emissions are
likely to be produced from the decomposition of carbonate-based raw materials, such that small-
scale CCUS might be required.
7.4 Flexible Hybrid fuel scenarios

Due to the challenges, costs and risks associated with each individual fuel scenario, it is most
likely that glass plants will adopt a flexible hybrid furnace approach, which will vary across
sectors and from region to region. Such scenarios will enable manufacturers to take advantage
of fluctuations in fuel costs (e.g. could overheat furnace at night with cheap green electricity
then switch to combustion fuels during the day, at a reduced rate due to the residual heat stored
in the glass tank).
Such scenarios would also bridge the transition from natural gas to new fuels, a likely necessity
to allow the industry to build the required knowledge and operator experience of these new fuels
and associated technologies in a more manageable manner than a 100% switch in one go.
Although hybrid furnaces may have a higher CAPEX, the reduced risk to fuel supply disruptions,
coupled with the cost-savings potential from controlled fuel-switching will potentially more than
off-set these costs.
It is recommended that the remit of future studies should initially focus on the following
combinations of fuels, each with significant scope for combining with electric boost (whereby up
to 80% furnace could be powered with electricity): (1) Biofuel + natural gas, (2) hydrogen +
natural gas, (3) Biofuels + hydrogen.
As well as aiding the transition to low-carbon fuels, this creates a route to begin off-setting CO2
emissions in the short-term whilst the respective supply chain is established. Hybrid furnaces
combining either natural gas or biofuels with hydrogen also addresses issues with the low
emissivity and high temperatures associated with a hydrogen flame, offering a route to achieve
a ‘yellower’ flame, burning at a lower temperature, thus aiding heat transfer and reducing NOx.
Given the uncertainty in the future availability and prices of fuels and the fact that most glass
furnaces are designed to operate for between 15-20 years, it is likely that furnaces with hybrid-
capability will be built within the next 3-5 years, or at least furnaces that can be easily up-graded
to a hybrid design. Even though the CAPEX might be higher, such furnaces will offer a route to
respond to future energy scenarios and avoid the need for an early rebuild.
7.5 Compatibility with CCUS

In parallel to the above developments, developers of CCUS technologies (e.g. Apache, C-
Capture) need to be engaged to identify the most suitable technologies for capturing CO2
produced by the glass furnace in order to direct future research towards finding a CCUS
November 2019 37
solution for the glass sector. It is expected that it may take up to 20 years for such solutions to
be realised. It is recommended that future R&D studies undertake detailed characterisation of
the emissions. This will reduce this time-frame and potentially identify routes to optimise fuels
and/or combustion parameters and/or glass composition in order to make it easier and quicker
to employ CCUS technologies in future, thus reducing the development timeframe, which could
enable the glass sector to become carbon negative by 2050. It will also provide inputs into
future furnace designs so that even if the CCUS technology is not available at the time of the
build, the potential for retro-fitting CCUS technologies can be included in the design.
Given some glass container plants are located near carbonated drinks filling lines some of this
CO2 could be utilised in down-stream processes; even if not, 80% glass plants are based along
M62 corridor and close to planned future CO2 pipelines for carbon storage projects.
7.6 Summary of possible implementation timeframe with sufficient R&D

investment
In summary this study indicates that, with suitable R&D investment, biodiesel could enable the
glass industry to eliminate up to 90% of the CO2 emissions associated with heating glass
furnaces by 2030. If combined with CCUS, this could offer a route to net-carbon negative
emissions.
However in the longer term, it is likely that the industry would want to move towards a hydrogen-
electric hybrid (once such fuel sources are available), possibly with a small amount of bio-oils
(e.g. either to optimise the hydrogen flame or as a back-up when electricity is in high demand),
such that the fuels can be delivered to site without the need of a road-based haulage network.
This also reduces the reliance on the biofuel network, freeing it up for other sectors.
Figure 22 below provides an outline of the implementation timescales that might be achievable if
the industry fully backs R&D activities proposed in this report.
November 2019 38
Figure 22: Estimated implementation timelines for the various fuel switching Options
8 Conclusions and Recommendations

