Chapter 8: Manufacturing and pathways to deployment
If liquid air as an energy vector benefits from a pre-existing ‘fuel’ distribution network (chapter 6), it may gain further advantage from the characteristics of the equipment that would run on it. Liquid air devices can generally be made substantially from existing components drawn from mature supply chains with few bottlenecks to hamper expansion. And unlike many other low carbon technologies - such as EVs - liquid air technologies require no rare earth or other precious metals. In this chapter we assess the manufacturability and sustainability of liquid air technologies against a number of criteria, and where relevant compare them to competing technologies.
1. Grid storage
Liquid Air Energy Storage (LAES) systems such as that developed by Highview Power Storage rely on equipment widely used in the energy and industrial gases sectors that has been developed over the last century. The liquefier, for example, is one section of a standard Air Separation Unit (ASU) – a process developed by the German chemist Dr Carl von Linde in 1902 - while the turbines and generators, compressors, pumps and electric motors required are commonly used in the power and process industries. In this section we survey the global supply chain for the key components of a Liquid Air Energy Storage system, which can be broadly divided into three main elements: the charging device, storage components and discharge device.
Global supply chain for LAES components
Because the LAES charging device is effectively one section of an ASU, a ‘first to market’ system could be drawn substantially from units commonly used in the industrial gases industry today. For example, as shown in Table 8.1, both the main air compressor and recycle air compressor fall well within what is currently available in standard designs, and can readily be sourced from several manufacturers globally. Other areas, such as plant controls, offer no significant requirement over and above that of a typical liquefaction or industrial process plant. Unsurprisingly the same is true of the expansion turbines, APU and coldbox, including the main coldbox heat exchangers and vessels.
Compressor and turbine efficiencies improve with scale, peaking at a capacity of several hundred tonnes per day, a plant size that is regularly built today. This fits well with a grid scale LAES with a power output 50-100MW and an energy storage capacity 200-400MWh, which is considered the optimum size for the technology.
The storage components of a LAES system include a low pressure cryogenic liquid storage tank and a separate high grade cold store, which stores the cold captured in the discharge phase for later use during the charging process.
The cryogenic liquid air store would typically consist of either a single tank or a series of smaller tanks. An early system would probably use established twin walled vacuum insulated tanks commonly used in the industrial gases business for storing cryogenic fluids. These tanks consist of a stainless steel low pressure tank within a carbon steel tank, in which the space between the tanks holds a vacuum and perlite insulation. For much larger capacities of up to 200,000m3 it would be possible to use bespoke tanks with concrete walls and stainless steel lining similar to those used in the LNG industry.
The high grade cold store is also a low pressure device, which not only reduces any safety risk but also the cost and difficulty of manufacture at scale. The store uses granite shingle, a cheap and widely available material, as the thermal store medium, which means that costs at scale are kept within acceptable limits.
The discharge device consists of a power turbine and generator. Power turbines of the design and type required in the LAES system are currently available from leading turbo machinery manufacturers up to 45MW. These machines use a standard gearbox design which helps reduce costs. Larger total outputs can be achieved by means of multiple machines, adding increased flexibility in turndown for a given efficiency drop, as well as increased resilience in case of breakdown and unplanned maintenance. Larger individual machine outputs are possible with the use of a direct coupled design. Unsurprisingly, the size and availability of suitable generator technology is closely related to the power turbine range which drives it, and therefore matching components are available.
Pumping cryogen to high pressure while maintaining required flow rates represents the greatest challenge to existing technologies in the supply chain. In order to maximise the energy retrieved from the cryogen it is necessary to pump the liquid to a high pressure, but there is a trade-off between the higher energy return and the pumping work needed to achieve it. Cryogenic pumps that are currently available operate up to 120 bar at 20kg/s, which falls short of 200 bar optimum. However, the optimum is considered achievable with further development in the supply chain. Larger flow rates can simply be achieved by using multiple pump configurations.
Heat exchangers for evaporation, reheat and superheat are all components that present little challenge to the current supply chain, with both pressure and temperature extremes within operating ranges of existing heat exchanger technology.
