Tokyo in a power crunch

On March 22, 2022 the Tokyo Electric Power Corporation (Tepco) warned electricity consumers in east Japan about the risk of rolling blackouts from a tight supply situation. The recent M7.3 quake near Sendai had knocked several of Tepco’s thermal power plants offline, which left the company in a difficult situation when a cold spell with snow flakes hit the region of the capital. Demand at times exceeded generation capacity and only the availability of pumped hydro storage saved the day before measures to curb demand such as turning down heating and switching off lights averted an outage.

No doubt this experience will increase pressure to restart more nuclear power stations that have been shuttered since the tsunami and nuclear meltdowns in Fukushima in March 2011. Before the nuclear disaster about 30% of Japanese generating capacity were nuclear; now only about 10% comes from restarted nuclear reactors. The current high prices of natural gas will further enhance the attraction of nuclear, at least in the eyes of anyone whose financial interests are tied to the balance sheet of the utility companies, such as their individual and institutional shareholders.

However, that is not the whole story.

While eastern Japan was in a power crunch, western Japan has ample spare capacity, as did Hokkaido. Why could this power not be used in Tokyo? You would have thought Japan would have learnt its lesson from the 3/11 disaster in 2011 and addressed it in the decade since then, but you would be wrong: Japanese electricity markets are still split between a handful of regional near-monopolies with minimal interchange capacities between them. For example, the Hokkaido grid has a generating capacity of 7.5 GW but only 0.6 GW of interchange capacity with Honshu (8% of the total). Tepco supplies up to 47 GW to customers in its area but can only exchange up to 1.2 GW with major utilities in the west of Japan. This leaves little margin when earthquakes or weather events with a regional impact hit supplies.

By contrast, China has built huge high voltage direct current (HVDC) transmission lines between the industrialized coastal cities on one side and hydroelectric power stations near the Tibetan plateau and solar and wind farms in the arid north on the other. Many of these lines are longer than the distance from Tokyo to Hokkaido, let alone Tokyo to Kansai. The Chinese government understands that if it wants to wean itself from the dependence of dirty coal or imported oil and gas then it will need to vastly increase power transfer capacity from the interior of the country where renewables are available to the densely populated urban areas near the coast lines.

Japan is actually in a similar situation. The elephant in the room that nobody wants to talk about is offshore wind. While European countries and the US are building up tens of Gigawatts of offshore wind power capacity, Japan has very little installed capacity, particularly offshore. The entire conversations seems to be about nuclear vs. solar vs. gas vs. coal, leaving out one of the most promising renewable energy sources available to Japan. So far the regulatory hurdles for erecting and connecting wind turbines in Japan have been high and that has left wind as an also ran compared to much more widely deployed solar. However, solar does not provide power at all hours. Wind would complement it.

Much of the European wind power capacity is installed offshore where wind speeds tend to be high and more consistent than onshore. This is where the largest and most economical turbine models tend to be used. By contrast, almost 99% of Japan’s wind power capacity is still onshore. A cumulative total of only 51.6 MW of offshore wind capacity was installed at the end of 2021 while total installed wind power capacity was 4.6 GW. Meanwhile the UK had 24.7 GW of wind power capacity, Spain 27.1 GW and Germany 62.2 GW. China is in a league of its own with 282 GW, more than all of Europe combined. Japan’s installed wind power base is less than that of small European countries such as Belgium (4.7 GW) that have relatively short coast lines and tiny EEZs: Japan’s EEZ of 4,479,388 km2 is over 1000 times larger than Belgium’s at 3,447 km2!

Japan is really only starting to build up offshore wind capacity, with projects off the coasts of Akita, Chiba and Nagasaki getting under way in the last two years. By 2030 its goal is for 10 GW of offshore capacity either installed or under construction which is still tiny compared to the already installed base of Germany, Spain or the UK.

Unlike fossil fuel or nuclear power stations, wind turbines are not location independent. They will be installed where wind conditions are favourable, where the sea is not too deep and connections to the coastal grid are cost-effective. To make the most of the wind conditions, the grid will need to be greatly expanded to allow large amounts of power to be transferred from regions with plenty of wind to regions with many consumers. This will be quite different from the current model where utility companies try to generate all the power they need within their own region, which is why there is only limited interchange capacity to help out if one company loses a large part of its generating capacity as happened in the recent quake or after 3/11.

