Beyond the ICE: Energy Storage for Consumer Vehicles

This was originally written in 2012.?I thought it would be interesting to see how the predictions played out in the ten years since, considered practicalities such as cost a little more (can you tell it was written by a student?), and made some worthwhile style alterations while I was at it.

---

Internal Combustion Engine (ICE) cars have been a pretty big deal.?They've been around since 1886, and today we buy (80 million per year)[26].?That could change, due to the (scribbles some notes - 10 year lifespan, 10,000 miles per year at 25 miles per gallon, 250 g per mile) 2 gigatonnes CO2 / year the sector emits.?The UK government has stated an intention to ban the sale of new ICE vehicles from 2035, and is far from acting alone.

Unlike the grid as a whole, methods employed for transport must bear in mind the need to minimise mass in a moving vehicle, so creating something with an energy density akin to the engine (ICE) vehicles is a challenge.?The second challenge is the adequate utilisation of energy from renewable sources in the national grid – much has been made of the limited contribution unreliable sources (e.g. wind and solar energy) can make to the grid, as non-renewable sources must be kept online to fill any gaps in supply.

Advances in Lithium-based cells

### Update

Lithium-based cells have pretty much won the contest to be the replacements for Internal Combustion Engine (ICE) cars.

?* Tesla have managed to (somewhat) mass-produce battery packs, which means not just the problem of primary battery chemistry, but actual engineering.?A model 3 LR (now discontinued), using a LiNiCoAlO2 cathode and silicon/graphite anode has 167 Wh/kg [27].?That works out as 480 kg and 80 kWh, which EPA testing states gives a 500 km range [25].?So, a fairly aerodynamic car is expending 160 Wh/km. A renault zoe manages similar, although that appears to be the higher end of efficiency for EVs.

?* The cost of these batteries is also coming down.?General Motors claims $100/kWh, for reference, which means $5000 - $10000 of the sticker price is the battery. Not great, not terrible.

?* A formula 1 car carries 100 kg of fuel, which works out to be ~1300 kWh. Given thermal efficiency, let's say 350 kWh makes it into useful work. For the benefit of someone wondering where formula E will be in 5 years: to match overall drivetrain weight with comparable range in an electric racer, you'd need over 1 kWh/kg. I'm putting money on Nope, not happening.

?* The carbon intensity of electrical grids has fallen, in the UK I'd probably estimate 250-300 gCO2/kWh. Good news, although France is still an order of magnitude better. *For some reason*.

?* Fast chargers may quickly become the limiting factor. Going back to our recharging dilemma - if a typical fuel pump - a fairly simple device - shovels enough energy in to "recharge" a car in 2 minutes, and a "fast charger" - containing power control circuitry, big transformers, and fat copper wires -?takes 30, and needs using twice as often, we're going to need ~30 times as many chargers and spend a lot more on them. This may be ameliorated by home charging (which can be slow, so the charger is much cheaper).

?* Solid-state lithium batteries haven't beeen seriously covered here, but are being invested in and are likely to increase that critical energy density figure to near 500 Wh/kg.

Rechargeable batteries for electric vehicles (EVs) have the advantage of being both the primary energy storage medium and providing an easy implementation of energy recovery (aka regenerative braking, aka "KERS") using a motor-generator.?The footprint of EVs is going to depend on where the energy actually comes from: this is called well-to-wheel (WTW) performance, and for the UK grid producing about 445g CO2/kWh representative of the current UK grid performance[2], they're already outperforming ICEs[1], with more potential to improve. A typical modern Li-ion cell intercalation of lithium in graphite for the anode, and uses a cathode made of lithium ions intercalated in cobalt oxide (CoO2) with an organic electrolyte; an anode made of pure lithium with a very high specific energy is possible, but dendrite formation on charging presents safety issues due to short circuits, and solid electrolyte interphase layer formation causes poor cycling performance, losing up to 3% of capacity each charge cycle[4],[5].?

Batteries do have several major flaws however:

?* Very low specific energy compared to petrol, however, partly due to the need to carry the oxidant and electrolyte on board.?On average (in 2012), they averaged 150 Wh/kg [3].?Petrol has a specific energy of ~13,200 Wh/kg, diesel ~12,700 Wh/kg [1].?So, EV batteries have 1% the specific energy of fuel tanks - though that's partly mitigated by having a *much* simpler, lighter drivetrain.