Balance of plant represents a significant proportion of materials required. Components include valves, circulation pumps, instrumentation, ancillary circuits and systems etc, all of which can easily be sourced from the local industrial utilities supply chain.
Table 8.1: Supply chain evaluation of key components of liquid air energy storage systems
Sustainability of materials
The main construction materials in air liquefiers are aluminium and steel of various types, including high alloy steels for compressors and turbines and low-temperature sections of the process. There is clearly no shortage of steel or bauxite, but the high energy cost of the conversion to aluminium has to be considered in any sustainability analysis.
Copper is more of a concern, however. Copper-based metals have been widely used in heat exchangers because of their heat transfer and corrosion resistant qualities, and a global shortage is pushing prices higher. This has driven the search for alternative materials and it is now common for ASUs to use aluminium and stainless steel in heat exchangers, columns and pipework. Copper is also significant in electrical machines such as motors and generators because of its excellent conductivity.
For safety and operational reasons some parts of valves, compressors and turbines may need to be made from high-alloy steels containing chromium, nickel, zinc and molybdenum. However, the quantities used are very small compared to the steel and aluminium. Unlike renewable energy technologies such as permanent magnet wind turbines, batteries and PV cells, LAES systems require no rare earth metals or other exotic materials, which are costly and finite.
Steel, aluminium and concrete are the main materials of construction and are all easily recyclable; when an ASU is demolished it is quite normal for 95% of the materials to be reused. The used adsorbents and oils would need to be disposed of by a specialist contractor and large foundations below ground level can be covered with soil and landscaped or rebuilt on. If necessary the foundations can be broken up and removed altogether.
UK manufacturing, the global supply chain and pathways to deployment
As liquid air plant sizes increase, the selection of turbo-machinery should become easier rather than harder; the equipment at the Highview Power Storage pilot plant is smaller than the standard range of many manufacturers. This does not imply that larger equipment is available ‘off-the-shelf’; indeed it may take many months to manufacture. However, the basic designs will already exist and components will be selectable from a standard range.
The major components for a LAES facility such as the turbines and generator sets, compressors and motors, cryogenic pumps and plate fine heat exchangers in the cold box would have to be manufactured overseas. The inlet air cleaning skid, storage vessels, cold store vessels, high voltage electrics, cold box manufacture and all interconnecting pipework could be sourced from the UK. For larger LAES plants, the number of manufacturers capable of fabricating large cold boxes, heat exchangers and pressure vessels becomes more restricted, but if greater capacity is required, the UK still has shipbuilding or oil rig construction plant capable of handling the largest components.
A round-table discussion of industry experts held at the Institution of Mechanical Engineers in March 2013 concluded that if design, civil engineering and construction work is added to domestically produced components, around 50-60% of the value of a LAES installation could originate in the UK.a This is not to say the UK would necessarily capture so much of the value, however; after 30 years of globalisation, it is now commonplace to outsource manufacturing to countries such as Brazil, Russia, India and China, where costs can be half those in the UK. However, other factors such as transport, efficiency, reliability and communications may affect the balance.
Although a large amount of manufacturing capability has been relocated overseas, the UK does maintain a significant capability to design, integrate and package individual process units and machines into complete operational facilities. It is envisaged that from start to finish one of these plants could be brought on stream in 20-22 months. If the intention is to build multiple units of the same design, the round-table concluded total installation costs could be reduced by some 15% (see chapter 3, Appendix 1 and round table reports). This would also allow operators to hold low volumes of centrally held stocks of major spares. LAES plants should also be long-lived since they generate no corrosion or combustion products; all the major industrial gas companies operate large scale air separation plants that are more than 40 years old.
The equipment used in a LAES comes from a mature manufacturing and operational background and so would be expected to operate with very high levels of reliability; plant on-line times in the air separation industry are in the region of 99.5%. If a LAES plant was installed at an existing industrial gas production site the manpower and expertise required to run it could be provided from within the current workforce. For a LAES plant installed at a standalone site, the unit would be stopped, started and monitored from a remote location. A technician would only be required to visit the site once a week or to deal with a site emergency.