Japan needs to start building high capacity long distance HVDC power lines like China has in order to enable a transition to zero carbon electricity. The fragmented power markets dominated by local utility companies are an obstacle to this transition as the interests of the regional companies seeking profits from existing investments in their area are not aligned with the interests of the consumers who want reliable green energy regardless of where it comes from.

Japan quickly needs to remove regulatory obstacles to expanding wind power and then invest to build a HVDC backbone to connect renewable power generation with consumers.

Hokkaido wind power for Japanese energy

Nikkei reports (“Japan pushes for undersea cables to solve wind power puzzle”, 2022-01-02) that the government is allocating 5 billion yen (about US$43 million) in its supplementary budged for a feasibility study for a 4 GW high voltage direct current (HVDC) link between the power grids of the northern island of Hokkaido and the main island of Honshu, where most of Japan’s population lives. This would be by far the biggest HVDC link ever built in Japan. The Japanese government wants to generate 45 GW of power from offshore wind in 2040, up to about a third of which (14.65 GW) is to be produced in Hokkaido. The development plan lists several promising offshore areas along the southwest coast of Hokkaido.

For this power to be available to consumers outside the northern prefecture, it would need to be exported via a HVDC link. This is the preferred technology for shifting large amounts of power over long distances, especially between AC grids not synchronized with each others or operating on different frequencies. Since 2019 there have been two 300 MW HVDC links between the two islands. Their combined capacity is to be doubled to 1.2 GW by 2028.

Japan has relatively little capacity for transferring power between its regional grids. This is because its grids used to be operated by regional monopolies that had little incentive to ever import or export power. This lack of interconnect capacity became a major problem following the power shortage after the 2011 Tohoku earthquake and tsunami when less affected areas could not help out the most affected region. There is a conflict of interest between the local utility companies and the country as a whole. Tepco owns a lot of nuclear power stations, expensive infrastructure with huge sunk costs. It would rather generate power from these plants than pay another supplier from outside its area for renewable energy. However, many of these power stations have yet to be restarted since their shutdown following the Fukushima meltdowns. By restricting how much power can be imported from other grids, Tepco can put pressure on regulators to allow it to restart more reactors to ensure a stable supply of power. On the other hand, expanding interconnect capacity would ease the pressure. Which side will the Japanese government take?

A related issue is the variable output of renewable power sources. Long distance transmission will make it easier to compensate for local weather patterns by shifting power between different regions, which allows a larger share of renewable energy to become part of the mix without having to resort to either energy storage or peaker plants (e.g. gas turbines to cover peak loads). That again means Tepco loses leverage to maintain coal and other fossil fuel powered generating capacity as insurance against shortfalls of renewable energy.

China, one of Japan’s main economic rivals in the world, has pursued a completely different course. Over the past decade it has aggressively expanded long distance HVDC links to stabilize its grid. Japan operates a single HVDC link of at least 1 GW, a 1.4 GW link between Honshu and Shikoku that started operating in 2000. All other links are only in the several 100 MW range and most of those are not long distance lines but back-to-back local interconnects, for example between the 50 Hz grid of eastern Japan and the 60 Hz grid of western Japan near Nagoya. By contrast, China has built over 20 HVDC links over 1 GW, mostly with a capacity of 3 GW or more. Many of the biggest projects cover distances of 1,000 to 2,000 km. This allows China to supply it coastal megacities with hydroelectric power from its southeastern mountains or from other power sources from its arid central parts. China is the world leader in wind power. Its windiest parts are along its border to Mongolia and on the Tibetan plateau. Large scale HVDC is key to China’s energy policy for the 21st century.

An alternative to shifting power long distance is to use it to locally generate hydrogen from water (“green hydrogen”) and feed it into pipelines or use it to make ammonia. This makes some sense for applications that already use hydrogen, such as the fertilizer industry or for carbon free alternatives to existing technology, such as direct reduction of iron ore for steel making without using coking coal. However, it makes little sense to use green hydrogen for power generation: if you convert electricity to hydrogen which you then use to generate electricity, more than 70 percent of energy is lost in the process while less than 30 percent remains. By contrast, batteries are 90 percent efficient. Therefore, if excess wind or solar power is used to produce hydrogen, that resource should best be used by industries that directly consume hydrogen, until all fossil fuel currently used for such purposes has been replaced.

If Hokkaido had a surplus of hydrogen from wind power, it would make more sense to have it consumed by steel works and fertilizer plants built in the prefecture rather than sending it through a pipeline to Honshu.

