Category: Battery storage

The battery that made headlines

dchone July 14, 2017

It is hard to imagine that something as mundane as a battery could make national headlines, but that is what happened last week in Australia. I was there for a family visit and as much as climate science gets debated in the media (see my previous post), it barely rates against the energy debate that has dominated the Australian headlines in recent months. The issue is the cost of electricity and the reliability of supply, both of which continue to be contentious subjects.

The cost of electricity has risen sharply in Australia over the last two years, but this has been coupled with supply problems that have led to blackouts in some parts of the country. This has been most visible in South Australia, which has closed all its coal fired power plants and invested heavily in renewable energy. The result is a wind / solar / natural gas power mix, backed up by an inter-connector with Victoria. For the most part this has been a workable solution, but when particular conditions coincide the result can mean blackouts or load shedding by some industrial customers. This usually involves periods of high load coinciding with low renewable energy availability at a time when some other part of the network is under stress, e.g. supply from the inter-connector. The problem is that this has happened enough times to become a major issue, leading to extensive media coverage, finger pointing and public outcry.

The reporting around this issue can be likened to the climate science debate I discussed in my last post – i.e. overly prone to hyperbole and lacking in basic facts. The solution that has been proposed for South Australia is the rapid construction of battery storage and into this foray stepped Elon Musk (CEO of Tesla). Some months ago, after power supply problems in South Australia hit the headlines yet again, he made the offer to build a major Lithium-Ion storage facility in 100 days, with no payment required if construction was delayed beyond that period. Not surprisingly this galvanized the government and led to a public tender for such a facility.

Last week the South Australia government announced that Tesla had won a tender to build a grid scale battery. The project will incorporate a 100MW peak output battery with 129 megawatt hours of storage alongside an existing windfarm, near Jamestown. Elon Musk flew into Adelaide for the announcement, which added to the fervour. What was interesting was the reaction in the media, ranging from support bordering on adulation to downright condemnation. Contrast the July 9^th Sun-Herald, where their social commentator Peter Fitzsimons noted;

Oh, how sweet it is. After all the haters, all the pile-ons, all the craven dinosaur politics which maintains that coal really does have a future, the SA government shimmies, shakes, steps left, steps right, bursts through into clear and announces its lithium battery deal.

. . . with political commentator Tim Blair in the Daily Telegraph on July 10^th, who starts a near full page article with the headlines;

Charge of the left brigade – People are falling over themselves to fawn over magic man Elon Musk’s battery absurdity but he is just the latest saviour to get the green light from eco worriers.

Neither are energy commentators of any note, but such is state of the energy transition debate in Australia that this doesn’t seem to matter. The controversy highlights the need for clear policy and thoughtful steps forward in implementing an energy transition, irrespective of the reasons driving such a transition, be they climate change, energy costs, local air quality or some combination of these and other needs.

In the case of South Australia, exuberance and some technology bias led to a rapid shift in the electricity system to a point where stability became a problem, which will now take some time to correct. The battery solution isn’t a full solution at all, but one of several measures that will be required to correct the imbalance that has been created. Alongside the Musk excitement, there was also the recent announcement that the government will build additional gas fired generation capacity at a cost of over A$300 million.

The South Australian electricity system operates at around 2-3 GW, with peaks and troughs depending on the time of day and year. Annual demand is some 14,400 GWh, or about 40,000 MWh per day. The initial battery capacity is 129 MWh. The gas capacity can add over 5,000 MWh per day if it operates continuously, or 13% of the demand. While the battery commanded all the headlines, it is clearly not a solution for extended periods of high demand or reduced supply, given that it can hold only 5 minutes of South Australian demand. However, it is an important step in the quest for a more balanced system that has very high levels (often >70%) of intermittent renewable energy. Further battery systems are likely, but a better balance between natural gas and renewable energy would appear to be the more achievable outcome in the short to medium term. It is also likely to be the more cost efficient outcome, given the ~A$100 million investment required (according to a local private source) for what will be the largest Lithium-Ion battery system in the world.