Due to uncertainties and differences between subsectors and predicted future variations in
availability and affordability of different fuels across the UK (e.g. hydrogen supply may be
localised; local grid capacity for electricity supply limited), no single low-carbon fuel scenario is
likely to be suitable for all 17 of the largest glass manufacturing sites which account for the large
majority the UK’s glass manufacturing output (and associated CO2 emissions from glass
melting).
Of the four potential fuel-scenarios investigated (biofuels, hydrogen, large-scale electricity,
flexible-hybrid) this study has found that each has potential to be technically feasible, with the
potential to fully decarbonise the glass furnace heating process whilst meeting regulatory
requirements, if the fuel could be supplied at an economic price.
8.1 Biofuels
Biofuels (i.e. fuels derived directly or from wastes from 100% renewable bio-sources, so not
including blends with standard diesel) could be a good option for fuel switching, given the
similarities of some biofuels to gas oil, the industry's preferred fuel before natural gas, and that
there is further potential to use lower cost bio-oils. The UK has biofuel capacity to upgrade to
supply the entire glass sector and this solution could be strengthened by potential later
application of CCUS to mitigate process emissions and provide negative emissions for the
sector. However, it was identified that there is no understanding of how biofuels will perform in a
glass furnace compared to natural gas and standard diesel, in particular in terms of their effect
on glass melting behaviour and on emissions as there are no recorded cases globally of firing a
glass furnace with bio-oils. As such, further R&D is required into these areas.
8.2 Hydrogen
There is little understanding of how hydrogen will perform in a glass furnace. Key concerns
include the heat transfer mechanism, the volumes of airflow through the furnace, H&S
implications such as ATEX rated equipment, effects on glass melting and furnace refractories,
whether furnace geometry is suitable, effects on emissions, e.g. higher NOx due to hotter flame.
A significant R&D programme is required to build sufficient understanding of these and other
technical challenges.
Significant effort also needs to be invested into training programmes and into building a proper
understanding of the requirements for a site to implement hydrogen fuels.
The study suggested that it would currently only be possible to commercially source suitable
volumes of hydrogen to provide 3-5% fuel for a typical glass furnace (from conversations with
BOC), so it would be challenging to undertake a meaningful large-scale trial now, but suitable
volumes may be available in future. Larger supplies of hydrogen would be required to enable
the glass sector to undertake meaningful large-scale trials.
8.3 Large-scale 100 % Electric melting

The study identified that furnace designers are reasonably confident they can design larger-
scale (>300 t/day) electric furnaces, despite technical unknowns, such as how efficient a semi-
hot top furnace might be.
November 2019 39
Due to lack of interest from the industry little, if any, modelling of such designs has been
undertaken and so this should be the focus of short-term R&D efforts.
The greatest barrier to implementation surrounds the economics of electric melting (i.e. the
higher cost of electricity compared to natural gas). There are also uncertainties around the
scope/cost of upgrading the supply to each site in the UK to facilitate full-electric melting and the
size of CAPEX investment required in new furnace designs and potential changes to plant
layout. Significant engagement with government would be required to provide suitable
incentives and investment into national supply infrastructure if 100% electric melting were to
become viable across the UK.
8.4 Flexible Hybrid-fuel scenarios

Beyond the widely employed natural gas-electric furnaces, little work has been done into hybrid
scenarios, nor into dynamic fuel-switching systems nor the impact that such a system might
have e.g. emissions or CCUS.
The following three hybrid scenarios have been identified as having the greatest potential: (1)
biofuels + natural gas + electric, (2) hydrogen + natural gas + electric, (3) biofuels + hydrogen +
electric.
Further R&D studies and furnace modelling is recommended to identify the most suitable hybrid
furnace designs, which should then be worked up into pilot furnaces for larger-scale trials.
The longer-term impact of a UK industry that has specialist knowledge in advanced furnace
control could be highly advantageous to both the economics of UK-based glass manufacture as
well as UK based specialist knowledge that can be exported globally.
8.5 Bio-methane
Although bio-methane was not covered within this study (due to it being out of the competition
scope), it does offer a potential route to decarbonise the glass manufacturing process and so
should be considered in future studies.
8.6 CO2 reduction potential
If successful, our study has concluded that these low-carbon fuel technologies have the
potential to remove up to 1.2 million tonnes of CO2 emissions per year by 2030, totalling
more than 20 million tonnes by 2050. Without continued funding in this area the industry is
unlikely to explore these fuel scenarios until after 2030 and these new technologies are
unlikely to be implemented widely until after 2040.
This study also identified the need to develop a research infrastructure and expertise within the
UK that can support and drive rapid implementation of these low-carbon fuel technologies. This
would have a knock-on benefit to the UK economy of creating new, high-skilled, jobs and
leveraging significant international R&D investment.
It has been identified that the glass industry needs to review the 2014 British Glass
decarbonisation roadmap, to update plans in accordance with findings from this project to
ensure that the industry is not only aware but signed up to implementing the most promising
decarbonised fuel technologies.
November 2019 40