Although green hydrogen or ammonia can be used as fuel in thermal power plants in place of coal or LNG, it would be a terribly wasteful use. Because of the huge conversion losses, we would need three times more wind or solar power to end up with the same amount of usable electricity than if we used grid-scale battery storage to absorb any surplus and make it available when needed. This advantage makes grid-scale battery storage a strategic technology.

Most existing Li-ion batteries depend on relatively scarce resources such as cobalt, nickel and lithium. Lithium-iron-phosphate (LFP) batteries only require lithium and widely available materials, while sodium ion batteries use only readily available raw materials. Japan will need to invest in high capacity long distance HVDC links as well as in battery storage to speed up its transition to a carbon neutral economy.

Toyota Hydrogen Combustion Engine Cars

Since 2014 Toyota has sold a little over 10,000 Toyota Mirai, a hydrogen fuel cell vehicle (FCV). The starting price of this 4 seat sedan model in Japan is about 7.1 million yen (currently about US$63,000) which is more than 50% more expensive than a battery electric Tesla Model 3 which seats 5 adults. And it seems unlikely that Toyota can make a profit on a car being made in such small numbers as the Mirai, unlike Tesla does with the cars it makes in large numbers in its plants on three continents.

Tesla sold about half a million battery electric vehicles (BEVs) last year and looks set to sell somewhere between 900,000 and 1 million cars in 2021. This means Tesla will have sold twice as many BEVs every week in 2021 than the total number of FCVs Toyota has sold since 2014. The sales gap between BEVs and FCVs is getting bigger and bigger.

Recognizing that the high cost of fuel cells makes it difficult to compete, Toyota has announced that it sees a market for cars with internal combustion engines (ICE) that burn hydrogen instead of gasoline. They should be cheaper to make than fuel cell cars and will not produce any CO2 if hydrogen is made from non-fossil energy sources.

It’s not a novel idea though. BMW tried it in its BMW Hydrogen 7 technology carrier based on its 7-series back in 2005-2007. It never went anywhere. Besides the absence of a fuel supply network, there were also issues with emissions. Hydrogen flames burn extremely hot, which means you end up with a lot of smog-forming NOX emissions — worse than diesels.

In terms of efficiency, hydrogen ICEs are worse than FCVs which are much worse than BEVs. While BMW used cryogenic tanks with liquefied hydrogen at -253 °C, Toyota most likely will use high pressure tanks like in its Mirai for its hydrogen ICEs. They hold hydrogen gas at pressures of up to 700 bar. Both liquefaction and compression require huge amounts of electricity that can not be used for propulsion but is effectively wasted. An FCV consumes three times more electricity for electrolysis to make the hydrogen fuel it consumes than a BEV uses to charge a battery to drive the same distance. A hydrogen combustion engine is even less efficient. Where will this hydrogen come from? We don’t currently have a surplus of solar panels or wind turbines to produce this electricity. That means a hydrogen economy will need significantly larger investments in renewable energy than with battery vehicles. Hydrogen for cars makes no economic sense whatsoever.

It makes even less sense for hydrogen ICEs than for hydrogen FCVs. Fundamentally, it’s no more than an excuse for not giving up on building internal combustion engines, pretending that nothing has changed even in a world that is facing climate change that we need to address as soon as possible.

I am afraid Toyota will not make a turn-around and face the reality that the industry is switching to BEVs within the shortest time possible until it replaces Toyoda Akio, its current company president. Mr Toyoda is the grandson of the founder of the company and a keen race car driver. He lacks the vision that Toyota will need in the transition to a carbon free future. Mr Toyoda needs to retire, along with the dead-end technologies he is committed to.

Subaru announces the Solterra, it’s first battery electric car

Perhaps not by coincidence Subaru chose the week of the COP26 climate summit in Glasgow to launch its first battery electric car, the Solterra (the name is a portmanteau of the Latin worlds for sun and earth). To say that it’s based on the same “e-TNGA” electric vehicle platform as the Toyota bZ4X understates how much the two cars have in common: They are basically one and the same car fitted with different badges. Even the wheels are the same. You have to look very carefully at this pair of genetically identical twins until you find a minor detail that distinguishes them: Yes, the rear lights are a bit different.