The cobalt rush?

dchone March 6, 2017

It was interesting to see the FT report recently that there may be a cobalt crunch looming, with hedge funds actually stockpiling physical cobalt. The FT reported that half a dozen funds have purchased and stored an estimated 6,000 tonnes of cobalt, worth as much as $280m. This is equivalent to 5 per cent of annual global production of the metal.

I talked about cobalt in a recent post and noted that the current Tesla S battery contains some 7 kgs of Cobalt; so the stockpile reported by the FT is sufficient for 860,000 cars. This is equivalent to ten times the Tesla 2016 requirement for the 80 thousand electric vehicles (EV) they manufactured. In 2016 total global sales of EVs neared 800,000, although not all of these would use cobalt to the extent that a Tesla S does.

Going forward, the EV outlook appears strong, with the IEA Global EV Outlook 2016 showing a “Paris Declaration” (Figure 3) EV stock build of some 40 million cars globally over the coming eight years. If each of these vehicles required 7 kgs of cobalt, then total additional demand for cobalt between now and 2025 would be 280,000 tonnes. This compares with global production of 123,000 tonnes per annum, i.e. it represents more than two years of global cobalt production. The website mining.com reported in October last year that the Democratic Republic of the Congo has a healthy pipeline of new cobalt projects and could see production rise by nearly 5% per annum in the coming years. If this continued through to 2025 then about half the shortfall would be covered.

Cobalt appears to be a critical component of the current generation of batteries. While there are many electro-chemistry formulations for rechargeable batteries, the use of cobalt in the cathode results in high energy density, important for vehicle range. But alternative chemistries offer other benefits, for example Lithium Iron Phosphate batteries have much better power output. Nevertheless, cobalt seems to feature in many of the battery chemistry formulations now being used for EVs.

Cobalt is primarily mined as a by-product of nickel and copper production, which is good news in that there will likely be significant additional demand for these metals as EV production ramps up. The same Tesla S battery discussed above has about 35 kg of nickel and the electric motor itself will have copper in the coil windings. Nearly half of global copper production is for electric motors, although alternatives may appear in the form of exotic material, e.g. carbon nanotubes.

Finally, there is a human rights issue associated with cobalt. About half the global production comes from the Democratic Republic of the Congo, where concerns have been raised regarding labour practices in the cobalt industry. Sky News reported on this recently but only mentioned the link with smart-phone batteries.

In recent months the cobalt price has doubled (Source: infomine.com), although it is clear that there is considerable long term volatility in this commodity, partly related to the fact that it is mined as a by-product of another process, so matching supply and demand can be problematic.

Making the most of hydrogen

dchone February 24, 2017

As hydrogen refueling stations begin to appear, including one from Shell opening near London, Toyota makes its Mirai hydrogen fuel cell vehicle more widely available and iron ore smelters even look to hydrogen as an alternative reducing agent (rather than coal) that doesn’t involve carbon, it would appear that this important energy carrier is beginning to show its potential.

The final energy (i.e. energy we use) mix in use today is less than 20% electricity, in large part dictated by the need to transport, store and use energy well away and disconnected from the point of production. Vehicles are a common example of this, with the requirement to power them on the move. While electricity storage is coming of age with batteries, hydrogen offers an alternative route forward that provides much of the flexibility we have come to rely upon with liquid fuels.

Hydrogen also offers the possibility of providing intense heat very quickly in a confined area, such as required by various industrial processes. For these reasons, and as society looks to shift away from direct use of fossil fuels for the purpose of carbon dioxide mitigation or simply needs to augment the energy system with other carriers, hydrogen shouldn’t be ignored.

Hydrogen production comes via three routes. The first is as a by-product from some industrial processes (e.g. chlorine manufacture). When I first started work in a Shell refinery I was assigned to follow the operation of the platformer, a unit which increases the octane number of naphtha by creating cyclic ring compounds (aromatics). These have a lower hydrogen content than straight chain molecules, so hydrogen is a byproduct. For the most part this source of hydrogen is in full use, either within other processes in refineries or in neighbouring facilities.