?* A long recharge time compared to the fuel tank of an ICE car; let's say I fill my 40 L petrol tank in about one minute.?Given a density of 750 g/L for petrol, that's 30 kg/min - or 24 *megawatts* (MW) going into my car.?A good EV charger, something like Tesla's supercharger, delivers 250 kW (i.e. about 1% as fast)

?* The cell degrades over large numbers of cycles (cycling performance).?A few hundred to a thousand cycles is typical.?Again, this is partially mitigated by the rest of an EV drivetrain being simpler.

### Better Anodes?

The maximum theoretical specific energy of the battery is determined by the anode and cathode, and voltage difference between them – in this case, the potential of the cathode with respect to lithium metal.?The Li-ion design shown above has an theoretical anodic capacity of 372 mAh/g [6],[7]; with a typical 4V cobalt oxide cathode, the specific energy is 387Wh/kg [4].???Much of the development has been focussed on improvements or replacement of the carbon anode; recently,?vapour-grown carbon fibre cells within 10% of the theoretical limit with excellent cycling performance have been grown and used commercially in an 8 Ah electric bicycle battery[8].?Other materials can break this barrier; e.g. Si has a much a higher capacity of 4.2 Ah/g [6], but a large volume change upon lithium insertion (300%[5,6]) presents longevity issues;?Zhou et. al present a means of reducing the damage, whilst improving electrical conductivity; by inserting silicon nanoparticles into graphite oxide sheets, then reducing the graphite oxide to graphene and freeze-drying, an encapsulation of silicon nanoparticles in graphene with a capacity of 1.8Ah/g falling to 1.1 Ah/g after 100 cycles [6]; this is still relatively poor cycling performance, but if capacity levels off after 100 cycles it is still an attractive alternative to graphite anodes.?A similar situation exists with Ge; Chan et. al report using Ge nanowires on a metal substrate to achieve similar 100-cycle performance (~1Ah/g) to Zhou, with the benefit of a much higher diffusivity, and therefore a faster recharge time[7].

Titanate (Li4Ti5O12) is suggested as an anode material; there is zero volume strain upon lithium insertion,giving excellent cycling life [9],[10].?Titanate anodes have the rare property among lithium anodes being researched of having a lower specific energy than graphite (175 mAh/g [9],[10]), but have another advantage; the spinel structure gives it a high diffusivity, leading to a fast recharge[9].?Consumers may therefore tolerate a lower range if the recharge time is short.

### Better cathodes, and the Li-air cell

A potentially high-capacity approach is the Li-air battery.?In this method, 'the intercalation cathode of the Li-ion cell is replaced by a porous, conduction substrate... flooded with electrolyte.?Upon discharge, Li-ions form by oxidation of the lithium metal anode and pass across the electrolyte [to the cathode, where atmospheric oxygen is reduced] [4].?The theoretical energy density of this technology is much higher than that of a Li-ion cell at 3.5 kWh/kg, and 'practical' specific energies of 500 Wh/kg are possible.?The Li-air battery comes in two forms: aqueous (sometimes referred to as Li-water), and non-aqueous (the convention refers to the nature of the electrolyte; the non-aqueous cell uses an organic electrolyte).?Non-aqueous batteries have the problem of a lithium metal anode, with the problems discussed earlier with electrolyte decomposition on discharge, and with charging causing both; an over-capacity of lithium is recommended to counter this, but at the expense of specific energy and greater cost due to the rarity of lithium[1].?Aqueous batteries utilise an ion-permeable membrane to stop the lithium reacting with the water, but this creates the risk of corrosion between the solid and aqueous electrolyte, leading to many of the same problems as the non-aqueous cell; Stevens et. al propose a third electrode as a remedy[4].

Other alternative cathodes for the Li-ion cell include iron phosphate (FePO4), manganate (Mn2O4), and the sulphur cell.?Of these, manganate and iron phosphate prove more environmentally friendly, and use more readily available materials than cobalt oxide[11], with similar capacities and voltages (98-155 mAh/g for CoO2 against 111-168 mAh/g for FePO4 and 40-138mAh/g for manganate, all in the range of 2.5-2.0 V SHE)?with FePO4 outperforming CoO2 after cycling [5].??Manganate also has the desirable characteristic of a flat charge/discharge profile[11], although at the cost of poor cycling life. Hybrids such as CoMnO4 cathodes may be profitable, with experimental capacities of 84 mAh/g developed[12] – the variation in capacities is due to the high dependence of capacity on nanostructure, so developing processes to create the correct nanostructure cheaply and efficiently will be the key challenge regarding cathodes in the coming years.