Toyota owns 20% of Subaru and they have shared models before (Toyota 86 / Subaru BRZ), but I did not expect to see so little recognizable Subaru DNA in their first battery electric vehicle. Yes, there is a four wheel drive model of both the Solterra and the bZ4X and one assumes that Subaru had a hand in design choices for this, but 4WD is by no means unique for BEVs, as models ranging from the Tesla Model 3 to the Volkswagen’s ID.4 are also offered in dual motor 4 wheel drive configurations. Even the hybrid Prius is available in an electric 4WD version.

What seems a little odd is that the non-4WD model is front wheel drive (FWD). In internal combustion engine (ICE) cars, FWD offers some advantages as it saves having to have a long drive shaft between the front engine and the rear differential. The engine and the gearbox can be bolted together and directly drive the nearby front wheels. At the same time the weight of the engine and gearbox provides good traction for the driving wheels, especially in wintry conditions.

With a BEV however, the bulk of the weight is not in the engine but in the battery under the passenger compartment. Thus there is no real advantage in driving the front wheels as opposed to the rear wheels.

An electric motor driving the rear wheels can be very compact, not much bigger than the rear differential and exhaust system in rear wheel drive (RWD) ICE car. Without the traction advantage of the engine over the wheels, it would be better to go for RWD to get more weight on the driving wheels when going uphill or when accelerating. The turning circle would benefit too if the driving wheels don’t have to steer. It is no coincidence that both Tesla and Volkswagen use RWD for their BEVs, in the case of Volkswagen despite the fact that its best selling models such as the Golf and Passat are FWD. So why not Toyota and Subaru? It’s a mystery to me.

Another detail that surprised me was that even though DC charging on this car can reach a respectable 150 kW, AC charging at home is limited to mere 6.6 kW, which is less than for a compact Chevy Bolt. A Golf-sized ID.3 actually handles up to 11 kW. Some of this may be due to the Japanese Chademo charging standard and domestic grid considerations, as Japanese households only have access to 100 V and 200 V single phase current while the US and Europe use the CCS standard and 120 V / 230 V respectively, with 400 V 3-phase AC available anywhere in Europe. So even if there were technical reasons for limited AC charging speeds in Japan, export models should be able to do much better. Toyota may have specified its home charging module to the smallest common denominator, which if true is a bit disappointing.

As for the looks of the Toyota bZ4X / Subaru Solterra, to me they look like a close cousin to the existing Toyota RAV4 that I personally do not find very appealing. However, it is a big seller in the US market and this similarity may help move existing RAV4 owners over to BEV models once they become available some time in 2022.

Toyota has never been enthusiastic about battery electric vehicles. Its official line has been that hybrids are good enough for today and tomorrow we’ll get hydrogen fuel cell cars like its own Toyota Mirai, with all the benefits of battery electric but none of the drawbacks. There was no real space for battery electric in this vision. Toyota clearly over-promised and under-delivered on this strategy: Hybrid cars still spew CO2 into the atmosphere while almost all hydrogen today is made from fossil fuels. Battery electric does much better than that.

In Japan Toyota could rely on the government to help promote its “hybrids today, hydrogen tomorrow” story but in international markets that won’t fly. There the war for the future of the car is over and battery electric won hands down. No other country has a comparable push for hydrogen refuelling infrastructure as Japan has. Even if there were a domestic market for hydrogen cars in Japan, there won’t be any export markets.

Most experts agree that hydrogen vehicles are at least three times less energy efficient than battery electric vehicles, a flaw that would kill them even if the cars and the necessary fueling infrastructure could be built for the same cost, which isn’t the case. Batteries are far cheaper than hydrogen fuel cells and DC chargers are cheaper than electrolysers and hydrogen fuel stations. With battery prices falling further and further, within a few years BEVs will become cheaper to build than hybrid cars. Then the speed of conversion will only be limited by battery production capacity. It’s not clear Toyota will have the right investments in place by then, since it says its future BEVs will eventually be using solid-state batteries, an as yet unproven technology that only exists in the lab.

Until now Toyota had been avoiding BEVs except for the Chinese market, as it hoped buyers would keep buying its existing more profitable hybrid models. That is becoming a risky bet. Drastic changes needed to avoid the worst of a climate disaster no longer seem so radical compared to worldwide measures taken to deal with SARS-CoV-2. Huge numbers of consumers are ready for change. New BEVs by competitors are picking up market share in the US and in Europe. Toyota can no longer afford to wait on the sidelines or it will be seen as becoming irrelevant due to obsolete products.