The second is steam reforming of natural gas, which produces hydrogen and carbon dioxide. Most industrial hydrogen in the world today is produced via this route. The carbon dioxide is relatively pure and can be used or captured and stored, making the process very carbon efficient. In the Shell refinery in the Netherlands some carbon dioxide from its hydrogen manufacturing unit is sold to local greenhouses to provide the enhanced growing conditions that they need. In Alberta, Canada, one million tonnes per annum of carbon dioxide from the Shell Scotford Refinery hydrogen unit is captured and geologically stored.

Finally, hydrogen can be produced by electrolysis of water, an experiment almost everybody has done in chemistry class at school. This route has renewable energy enthusiasts very excited as it can provide a ready use of electricity when wind and solar production exceeds demand, a situation many grids may face as renewable energy penetration reaches very high levels. But this process suffers from a relatively low efficiency, with only a few percent of global hydrogen being produced via this route. Generating hydrogen by electrolysis is only (optimally) about 60% efficient and the use of this hydrogen in a car is (optimally) also only about 60% efficient, so two thirds of the energy input is lost. This can be improved to some extent if heat recovery methods are employed.

The economics of the electrolysis route can be difficult, particularly if it relies on renewable energy surpluses, which won’t always be available. That would leave the electrolyzer idle or having to use purpose generated electricity. Further, most current electrolyzers vent the oxygen that is made, which is effectively waste energy and explains why the process can only be 60% efficient if hydrogen is the only product being made.

The global oxygen industry is very large; today it is around 500+ million tonnes per annum, with the metallurgical industry being the largest consumer. Oxygen is typically manufactured by cryogenic distillation of air. In the context of a hydrogen economy based on electrolysis rather than steam reforming of natural gas, oxygen production would be significant. According to Toyota, the Mirai consumes 0.76 kg H2 per 100 km. Using these numbers as a basis, a 200 million vehicle hydrogen fuel cell fleet might consume 23 million tonnes of hydrogen per annum, which when produced by electrolysis would also generate around 180 million tonnes of oxygen, or around one third of the current demand. This of course presumes it can be efficiently captured and transported.

Significant production of oxygen may also benefit another facet of an energy system striving for net-zero emissions. Biomass could be used for power generation, but in an oxygen combustion system (i.e. combusting the biomass with pure oxygen rather than with air which is 79% nitrogen). This means that the waste gas would be mainly carbon dioxide, which in turn would lead to easier capture and storage. This could then be an important negative emission technology in the net-zero sum for the economy as a whole.

The hydrogen economy offers many new opportunities and it is good to see that it is starting to move forward. The Shell hydrogen refueling station that opened in Cobham, uses an on-site electrolyzer to produce its hydrogen, with the electricity coming from a certified green source. It is the first of three hydrogen stations Shell plans to open in the UK in 2017.

Materials in a transition

dchone December 13, 2016

In thinking about the energy transition, it is easy to look at the status quo and compare it with a clean energy future of renewables, electric vehicles, bio-polymers and recycling then naturally opt for that future. After all, that future looks very attractive. What is often missing in this thought process is the required change in the global stock of materials to get there.

A good example of this comes from looking at the electric vehicle (EV). There are about 2 million globally or 0.2% of the passenger car fleet, but we might imagine a world in just a few decades where all the cars are electric. Some months ago I looked at what a highly accelerated EV transition might look like. It meets the 2020 goal of the Electric Vehicle Initiative of the Clean Energy Ministerial and is compatible with the Paris Agreement in terms of a 1.5-1.8°C goal (i.e. well-below 2°C with a stretch to 1.5°C). The outcome has internal combustion engine (ICE) vehicles largely departing the scene by 2060 and peak passenger vehicle stock of around 1.6 billion, compared with about 1 billion in 2016. This is a faster transition than many other outlooks, but it is a normative scenario designed to show the rate of change required for the Paris outcome to be realised.