Alternatives to the Li-ion cell

---

### Update

As mentioned, I think we can pretty much say Lithium has won. Hydrogen's difficulties are just too much for practical handling, and flywheels don't have the cost effectiveness. I'm not aware in any major advances in MOFs.

There may still be exceptions outside of cars - principally because, in the North, wind is the best source of renewable power, and if you want enough turbines to cover a cold still winter, you're going to have easily 100 GW of unpredictable excess, and ideally you don't want to throw that away.

### Flywheels

Flywheels have already had a limited success in automotive applications, though so far only at high expense – being used in the williams formula 1 car[13] and the porsche 997 GT3 R endurance car [14], and in some public transit schemes.?A flywheel's energy is given by:

?E=Iω^2

Where I is the rotational inertia and ω is angular velocity.?Pérez-Aparicio and Ripoll [15] show that the performance index for minimising specific energy is (strength/density), since a reduced moment of inertia from a less dense material can be compensated for by dimensional changes and increased angular velocity.?Although the specific energy of the flywheel is not competitive with existing lithium battery technology[16], and is unlikely to improve (CFRP, with an specific energy of 160 Wh/kg [15], is the highest realistically achievable), and it has a high self-discharge rate (15 mins[16] c.f. weeks for a Li-ion battery) making it impractical for overnight energy storage, it has a high specific power and excellent cycling performance[1], making them well suited to ESS applications (where the power from regenerative braking usually exceeds the capability of a battery?).?Doucette and McCulloch[16] give a comprehensive analysis of current flywheel, battery, and ultra-capacitor technology; using a 540 kJ flywheel weighing 15kg supplied by flybrid systems – mass-market solutions with a CVT (continuously variable transmission) needed to supply drive to the wheels will weigh approximately 48kg for the same energy storage.?

### Hydrogen storage

Hydrogen is an initially promising candidate as a primary energy storage medium for EV's, having a high inherent specific energy when reacted with oxygen in a fuel cell (39 kWh/kg?).?However its high specific volume when stored under standard conditions means it would require an impractically large container.?It can therefore be stored under elevated pressure, liquefied, or chemically adsorbed – or sometimes a combination.

Pressurised type IV tanks can hold up to 700 bar at ambient temperatures, but are expensive; and the performance, like that of flywheels, is unlikely to improve in the near future (though cost may decrease).?Liquefied tanks can achieve a greater volumetric efficiency, but at the cost of a high self-discharge rate.?Storing in the form of a hydride is possible at room temperature, and, as will be seen later, the most active front of materials research for hydrogen storage.

Paster et. al [17] conducted an analysis of various hydrogen storage methods; solely pressurised storage at 350 bar and 700 bar, 'Cold gas' (500 bar and approx. 193 K on filling), CcH2 (liquefied hydrogen at 275 bar and 30 K in a type III tank) and the adsorption material MOF 177 under 150 bar pressure and 100 K, and found that:

* CcH2 had the best energy density and specific energy (5.5wt%, 42 g/L), a reasonable dormancy (4 Wd), and the lowest capital cost ($2219), but poor WTW performance owing to the need to liquefy the hydrogen (174 gCO2eq/km);

* MOF 177 had worse energy density and specific energy (4.8wt%, 34 g/L) than CcH2 but a higher volumetric efficiency than any of the gas systems, similar WTW issues to CcH2 (174 gCO2eq/km), but better dormancy than CcH2 (16 Wd) and a slightly higher capital cost of $2984;

* Cold Gas has the lowest specific energy, intermediate energy density, dormancy and WTW performance with a high capital cost (3.8 wt%, 27 g/L, 12 Wd, 136 gCO2eq/km, $3431)

* 700 bar pressurisation achieved the same specific energy as MOF 177 (4.8wt%), lower energy density (26 g/L), and the highest capital cost ($3506), but a lower WTW energy usage (129 gCO2eq/km), surpassed only by 350 bar, and infinite dormancy;

* 350 bar pressurisation had the lowest WTW energy usage and no dormancy issues, but the lowest energy density and poor specific energy (unsurprisingly given the performance of 700bar storage), and intermediate capital cost (123 gCO2eq/km, 17 g/L, 4.0wt%, $3096)

???(all figures for the entire system, per vehicle.?An ICE vehicle and ICE/battery hybrid can??achieve 204 gCO2eq/km and 124 gCO2eq/km WTW respectively)

?A similar study conducted by Jorgensen[18] gave LH2 as the highest specific mass, at 15%; the specific energy is then competitive with an ICE vehicle.?The WTW equivalent emissions of all hydrogen technologies are found to be disappointing, however, with little to no advantage over ICE vehicles.