This new BEV model is a very cautious move by Toyota and Subaru. Instead of competing head on with Tesla or Volkswagen, Toyota and Subaru are entering the BEV market only about as far as they absolutely have to, to still be a credible global player in 2022. The two companies will have to up their stakes to keep up with market developments.

METI and Japan’s exit from the Carbon Economy

On the eve of COP26, the UN Climate Conference in Glasgow, Scotland, the Japanese government took out a full page ad in the Japan times to talk about “beyond zero”, a series of events and initiatives related to Climate Change. It struck me that none of them were specifically about renewable energy, the essential ingredient for a carbon-free economy.

The title of “Tokyo Beyond Zero Week” already had me confused: It reminded me of the Toyota bZ4x, a battery electric SUV that is the first mainstream battery electric vehicle for the Japanese market that Toyota has announced. Toyota has become notorious for bucking the Battery electric trend by plugging hybrids and hydrogen fuel cells, despite hydrogen fuel from renewable sources being 3 times less energy-efficient than battery electric vehicles. The bZ4x is too little, too late when Toyota is telling potential customers that they should really be buying hybrids like the Prius or hydrogen fuel cell vehicles like the Mirai.

METI, the Japanese Ministry of Economy, Trade and Industry has been sponsoring vehicles based on hydrogen fuel cells using hydrogen made from Australian brown coal (lignite), with the resulting CO2 emissions sequestered using “carbon capture and storage” (CCS) and the hydrogen shipped to Japan in cryogenic tank ships developed by Japanese shipyards with METI funding. Essentially it’s a massive pork barrel project, designed to pay industry players to go along with a Rube Goldberg project that will not be economically viable. It’s a way of keeping ecological laggards such as Toyota and the huge Japanese shipbuilders and trading companies relevant. Some of the initiatives sponsored by METI are:

  • LNG (Liquified Natural Gas) Producer-Consumer conference
  • International Conference on Carbon Recycling
  • International Conference on Fuel Ammonia

There is no place for LNG in a zero carbon economy. “Carbon Recycling” aka CCS is a fig leaf to keep burning fossil fuels. Ammonia may be a necessary fuels for ships and airplanes, but if it’s made from coal it won’t be green energy.

Why is the METI ad not talking about offshore wind and geothermal power, two of the most important energy sources for green baseload electricity? It’s because they are primarily concerned about creating and maintaining business opportunities for Toyota, trading companies making profits from fossil fuel imports and other companies wedded to the fossil fuel industry and not about how to get Japan ready for the zero carbon age.

I find this very sad. As a country with limited fossil fuel resources, Japan could become a prime player in the post-carbon era, developing new technologies to help other countries move beyond fossil energy sources. Japan has huge opportunities in offshore wind, onshore wind, solar and geothermal but its government has been largely turning a blind eye to them because those energy sources can not be controlled by its big trading companies. Likewise, its biggest automobile manufacturer is a laggard in battery electric vehicles which is determined to sabotage the switch to BEVs.

Tesla 4680 cells and bad journalism

Tesla and Pansonic have introduced the new 4680 battery cell that future battery packs for the Model Y and the Cybertruck will be based on. These larger cells will replace the previous 2170 form factor that current Tesla packs are based on, which in turn replaced the 18650 cells that Tesla inherited from the laptop industry.

Some of the articles about the new cell have talked about the 5 times higher capacity of the cells saying it would address the problem of “range anxiety”:

5 times more energy means less range anxiety and more drive time. It means fewer stops on a road trip and a more enjoyable experience.
(Why The Tesla Tabless Battery Is So Good, torquenews.com, 2021-03-30)

Actually, this claim is embarassingly disingenuous.

Yes, the new cells have higher capacity but that’s because they’re bigger, which means a battery pack of a given capacity will be built from fewer but larger cells. The bottom line of capacity by weight or by volume is largely unchanged.

The new cells are 2.2 times the diameter of their predecessors, meaning they will have a cross section 4.8 times larger, so a given number of square meters of floor plan for a particular vehicle will fit 4.8 times fewer of these larger cells with each storing about five times as much energy as their smaller siblings. If you think this makes for 5 times more range then I have a bridge to sell to you 😉

The cells are also 80 mm long instead of 70 mm, but for energy density it’s basically a wash: The energy density per liter or per kg is unlikely to be vastly different.