Populating the world with 1.6 billion EVs requires building that many batteries and producing the materials that those batteries require. Although we can visualise a world of near 100% recycling of the battery components, the benefits of recycling don’t come until much later, as in a fast growth scenario the number being recycled after 15 years use is far lower than the number being produced. Even in 2060 in my rapid EV scenario, scrappage of EV vehicles still lags EV production in 2060.

The current Tesla Model S has a Nickel-Cobalt-Aluminium Lithium Ion battery. The cathode of the battery consists of about 35 kgs Nickel and 7 kgs Cobalt. While there are many different battery chemistries available, each offering different properties in terms of energy density, charging rate, hysteresis etc., all the chemistries depend on particular combinations of metals.

In a world in which the current Tesla chemistry dominated, the shift to a 100% EV fleet would require an on-the-road stock build of some 50 million tonnes of Nickel and 10 million tonnes of cobalt. This assumes an eventual global fleet of some 1.6 billion cars, as per my EV scenario. This stock is never recovered unless battery chemistry changes or the EV car population falls.

Current global production of Nickel is around 2 million tonnes per annum, with about 3% of that used for batteries. Nearly 70% of Nickel is used in the stainless-steel industry. Global cobalt production is around 110,000 tons per annum, with nearly two thirds of this coming from the Democratic Republic of the Congo. The remainder comes from about ten other countries.

Given the above, such a stock build would require 25 years of global Nickel production and 91 years of Cobalt production – assuming 100% of production is directed to the new EV battery industry. An accelerated EV scenario requires this level of stock build in under 35 years, which either means a rapid escalation of production in these metals or many different battery chemistries, but probably both. Competition will also come from grid batteries, home batteries such as the Tesla Powerwall and many other new battery applications, although these may also use alternative chemistries.

The initial proposition of the Tesla chemistry dominating isn’t simple conjecture either. As I reported in my blog from COP22 in Marrakech, Professor Jeff Sachs from the Earth Institute at Columbia University made the strong claim that there would be no further production of internal combustion engine vehicles after 2030. Scaling up battery production so rapidly would likely depend on an existing chemistry; there simply wouldn’t be time to wait for new chemistries to be researched, developed and perfected for mass production.

A potential energy transition outcome is one of significantly increased production of certain materials, diversity in battery chemistry, a slower than desired uptake in EVs and perhaps a much smaller EV fleet than anticipated thanks to autonomous driving and vehicle sharing. Nevertheless, some simple calculations quickly show that the transition is likely to be a complex one, with an end result possibly far from the expected.

Baby steps through Marrakech

dchone November 20, 2016

COP22 has come and gone and despite a surprise US election result in its early days which reverberated throughout the temporary site in Marrakech that was built for the meeting, progress was made in starting to turn the Paris Agreement into an operational treaty. This was of course helped by the early entry into force of the Agreement on November 4^th. That event also meant that COP22 became the first meeting of the Parties to the Paris Agreement, or CMA.1.

The COP had a clear division into two parts. The first seven working days were spent agreeing on the steps to take to reach a conclusion on the many different aspects of the Paris Agreement that require further detail, rules, modalities and procedures. This effectively sets the work plan for the discussions over the next two years to COP24 which will see two important aspects of the Agreement come into play;

The IPCC will present its findings on a 1.5°C goal in accordance with paragraph 21 of the accompanying decision text.
As per paragraph 20 of the decision text, a facilitative dialogue among the Parties will take place to take stock of the collective efforts in relation to progress towards the long-term goal.