The visiting technician would be expected to undertake routine maintenance tasks; major maintenance on this type of equipment would only be required every 5 years at which point a shutdown of five to seven days would be required. The main drivers for maintenance frequencies tend to be safety valve testing and the cleaning of cooling systems.
In overseas markets, the main potential export would probably be engineering design and project management, which can be high value. However, the approach would differ between markets. Some countries where technical capabilities are low may require ‘turnkey’ plants; others may need only technology licencing and engineering consultancy.
In summary, the supply chain for liquid air technology is mature, global and extensive and the UK has the industrial capacity to deliver more than half the value of a LAES plant. There is no reason why the international supply chain should not deliver a target of 500MW of LAES capacity in the UK by 2020 (Summary Report and Recommendations), or supply the UK market potential of 14GW by 2050 (chapter 3). To achieve the earlier target, orders would need to start to be placed this year, but the current international supply chain is capable of delivering these levels of capacity without creating a bottleneck.
The economic value to UK manufacturing
At this early stage it is clearly not possible to quantify with any certainty the potential value of liquid air to the entire UK economy. However, it is possible to make a high level estimate of the value of grid-based LAES technology to UK manufacturing, on the basis of the results presented elsewhere in this report and a number of simple assumptions.
Other recent advances in low carbon technologies - such as offshore wind, for example – have delivered disappointingly little economic benefit to the UK, because we lack the relevant manufacturing base. However, liquid air plays to the UK’s traditional strengths in mechanical engineering and cryogenics and therefore has the potential to achieve a relatively high proportion of UK content. If the technology proves cost effective the economic benefits to the UK could be significant.
This analysis considers the potential for liquid air to increase Gross Value Added (GVA) – one measure of output in an industry or sector - and to create jobs. We adopt three different approaches but with a single set of assumptions. We assume the market for grid storage achieves the potential identified in chapter 3, and that LAES captures 25% market share (Table 8.2). The analysis is calculated in today’s prices - ie excluding inflation - or ‘real’.
Table 8.2: Assumed storage market and liquid air capacity
It is not yet clear how highly the electricity market will value the additional flexibility brought by liquid air, particularly in light of the continuing uncertainty around Electricity Market Reform. Detailed analysis has been undertaken by Strbac and colleagues1, but this leaves us with a wide range of scenarios and variables. In any event, Strbac considers the issue from a whole system perspective and not from the point of view of technology developers looking to generate an economic return. At this stage, therefore, any assessment of economic benefits must necessarily be broad brush.
For the purposes of this indicative analysis, we have considered three approaches. These produce a wide range in terms of potential impact on GDP, although given the large number of variables it should be stressed that outcomes could fall outside this band. It should also be noted that this analysis derives from the Strbac/DECC ‘Grassroots’ pathway and therefore assumes a ‘high wind’ scenario.
Model 1: Annualised cost of storage
Strbac calculates a range of annual savings in 2020, 2030 and 2050, based on a spread of annualised costs of storage. Liquid air is assumed to have an annualised cost of £150/kw/year and a market share of 25%, except in 2020, when Strbac’s cost of storage range falls below the projected cost of liquid air and the market share is therefore assumed to be zero. On these assumptions, liquid air revenues are estimated to be £244 million in 2030 and £525m in 2050.
Model 2: Investment cost model
We assume that at the lower end of the expected cost range (£750/kw) there is sufficient value in the liquid air proposition for investors to generate a real return on capital of 9%. In addition, revenue will be generated through operations and maintenance of the facilities. We have assumed that this will equal around 10% of the capital cost. This is higher than the 1.5% - 3% quoted elsewhere in this report (chapter 6) in order to adjust for the relatively high expected labour intensity of the whole life cycle of the projects including the supply chain.