Another point of confusion is Tesla’s claim that the cells will have five times the capacity but 6 times the power output. Some articles have interpreted that as 20% more range which is not the case. The truth is that the new cells can be discharged 20% faster without overheating but the total amount of energy released is unaffected by that. It’s like saying a car with 120 HP will have 20% more range than a car with 100 HP because it can drive faster. In reality it will burn fuel more quickly while doing so. This is strictly about peak power (energy by time), not total capacity.

The reason for the higher output is that the new batteries are tabless. All cylindrical Li-ion cells consist of two layers with a separator layer in between, wrapped up as a roll. Think of a double ply roll of toilet paper. When Tesla switched from 18650 to 2170, they made the roll wider (65 mm to 70 mm) but also made made the rolled-up sandwiched layers longer, giving the roll 21 mm instead of 18 mm of diameter.

This increased capacity per cell but it also meant that when energy is released in the ion exchange between the two layers in the innermost part of the cell, the current needs to flow round and round the rolled up layers until it reaches the tabs soldered to the exterior from where the power is transferred to the two opposite end of the cell.

The tabless design does away with that. In it, all the top edges of one layer touch each other and the battery pole at the top while the bottom ends of the other layer touch each other and the bottom pole. That dramatically shortens the path of the conductor through which current needs to flow. Internal resistance and waste heat are greatly reduced.

The bigger diameter means that the exterior steel skin of the cell is lighter relative to the reactive parts inside for some weight savings.

Not directly related to the bigger format, the new cells also break new ground by making do without any cobalt in their anodes which rely on nickel instead. Unlike cobalt which is primarily sourced from the Democratic Republic of Congo (a troubled country with huge corruption and human rights problems), Nickel is available from sources worldwide.

Several online articles have also repeated a claim that the new cells have a capacity of 9,000 mAh vs the approximately 5,000 mAh of the 2170 cells. This is way off the mark and must be based on bad arithmetic. To be consistent with Tesla’s claim of 5 times the capacity per cell, it would have to have about 25,000 mAh of capacity. That is also consistent with the quoted capacity of a 4680 cell quoted by a Chinese supplier of Volkswagen, which is also looking at using this format in the future.

LFP cells and the 4680 form factor

Personally, I think it would be great to also see a LFP (Lithium Iron Phosphate) version of 4680 cells. Panasonic announced that they would not be making it, but some of Tesla’s Chinese suppliers might opt for this format, which would work well for entry level models. LFP is a very safe chemistry and has a long cycle life, even if the energy density is somewhat lower.

In any case, it makes more sense for BEVs not to have the highest battery capacities possible but instead for some of the battery inventory to be used for infrastructure to decouple quick charging from available grid capacity: A certain percentage of annual battery production should be installed in chargers instead of in cars. Actually, recycled batteries from scrapped BEVs make a lot of sense for this, but so do different chemistries such as redox flow batteries including iron batteries.

If for example, most cars travel less than 150 km per day it does not really make much sense that they have a large but heavy battery that gives them 400 km of range but costs a lot of money and whose weight increases electricity use when accelerating. More weight also means more tire wear.

On the few days that cars need to travel further than their limited range, they should be able to quickly recharge from supercharger stations that use on-site battery storage to be able to recharge cars regardless of whether the grid has spare capacity at that moment or not. This is a far more efficient use of scarce resources than giving all BEVs a huge battery and makes for a more robust electricity grid.

The TerraPower Natrium Reactor – a Quick Review

TerraPower, a company funded by billionaire Bill Gates, has teamed up with several partners to build a demonstration nuclear power station in Wyoming by the end of the decade. Several sites are under consideration. The plan is to re-use the grid connection of a former thermal coal power plant, of which Wyoming has many.

The Natrium reactor developed by TerraPower in cooperation with GE Hitachi Nuclear Energy is quite a departure from the design of the light water reactors (LWRs) that produce the bulk of nuclear power worldwide today. For one, its output is highly variable because it incorporates gigawatthour (GWh) energy storage using tanks of molten salt. The design is quite innovative, which creates both upsides and challenges.

After reviewing the company’s website and watching a webinar, I am quite impressed but also concerned. The reactor will still run on uranium and will produce radioactive fission products that will need to be contained and stored safely for thousands of years. This is still a largely unsolved problem. Countries that have been generating power from nuclear fuels are today sitting on thousands of tons of waste kept in intermediate storage, still without a proven long term storage solution. Eight decades since the start of the “atomic age” with the Manhattan Project that gave us nuclear reactors and atomic bombs we are only now seeing the first permanent storage site being used in Finland. Some consider this the Achilles heel of the nuclear industry. Proponents of nuclear power will argue that, given we already have existing waste, this is a problem we will need to address anyway and that the volume of highly active nuclear waste will remain relatively compact. Nevertheless, there is a lot that can go wrong there, especially if the volume keeps increasing.