Of particular interest to me was the discussion on Article 6 of the Agreement, which contains the potential foundation elements for a globally traded carbon market. There were also a number of side events and background discussions held over the first few days to help observers, primarily from the private sector, input their thinking into that process. The Parties agreed a process to share further views over the coming months, including a new round of more focused submissions and a round-table discussion at the inter-sessional meeting in Bonn next May. But the Article 6 discussions that were held and which observers were permitted to attend seemed to be held in the abstract, without significant reference to carbon market development and the wider adoption of carbon pricing policies at national level. For many of the Parties carbon pricing remains contentious, leaving others to walk on egg-shells as the Article is discussed.

The remainder of the COP involved the arrival of various Ministers and Heads of State, in what was called by one delegate more of a celebration of the Paris Agreement than adoption of measures to implement it. CMA.1 opened, but instead of closing as would normally happen at the end of a COP, it was suspended. The Paris Agreement requires some thirty work items to be concluded at the first meeting (CMA.1) following entry into force, but the unexpected early entry into force has made such a task impossible. Hence the suspension. This won’t go down as a memorable COP in terms of real progress, but it has effectively served to launch the process to operationalise the Paris Agreement.

Any COP, and this was no exception, is host to numerous side events, side conferences and functions to allow all the attendees to share views. At the Paris COP last year, carbon pricing was a particularly high profile subject, led by the World Bank and various other institutions. The Paris COP served as the launch platform for the Carbon Pricing Leadership Coalition (CPLC). While carbon pricing was discussed on the sides of this COP and the World Bank CPLC event was standing room only, low carbon technology was the more pervasive theme.

A Low Emissions Solutions Conference was held within the COP itself, with an interesting speech by Jeff Sachs of Columbia University as part of the opening. Professor Sachs made the bold statement that no more internal combustion engine (ICE) vehicles would be sold after 2030; this assertion was based on the pace of change he sees in the electric vehicle (EV) market. He asked the audience to hold him to his statement. I looked at a highly accelerated EV transition in a post earlier this year and even in such an extreme calculation it wasn’t until 2030 that the crossover between ICE and EV production occurred (i.e. 50/50 sales in 2030 rather than 100/0 as forecast by Sachs). This then led to a possible crossover in vehicle stock (i.e. more EV than ICE on the road) around 2040. The EV will require a range of new supply chains to be developed, such as for the specialty metals used in battery formulation (e.g. Nickel, Cobalt). Production on such a scale could require global production of some metals doubling in a short space of time.

The Moroccans did a splendid job hosting the event and should certainly be congratulated. Transport around Marrakech was excellent, to the extent that a free Uber service was available for COP participants. Next year the process reconvenes in Bonn, but with Fiji holding the COP Presidency.

100 Gigafactories

dchone November 8, 2016

Last week National Geographic launched its climate change movie, Before the Flood, online globally for free viewing. The film offers the opportunity to join Leonardo DiCaprio as he explores the topic of climate change, and discover what might be done today to address it. I sat down at home and watched it through YouTube on its opening night.

DiCaprio tours the world, talking to scientists, business people, Pacific islanders and many others. One riveting conversation is with Elon Musk, founder and CEO of Tesla, filmed as he and DiCaprio tour the Tesla GigaFactory in Nevada. This is a truly gargantuan building, with the largest physical footprint on the planet, designed to build batteries on an immense scale. The conversation moved onto the energy transition;

Elon Musk: We actually did the calculations. What would it take to transition the whole world to sustainable energy? What kind of throughput would you actually need? You need a hundred Gigafactories.

DiCaprio: A hundred of these?

Elon Musk: Yes, a hundred of these.

DiCaprio: That’s it?

Elon Musk: That would do the whole world, all energy!

DiCaprio: That’s it? That sounds manageable.

I agree with DiCaprio, building 100 GigaFactories seems like an entirely plausible venture. But what would this mean and what impact might it have?