Given the length of time expected for the storage market to develop, significant project activity will not start to take place until after 2020. As a result, annual revenues are projected to rise from £60m in 2020 to £200m in 2030 and £570m by 2050.
Model 3: Share of Benefits Analysis
Strbac projects a range of gross and net annualised benefits according to the projected annualised cost of the storage solutions. The net benefits are projected at around £1bn in 2030 and £8bn in 2050 for a cost of £150/kw/year. It is assumed that half of the net benefits generated are attributable to the storage solution and the remainder shared between other stakeholders. There is zero projected benefit in 2020 because the annualised cost range used by Strbac is below the expected level for liquid air, but there are benefits of £125m in 2030 and £1 billion in 2050.
GDP and Jobs Assessment
The economic analysis resulting from these approaches uses data from the Annual Business Survey 2011 (Provisional Results), showing annual revenues and gross value added per sector. We have taken the figures from the sub-sector Production - Electricity, Steam, Gas and Air Conditioning as the nearest proxy. The 2011 GVA percentage for this sub-sector was 24%. After calculating GVA in line with this sub-sector, we have assumed that 50% of the costs (revenues less GVA) are derived from a UK source and that there is an Induced Spending Multiplier (additional spending in other sectors that arises from this activity) of 120%. This produces the total GDP Added figures shown in Table 8.3. We have also calculated the potential for job creation by using the median gross annual earnings from the relevant sectors of the Annual Survey of Hours and Earnings 2011. The number of new jobs created in these scenarios ranges from 10,000 to almost 20,000, also presented in Table 8.3.
It should be stressed that this analysis is high level and indicative, and the results are highly dependent on assumptions about the size of the storage market. Nor does it consider the potential economic impact of liquid air in the transport sector.
Table 8.3: GDP and employment impact of LAES to 2050
The Dearman Engine is a reciprocating (piston) engine that operates at near ambient temperatures, and as a result it is unlikely to offer many unfamiliar challenges to vehicle engine manufacturers. The most unfamiliar part of the system is likely to be the part exposed to cryogenic working fluid – liquid air or nitrogen. However, cryogenic technologies are mature and have been used in the industrial gas and liquefied natural gas (LNG) industries for decades.
There is a wide variety of materials suitable for use in cryogenic systems that are plentiful and relatively low cost, including stainless steel, aluminium alloys, PTFE and polyethylene.2 The availability and cost of these materials compares favourably to some of those required for other low carbon vehicle technologies, such as platinum in hydrogen fuel cells and lithium and neodymium in battery electric vehicles.
The Automotive Council has created a framework for assessing technologies on the basis of their technology (TRL) and manufacturing (MRL) readiness levels.3 The Dearman Engine relies on the integration of sub-systems that are already in widespread use in existing vehicles around the core technology, and so has a relatively advanced manufacturing readiness level. We assess the MRLs of the key sub-systems below.
Manufacturing readiness of key Dearman Engine sub-systems
Working fluid storage. Cryogenic liquids such as liquid nitrogen can be stored in a variety of commercially available vessel types. Where LNG or liquid nitrogen are used in or transported by road vehicles today, the cryogens are typically stored in vacuum insulated stainless steel tanks, which are available in a variety of sizes.4 As a result, on a scale of 1 to 10, the MRL of the liquid air energy store would be about 8.
LNG costs considerably more per kilogramme than liquid air or nitrogen and is flammable when it boils off, so the design of these tanks is probably more complicated than may be required for liquid air. However, existing designs are likely to be suitable for early deployment of liquid air vehicles. Longer term there is a development opportunity for the cryogenic tank to become cheaper and simpler. For example, lightweight tanks made of plastic or aluminium are already commercially available for use in static applications in capacities up to tens of litres.5
Working fluid delivery. Pumps capable of pressurising cryogenic fluids to very high pressures are a mature technology in a range of static and mobile applications. In transport they have been used in LNG lorries, and high efficiency ultra-high pressure submersible pumps are deployed commercially by companies such as Westport.6 The MRL of this part of the system is therefore likely to be 7 or 8.