?Of these, the technology with the greatest potential for improvement is hydride storage; similar to Li-ion anodes, greater relative specific energy and energy density lies in a greater surface area.?Chae et. al[19] describe the principles behind metal organic frameworks (MOFs), and MOF-177 as described earlier; an infinite sheet of graphene has a specific surface area of ~3000 m2/g, but the specific surface area of benzene rings is much higher, at nearly 8000 m2/g.?MOF-177 is made by linking units of small groups of carbon rings (1,3,5-ben- zenetribenzoate. Or BTB) , with Zn4O(CO2)6 clusters, forming a porous crystalline framework with a specific surface area of 4500 m2/g, leading to an uptake of 1.25wt% at 1 atm and 77 K (16.1wt% is achievable at 250 bar and 100 K): Isoreticular MOF's, such as IRMOF-11 with up to 1.62 wt% have been created[20], and NOTT-112 can achieve 2.3wt% at the same temperature and pressure.?NOTT-112 can also achieve 10 wt% at 77 bar and a hydrogen density of 50 g/L, compared to 48 g/L for MOF-177 (at 1 bar).?The next challenge facing such surfaces is to increase the temperature at which they are useful; this, as seen previously, is instrumental in reducing the WTW emissions of the system.

Energy storage solutions for the grid

The key disadvantage of renewable energy sources, such as wind and solar energy, is unpredictability of supply; when wind power consists of over 10% of the supply the simpler mechanisms for controlling the grid are overwhelmed [21].?Traditional thermal power stations, nuclear in particular, can supply power consistently, but are inefficient under small loads, and are slow to respond to changes in demand.

nuclear power stations require hours to reach full power, and whilst fossil fuel stations are faster this brings a cost in both efficiency and longevity of parts[22]. Solutions capable of storing energy when it is in surplus and releasing them in times of deficit are therefore of key importance.?Grid storage presents different challenges and opportunities than automotive KERS – the energy stored is not kinetic, specific energy matters less than cost per unit energy, it is possible to foresee demand to a much greater extent,?and there is the possibility of increasing the amount of load that can be removed at will.?At a national level, the dominant form of energy storage is hydroelectric (total hydroelectric storage in the U.K. is 26.7 GWh, with a peak power output of 2.86 GW and efficiency of 75%[1]) – but whilst this is by far the cheapest method for bulk energy storage and likely to remain so, niches do exist for other forms – integrating energy storage directly into renewable energy sources, for instance, simplifies the balance of the grid enormously, and a great deal of decentralised energy storage may be possible through V2G.?Whilst energy storage for the grid is not primarily a materials science challenge, applications for advanced materials in grid storage do exist.

Load control and Vehicle-to-Grid (V2G)

Systems to remove sources of energy demand from the grid are already in place in an industrial context – large plant manufacturers can have their supply temporarily cut off§ – but thus far no such technology has emerged for private consumers.?A small contribution could be made from current household goods – refrigerators, for example, are a product that can tolerate deviations from its preferred load for small amounts of time.?The major addition, however, is EV's – V2G systems could provide up to 50 GW and 1200 GWh[1] to the grid for intervals on the order of minutes, at low cost once the vehicles themselves have been purchased, although this will be highly time-dependent; very little capacity available during rush-hour, for example, and almost all available during the night.?This may prove advantageous; since the peak demand from EV's fits troughs in demand for most other loads (with the exception of TV pickup), V2G even without active regulation of demand would allow supply to work more consistently, and thus more efficiently.?Heating, if electrified through the use of heat pumps or (at the cost of efficiency) simpler resistive heaters, is another potential source of household load control, – the Scottish island of Fair Isle has a second grid for heating, which renewable energy sources can transfer excess energy into at will[1].?Although this exact solution may not be practical for larger grids such as that of the UK, solutions involving smart heaters and fridges may be possible.?If it is economically possible for electrolysis facilities to run with such an unpredictable supply, load control might even be a useful companion for hydrogen production via electrolysis – as is already done on some small islands[23],[24] Whether this is feasible on a larger scale remains to be seen.