What most excited me about the reactor concept was its incorporation of the heat store using molten salt tanks, which it borrowed from concentrated solar power (CSP). Everything from the molten salt tank to the grid connection is basically the same as in this type of solar power plant. The major difference is that the heat source is not solar power focused onto a tower by thousands of mirrors but an underground nuclear reactor. This means the designers could use existing technology developed to maturity over the last 2-3 decades for use in solar projects in Nevada, Australia, Morocco and other locations.

This part of the plant is conventional technology that will not be subject to the same regulatory oversight as the nuclear portion, making it easier and cheaper to build. At the same time, the nuclear portion of the plant is much smaller and simpler, requiring a lot less concrete and steel than in a LWR per MW of output capacity.

By incorporating the heat storage, the electrical output of the power station can be varied considerably – the TerraPower presentation showed a range of about 240 to 500 MWe, with 345 MWe available continually without charging or discharging the heat store. Output that varies by 100 percent roughly covers the demand swing between day and night in many power markets. If combined with solar and wind, the stored heat can be used to smooth out fluctuations in power output from those natural energy sources. Heat from the power station may also have applications for desalination, industrial processes and residential heating.

Conventional nuclear power stations such as PWRs or BWRs can not vary their output very much. They basically can only run at 100 percent load or be switched off. Once shut down, bringing them back up again takes a very long time. That makes them suitable only for base load but not for demand peaks such as in the afternoon or evening. For that they would have to be combined with energy storage such as pumped hydro, opportunities for which are limited by geography. Due to the literally built-in output flexibility of the salt storage system, a zero carbon grid could theoretically incorporate a lot more Natrium output capacity than would be possible with existing LWRs. From an economic point, it means the operators in a competitive electricity market with bidding for supplies can sell more power at lucrative peak prices instead of having to try to find buyers at night when demand and prices are low.

So what’s the catch? The nuclear reactor itself is a sodium-cooled fast reactor (SFR), basically a Fast Breeder Reactor (FBR) without the breeding: Except for the absence of a breeding blanket made of depleted uranium that slowly turns into plutonium, the technology is very similar. Perhaps you remember the Monju reactor in Fukui, Japan that was shut down after a major accident in 1995. The operators attempted to hide the extent of a coolant leak and fire, which led to a 15-year shutdown. After a second accident in 2010 the reactor was eventually decommissioned. In 1966 the prototype Fermi 1 FBR in Monroe, Michigan suffered a partial meltdown. It was permanently shut down in 1972. Several other sodium-cooled fast reactors have been built around the world, such as the French Superphénix, the Prototype Fast Reactor in Dounreay, Scotland and the SNR-300 in Kalkar, Germany. All of the above have since been shut down due to high costs or troubles or, like the one in Kalkar, were never even started up.

While sodium has a high temperature range between melting and boiling point and is a good heat conductor, it also reacts violently with water and oxygen. Naturally, you can not put out a sodium fire with water. Normally the top of the reactor vessel is filled with an inert gas such as argon to prevent sodium fires but it needs to be opened for loading and unloading fuel, during which time the sodium has to remain heated above its melting point. You do not want to start a fire then.

If an LWR overheats, steam bubbles will form that reduce the criticality, interrupting the chain reaction. By contrast, control of the chain reaction in SFRs depends 100 percent on positioning of the control rods.

While the cooling pipes will not have to withstand high steam pressures as in a BWR, they will be subject to thermal stress: The coolant temperature in an SFR is much higher, around 550 deg C (1020 F) which is basically red-hot and hot enough to melt some aluminium alloys (and of course salt, for the heat storage). When SolarReserve decided to build a molten salt CSP solar power station in Nevada, it turned to Rocketdyne to make some of the metal parts, because of their metallurgical expertise in rocket engine nozzles that are also exposed to high temperatures.