Dearman Engine. This is the novel part of the system currently in development at Ricardo. The Dearman Engine is a reciprocating engine that operates between a few degrees below ambient and low grade waste heat temperatures of around 90C. Peak cycle pressures of about 200 to 300 bar are unlikely to be prohibitive in this temperature range; some diesel engines operate at this pressure and higher temperatures.7 First generations of the Dearman Engine are likely to be made from steel, aluminium and other alloys using known engine manufacturing techniques. In the longer term the ambient operating temperature range of the Dearman Engine as a prime mover could allow it to be made of lightweight materials such as plastics.
Heat exchange fluid system. The heat exchange fluid is a water/glycol mix much like that found in conventional vehicles today, and the storage and transfer of this type of fluid is mature low cost technology. The reheat arrangements are likely to involve standard vehicle radiator technology or, in the case of ICE-Dearman heat hybrids, standard liquid-to-liquid heat exchangers. These are commodity items supplied by companies such as Alfa Laval and SWEP, and so require no significant manufacturing process development to be integrated into a Dearman Engine.8
The main sub-systems of a Dearman Engine may present no significant challenge for Original Equipment Manufacturers (OEMs) or their Tier 1 Suppliers, but the question remains whether the industry would choose to develop the technology for mass markets. One way to assess this is to compare the characteristics of liquid air vehicles against the technology roadmap developed by the New Automotive Innovation and Growth Team (NAIGT) since 2008.9 The roadmap (Figure 8.1) represents the industry consensus around how the vehicle technology and manufacturing will evolve over the next 30 years, and has been used to develop a common research agenda (Table 8.2). Liquid air could address a number of the automotive industry’s key research challenges.
Figure 8.1: NAIGT Technology Roadmap Source: NAIGT
Table 8.4: NAIGT common research agenda summary. Source: NAIGT
Propulsion. The use of liquid air for waste heat recovery could help increase the thermal efficiency of the IC engine and therefore allow it to be downsized, a key aim of the research agenda for propulsion.
Energy storage. Liquid air can also contribute to meeting the industry’s energy storage objectives particularly on cost. Cryogenic tanks for energy storage can be produced for as little as £4,500 for a 200 litre ‘one-off’ in the UK, or £1,000 in China. A 200 litre tank will hold 160kg of liquid air and approximately ten kg are required to generate 1kWh of electricity. This means the price of energy storage using liquid air today is ~$450/kWh using a tank produced in the UK, which is already lower than the NAIGT’s medium term cost target; and $100/kWh using an imported tank, about half the long term energy storage target. There is also likely to be scope for further cost reduction if alternative materials or high volume manufacturing techniques are employed. Improvements in energy density could be achieved by elevating peak cycle temperatures and pressures, but the imperative to do so is reduced by the rapid refuelling times that liquid air can deliver.
Vehicle efficiency. Liquid air may be able to raise vehicle efficiency by reducing weight if plastic engines and tanks are developed. As a waste heat recovery device, it could also contribute to engine downsizing and significant improvements in aerodynamic efficiency if it allows designers to dispense with the need for a conventional radiator.
Pathways to deployment
Passenger cars. Vehicle manufacturers are slow adopters of new technologies because of the scale and risk of the necessary investment. It will cost an OEM such as Ford or GM more than £1 billion to develop a new passenger vehicle from scratch, and failure would be disastrous. Even incremental changes to existing designs take three to five years to introduce. The Euro 1-6 emissions standards that require modifications to existing diesel engine technology, have taken many years to implement. Euro 6, which was legislated in 2007, will come in to force for new vehicle sales in January 2015.10
A new passenger car typically takes seven years to develop: three years from engineering prototype to programme approval, then four years to volume production. Changes to an existing model typically take three years from initiation to volume production. The development timelines of commercial vehicles are similar, while off-road vehicles may be quicker.
It may take over ten years for a zero emission technology to be fully adopted as a prime mover on volume vehicles. However, liquid air may have some advantage because the main sub-systems are similar to those of IC engines. The NAIGT technology roadmap foresees the introduction of zero emission vehicles such as EVs from the 2020s, and it seems liquid air vehicles could be developed in broadly this timeframe.