Flow batteries

Flow batteries are a slightly left field idea, more akin to a fuel cell - or a traditional ICE.?The exchange is mostly similar to normal batteries, but with the electrolyte containing oxidised or reduced ions being continually replenished from a tank. This requires both oxidised and reduced form to be stable in solution (and therefore ionised) so transition metals with many stable ionisation states are desired.?The half-cell reaction for a VRB is:

V4/5+ +2e- → V2/3+ (E~1.26v)

?The ability to quickly replenish the electrolyte has obvious implications for EV's, but the heavy ions required result in too low a specific energy (40 Wh/kg[21]).?However, for grid storage this is not an issue; and vanadium has the advantage of being far less rare and valuable than lithium, which is needed not only for batteries but potentially for fusion reactors[1].?An additional advantage is the near-indefinite cycling performance[21], and the fact that energy and power are decoupled; power is proportional to flow rate, energy is proportional to tank capacity.?VRB's have already been successfully trialled in Sapporo, Japan (peak power 6 MW, 4 MW steady state, 6 MWh), King Island, Australia (0.2 MW, 0.8 MWh) and Castle Valley, Utah (0.25 MW, 2 MWh).?Whilst they are not cost-competitive with hydroelectric storage where it exists, close integration with wind farms can make achieving energy balance on the grid simpler, and on islands with little suitable land for hydroelectric storage and ample supplies of wind power, they can prove very useful.

Conclusion

Effective use of sustainable energy, both in developing EV's with sufficient range, power and recharge time to be attractive to consumers, and in fully utilising the sustainable energy available for the national grid, will involve several materials challenges, and the interaction and competition of different technologies – e.g. V2G technology is only applicable for battery EV's, not fuel cell EV's, and the presence of technologies such as V2G can make other storage systems such as VRB's largely obselete.?Developments in fuel cell EV's are likely to be hamstrung by the high cost and energy usage of the required infrastructure, leading advances in high-capacity anode li-ion battery EV's (and hence V2G storage) as the likely leading technologies going into the future – though, and potential advances in high-temperature hydride storage could even mean the return of the hydrogen economy.?In any case, globally, as we have already seen with the UK's ample hydroelectric reserves, Fair Isle's second grid, and Sapporo's VRB's, with a large range of needs and assets, almost all of these approaches are likely to find their niche somewhere.

Glossary

EV – Electric Vehicle

KERS – kinetic energy recovery system – used in vehicles to recover energy normally lost in braking.

CFRP – carbon fibre reinforced plastic – high specific strength material used for e.g. flywheels or pressure tanks.

ICE – internal combustion engine

CcH₂ – compressed cryogenic

Wd – watt-days of dormancy (how many days a full tank can store hydrogen before venting under standard conditions with a given level of heat leakage)

type IV – all CFRP, type III is CFRP with a metal liner.

WTW – well-to-wheel – the total energy required from primary production to use.

VRB – Vanadium Redox Battery

CHP – combined heat and power

MOF – metal organic framework

V2G – Vehicle to grid – battery EV technology where the vehicle is used as energy storage for the grid, absorbing energy when there is excess and vice versa.

References

1. Sustainable Energy without the hot air, David JC Mackay, UIT Cambridge 2009 (a free online version is available at https://www.withouthotair.com/)

2. 2011 Guidelines to DEFRA / DECC's GHG Conversion Factors for Company Reporting ,?Department of Energy and Climate Change (DECC) and the Department for Environment, Food and Rural Affairs (Defra) , 2011

3. A contribution to the progress of high energy batteries:?a metal-free, lithium-ion, silicon-sulfur battery

4. Lithium-air and lithium-sulfur batteries, Peter G. Bruce, Laurence J. Hardwick, K.M. Abraham, MRS Bulletin 2011 vol. 36, 506-512

5. Nanostrutured electrodes for lithium-ion and lithium-air batteries: the latest developments, challenges, and perspectives, Min-Kyu Song, Soojin Park, Faisal M. Alamgir, Jaephil Cho, Meilin Liu, Materials Science and Engineering 2011 vol. 72, 203-252

6. Facile synthesis of silicon nanoparticles inserted into graphene sheets as improved anode materials for lithium-ion batteries, Xiaosi Zhou, Ya-Xia Yin, Li-Jun Wan, Yu- Guo Guo, ChemComm 2012, 48, 2198-2200

7. High Capacity Li Ion Battery Anodes using Ge Nanowires, Candace K. Chan, Xiao Feng Zhang, Yi Cui, Nano Latters 2008 Vol.8 , No.1 307-309