Alternative Fuel Switching Technologies For The Glass Sector - Phase - 2 - 2019 - Final - Report

Uploaded by

Copyright:

Available Formats

Alternative Fuel Switching Technologies For The Glass Sector - Phase - 2 - 2019 - Final - Report

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Alternative Fuel Switching Technologies For The Glass Sector - Phase - 2 - 2019 - Final - Report

Uploaded by

Copyright:

Available Formats

BEIS Industrial Fuel Switching Phase 2

Alternative Fuel Switching Technologies

Date: November 2019

3 Current state of art 9

4 Overview of most attractive Low carbon Fuel Switching technologies 14

5 Key Challenges and Opportunities 22

7 Timeframes and implementation 35

8 Conclusions and Recommendations 39

1.2 Why more than one solution is required

Therefore, to effectively decarbonise the entire sector as fast as possible, it is recommended

1.3 Technical opportunities

1.4 Timescales and findings

• A unique multi-fuelled ‘Hot’ glass pilot facility in St Helen’s, Merseyside

2.3 BEIS Industrial Fuel Switching competition remit

2.4 Scope of Study

Figure 1 Inter-relationship of work packages within Phase 2 study

2.5.1 Literature reviews

2.5.3 Direct Industry Engagement

2.5.4 Energy networks and suppliers

• University of Sheffield, PACT • British Glass

3 Current state of art

1 Prodcom sales figures, 2015, Office of National Statistics

Figure 2 Manufacturing process of float glass 4

3.2 Energy consumption and emissions

4 Figure credit: University of Leeds

Gas oil Fuel oil Natural gas Electricity Total

[Accessed 10 June 2019]. Available from: https://www.britglass.org.uk

2019. [Accessed 10 July 2019]. Available from: https://www.gov.uk/government/statistics/energy-consumption-in-the-

3.3 Diesel and fuel oils

3.4 Oxyfuel furnaces

• Melting of speciality or high temperature glasses (Borosilicate for instance)

3.5 Praxair, Optimelt

4 Overview of most attractive Low carbon Fuel Switching

8 Image courtesy Inperpetuum Partners LLP

4.1.2 Lower grade bio-based fuels

4.1.3 Availability of biofuels

Figure 6.1 sources of Biofuels used in the UK 9

Figure 6.2 sources of Biofuels used in the UK by country 10

Figure 6.3 Types of Biofuel used in the UK 11

12 DfT: Renewable Fuel Statistics 2019 First Provisional Report

Figure 7 potential decline in future biodiesel use

4.1.4 Potential for reducing CO2 emissions

4.2.1 Applicability of hydrogen fuel for glass melting

4.2.2 Availability of Hydrogen fuels

4.3 100% Large-scale electric melting

4.4 Flexible hybrid fuelled furnace

5 Key Challenges and Opportunities

18 [Industry data, British Glass, Glass BREF Section 4.8]

Factor Impact of alternative energy Scalability potential & mitigation

build on this knowledge.

Table 13 Overview of bio-based fuels suitability for the glass industry

Factor Impact of alternative energy Scalability potential & mitigation

Hotter flames and higher volumes of

Impact on Hydrogen burns at a higher Existing NOx handling technologies

Pure hydrogen combustion chemistry needs to be studied across a number of

If combustion is steady and able to fulfil environmental obligations then

Table 14 Overview of hydrogen fuel suitability for the glass industry

5.3 100% Electric melting

Factor Impact of alternative energy source Scalability potential & mitigation