Another advantage of liquid air to the OEMs is that it could extend the time before it becomes necessary to replace the ICE altogether, by raising its efficiency and reducing its emissions through ICE-Dearman hybrids or concepts such as the liquid nitrogen split cycle engine (chapters 2 and 4). Billions of pounds have been spent to date on the development of internal combustion engine vehicles and liquid air could help OEMs retain some of that value. This could mean liquid air is developed in hybrid, emissions-reducing applications sooner than as a prime mover, zero emissions concept.
Vehicle fleets. OEMs are not the only route to market for liquid air. Operators of depot-based fleets such as buses or delivery vehicles could be early adopters, since the risks to them of product failure are much lower than for an OEM, so they are prepared to take greater risks on new technologies if the business case is strong enough. Production volume requirements for the fleet sector are also more compatible with early stage technologies. Fleet retrofit is also likely to be the fastest adopter of liquid air technology since installing liquid air production or storage capacity at depots can be achieved more quickly than creating an entire filling station network.
There is scope for the waste-heat-to-power and refrigeration applications of liquid air to be retrofitted, which could lead to rapid take up of the technology in the fleet sector. This could be achieved on the basis of the existing cryogenic supply chain (chapters 6 and 7) without the need for significant OEM or Tier 1 involvement. This could accelerate take-up in the mass markets through demonstration of the technology.
The cryogenic equipment supply chain is currently geared to produce a few to tens of thousands of units per year. This is because of the low level of current demand rather than any materials or manufacturing capability constraint. Companies such as Productiv, which is developing a ‘proving factory’ specifically to bring early stage vehicle technologies to the level of manufacturing 10,000-20,000 units per year, offer a route for liquid air technologies to be produced in low commercial volumes. This may then allow the technologies to be demonstrated in niche markets and perhaps justify the OEM investment needed to develop the production lines necessary to deliver mass market volumes.
From the discussion presented in this chapter we conclude:
- Liquid air technologies are based on mature components which can be readily sourced from mature supply chains.
- The key components of a liquid air energy storage (LAES) plant can be readily sourced from existing manufacturing capacity in the UK and abroad and 50-60% of the plant’s total value could in principle be sourced from the UK.
- Existing supply chains could, in principle, deliver 500MW of LAES capacity in the UK by 2020 if orders were placed soon, and could also deliver estimated UK storage market potential of 14GW by 2050.
- The Dearman Engine would be made mainly from components similar to those of a conventional internal combustion engine (ICE) and cryogenic components similar to those already used in LNG vehicles.
- The NAIGT technology roadmap calls for the introduction of zero emissions vehicles in the 2020s and liquid air vehicles could be developed in this timeframe.
- Liquid air technologies could extend the time before it becomes necessary to replace the ICE altogether by raising its efficiency and reducing its emissions.
Strategic Assessment of the Role and Value of Energy Storage Systems in the UK Low Carbon Energy Future, report for the Carbon Trust, Strbac et al., June 2012, http://www.carbontrust.com/media/129310/energy-storage-systems-role-value-strategic-assessment.pdf
Automotive Technology and Manufacturing Readiness Levels, Low Carbon Vehicle Partnership and Automotive Council, January 2011, http://www.automotivecouncil.co.uk/wp-content/ uploads/2011/02/Automotive-Technology-and-Manufacturing- Readiness-Levels.pdf
HLNG Vehicle Tanks Specification Sheet, Chart LNG, 2012 http://www.chart-ferox.com/getattachment/cbe66c42-2bd1-4ceb-8fc6-a42662ac2cad/14771738.aspx
An Independent Report on the Future of the Automotive Industry in the UK, New Automotive Innovation and Growth Team (NAIGT), February 2008, p45,
Reduction of pollutant emissions from light vehicles, http://europa.eu/legislation_summaries/environment/air_pollution/l28186_en.htm