8. Vapour-grown carbon fiber anode for cylindrical lithium ion rechargeable batteries, H. Abe, T. Murai, K. Zaghib, Journal of Power Sources 1999 vol. 77, 110-115

9. Electrochemical study of Li₄Ti₅O₁₂ as negative electrode for Li-ion polymer rechargeable batteries, K. Zaghib, M. Simoneau, M. Armand, M. Gauthiet, Journal of Power Sources 1999 vol. 81-82, 300-305

10. Single crystalline lithium titanate nanostructure with enhanced rate performance for lithium ion battery, Jun Lu, Caiyun Nan, Qing Peng, Yadong Li , Journal of Power Sources 2012 vol. 202, 246-252

11. Rechargeable Lithium Batteries (and Discussion), Peter G. Bruce, R. Cahn, N.E. Bagshaw, A. Hamnett, Philosopical Transactions: Mathematical, Physical and Engineering Sciences 1996 vol. 354, no. 1712, Materials for Electrochemical Power Systems 1577-1594

12. Synthesis of LiCoMnO4 via a sol–gel method and its application in high power LiCoMnO4/Li₄Ti₅O₁₂ lithium-ion batteries, Xingkhang Huang, Min Lin, Qingsong Tong, Xiuhua Li, Ying Ruan, Yong Yang, Journal of Power Sources 2012 vol. 15, 352-356

13. https://www.williamsf1.com/news/view/1341

14. https://www.porsche.com/uk/motorsportandevents/motorsport/racingcars/911gt3r-hybrid-997/technologyandconcept/

15. Exact, integrated and complete solutions for composite flywheels,?José Pérez-Aparicio, Lluis Ripoll, Composite Structures 2011 vol. 93, 1404-1415

16. A comparison of high-speed flywheels, batteries, and ultracapacitors on the bases of cost and fuel economy as the energy storage system in a fuel cell based hybrid electric vehicle, Reed T. Doucette, Malcolm D. McCulloch, Journal of Power Sources 2011 vol. 196, 1163-1170; see also https://www.flybridsystems.com/ for further information on the flywheel used

17. Hydrogen storage technology options for fuel cell vehicles: Well-to-wheel costs, energy efficiencies, and greenhouse gas emissions, M.D. Paster, R.K. Ahluwalia, G. Berry, A. Elgowainy, S. Lasher, K. McKenny, M. Gardiner, International Journal of Hydrogen Energy 2011 vol. 36, 14534-14551

18. Hydrogen storage tanks for electric vehicles: Recent progress and current status, Scott W. Jorgensen, Current Opinion in Solid State and Materials Science 2011 vol. 15, 39-43

19. A route to high surface area, porosity and inclusion of large molecules in crystals, Hee K. Chae, Diana Y. Siberio-Pérez, Jaheon Kim, Yongbok Go, Mohammed Eddaoudi, Adam J. Matzger, Micheal O'Keeffe, Omar M. Yaghi, Nature 2004 vol. 427, 523-527

20. Hydrogen Sorption in Functionalised Metal-Organic Frameworks, Jesse L.G. Rowsell, Andrew R. Milward, Kyo Sung Park, Omar M. Yaghi, JACS communications 2004, online.

21. The Vanadium advantage: Flow batteries put Wind Energy in the Bank, David C. Holzmann, Environmental Health Perspectives 2007 vol. 115, A358-A361

22. Electrical Energy Storage for the Grid: A Battery of Choices, Bruce Dunn, Haresh Kamath, Jean-Marie Tarascon, Science 2011 vol. 334, 928-935

23. An integrated approach to hydrogen economy in Sicilian islands, Fabio V. Matera, C. Sapienza, L. Andaloro, G. Dispensa, M. Ferraro, V. Antonucci, Internation Journal of Hydrogen Energy 2009 vol. 34 issue 16, 7009-7014

24. A Norwegian case study on the production of hydrogen from wind power , Christopher J. Greiner, Magnus Korp?s, Arne T. Holen, International Journal of Hydrogen Energy 2007 vol. 32, 1500-1507

25. https://www.teslarati.com/wp-content/uploads/2017/09/Tesla-Model-3-EPA-CSI-HTSLV00.0L13.pdf

26. https://www.statista.com/statistics/200002/international-car-sales-since-1990/

27. https://insideevs.com/news/338743/everything-you-ever-wanted-to-know-about-tesla-batteries/