According to the Tesla website, the GigaFactory will produce 35 GWh of battery capacity per annum. A 200-300 mile range Tesla currently has an 85 kWh battery, which means that a Gigafactory can produce about 400,000 such batteries per annum. 100 Gigafactories would produce at least 40 million such batteries per annum. There are also reports in the media of the current Tesla Gigafactory being expanded, which means there is upwards room in this estimate. Although Elon Musk made mention of 100 Gigafactories, we don’t know exactly what capacity per factory he was assuming. But on this basis and assuming improvements in Electric Vehicle (EV) efficiency and also allowing for smaller city vehicles which would have smaller batteries, the 100 Gigafactories should be able to supply batteries for all new vehicles in a 100% EV fleet, assuming some 80 million new cars per annum (currently 70 million cars per annum).

Such an outlook potentially meets all the battery requirements for cars, but would that put sufficient storage into the energy system to balance the intermittency requirements for a 100% renewable energy grid? And is this sufficient for all energy as Elon Musk stated?

Over the course of an energy system transition to net-zero emissions such as illustrated in the recent Pathways to Net-Zero Emissions publication from the Shell scenarios team, we could potentially see the electricity system grow from some 80-90 EJ today to 300 EJ, a several fold increase. This results from both a growth in the overall size of the energy system and a shift in electrification from some 20% of final energy now to over 50%. If that system is to be powered entirely by renewable energy, as imagined by Elon Musk, it would need long distance grid connection and some form of energy storage. Assuming grid connections are not truly intercontinental, e.g. joining the Americas with Europe and Africa, then each continental mass would need to manage its own intermittency.

In 2015, North American power generation was in excess of 5000 TWh, but in a highly electrified energy system, demand would likely be significantly higher. Assuming at least 300 million cars in North America, that might represent up to 20 TWh of storage if all are EV and available for backup (N.B. In a highly accelerated-EV scenario I developed for an earlier post, the global EV stock only exceeds 300 million in 2031 but reaches 1 billion in 2042). While it is highly unlikely that renewable energy generation would drop to zero across all of North America at any one time, this level of storage could theoretically power the grid for around a day. Given that over 90% of cars are parked at any given time, instantaneous vehicle use isn’t really a factor, however the backup requirement would need a significant percentage of the parked cars to be physically connected to the grid at any one time.

Is this enough storage? The answer to that isn’t simple at all and probably requires a complex computer simulation of the entire system. Certainly if the grid is heavily dependent on solar then in winter there are several hours a day when the entire North and South American continents are in darkness. But given a distribution between wind, solar and hydroelectricity, together with continental inter-connection, this amount of storage feels at least plausible.

But this isn’t all energy, as even in this very electrified scenario only 50% of final energy is actually electricity. The remainder is largely hydrocarbon based in services such as aviation, shipping, heavy ground transport and a wide variety of industrial applications (cement, metals refining, chemicals etc.). These hydrocarbons might still come from fossil fuels in which case CCS will be required or they could originate from biomass.

A world of 100 Gigafactories all building vehicle batteries for an all EV fleet, with a grid powered by wind and solar, reliably backed up by those same vehicles connected to the grid, is aspirational given where we are today. It would require considerably more than just building the Gigafactories and such a transition would be measured in many decades rather than many years. But it’s certainly an interesting outcome to consider.

“The New Lens Scenarios” and “A Better Life with a Healthy Planet” are part of an ongoing process – scenario-building – used in Shell for more than 40 years to challenge executives’ perspectives on the future business environment. We base them on plausible assumptions and quantification, and they are designed to stretch management thinking and even to consider events that may only be remotely possible. Scenarios, therefore, are not intended to be predictions of likely future events or outcomes, and investors should not rely on them when making an investment decision with regard to Royal Dutch Shell plc securities.

It is important to note that Shell’s existing portfolio has been decades in development. While we believe our portfolio is resilient under a wide range of outlooks, including the IEA’s 450 scenario, it includes assets across a spectrum of energy intensities including some with above –average intensity. While we seek to enhance our operations’ average energy intensity through both the development of new projects and divestments, we have no immediate plans to move to a net-zero emissions portfolio over our investment horizon of 10-20 years.