There are other viable solutions for base load in a zero carbon grid, such as geothermal power, utility scale battery storage, thermal storage using rock heated electrically with surplus wind and solar or green hydrogen powering fuel cells or gas turbines. Costs for wind, solar and battery storage have been falling rapidly for years. Once renewables are cheap enough, you can partially address issues of intermittent output by overbuilding capacity and simply idling some of it when not needed. Or you can use spare output when supply exceeds demand to produce hydrogen, for making ammonia and for use by the steel industry.

Some of these solutions depend more on geography than the Natrium reactor, which can be installed on any continent and provide power at time of day and in any season. However, it would definitely need to be safe and reliable. Ultimately, this new technology will first have to prove itself.

Expiring the Internal Combustion Engine Car

The US state of Washington has decided to ban sales of new cars with internal combustion engines (ICE, gasoline or diesel) by the year 2030. That is five years earlier than in the state of California.

There are two issues to overcome for a switch to battery electric vehicles (BEVs): supply and charging. Two common worries however will not stand in the way of BEVs replacing ICEs: cost and range. Let me explain.

Battery cost per kWh has been dropping for decades and this trend is expected to continue. THis is highly significant: Most parts of a BEV car other than the big battery cost either the same as in an ICE car or they’re cheaper. As a result, the cost of batteries will stop being a major obstacle to adoption of BEVs years before the end of the decade.

The same is true for range. Cheaper batteries mean BEVs with more capacity will become affordable. The higher the capacity, the more km of charge can be replenished in a given number of minutes. For example, a Nissan Leaf with a 40 kWH battery will fast-charge from 0 to 80% in 40 minutes. The Volkswagen ID.4 First Edition with an 82 kWh battery (of which 77 kWh are usable capacity) will go from 5% to 80% charge in 38 minutes, essentially double the charging speed (kWh added per minute) for a battery with twice the range. If you can add hundreds of km of range in the time it takes you to use the toilet and get a cup of coffee then BEVs will be just as viable for long distance trips as ICE cars.

By the middle of this decade there is likely to be a wealth of different battery electric vehicle models on the market, with even BEV laggards such as Toyota, Honda and Subaru having joined in. Production could increase to about 50% of new sales of several large makers (e.g. GM, VW). It will have to scale up further, with the necessary increase in battery production capacity, by the end of the decade to make this happen but it seems eminently doable. Right now, the major bottleneck to ramping up production is not lack of demand but limited availability of battery cells. Every big car maker getting into BEVs will have to build Gigafactories churning out battery packs, or team up with battery makers who make these huge investments.

The more BEV there will be on the road, the more the impact on the electric grid becomes an issue. If you have a car that can cover 300 km or more on a full battery and you can charge at home every night then most likely you will almost never have to seek out a charging station, unlike drivers of ICE cars who regularly will have to fill up at a gas station. BEVs parked in a driveway or garage with a nearby wall socket are much easier to accommodate than cars currently parking in the street or on parking lots, who will require capacity at paid public charging points, which are more likely to be used at daytime. The grid has plenty of capacity for off-peak charging (e.g. overnight), but if a lot of people want to do their charging at superchargers or other fast charging points, this could require an upgrade in generating and transmission capacity to cover a higher daytime peak load. Vehicle to grid technology would help to make this more manageable, as cars sitting idle in a driveway could provide spare power for the few cars doing the odd long distance trip.

In any case, I see a date roughly around 2030 as the Goldilocks target for a phase-out of ICE-powered new cars. For high income countries this goal is neither too unambitious nor too unrealistically aggressive. Japan’s goal by contrast for a phase-out by the mid-2030s that still allows hybrid ICEs like the Toyota Prius after that date is quite unambitious. By setting the bar that low, prime minister Suga pleases Toyota, as expected, allowing it to keep selling dated technology in Japan that they will no longer be able to sell elsewhere. That puts Japan in the company of developing countries, which will most likely continue using ICE cars exported from rich countries for years to come.

The sooner rich countries switch to BEVs, the shorter the long tail of CO2-emitting ICE cars still running in poorer countries will be.

Releasing Tritium-tainted Water from Fukushima 1

The Japanese government has approved a plan by Tepco to release more than a million tons of water stored in tanks at the site of the Fukushima 1 nuclear power station. The water is supposed to be gradually released into the ocean starting two years from now.

Currently about 1.2 million t of contaminated water are stored on site, an amount that is increasing by about 170 t per day. Tepco is expected to run out of space at the end of 2022. Water is being injected into severely damaged reactors on the site to cool the remains of nuclear fuel left inside. It leaks back out, mingles with ground water that seeps in and is then purified through a filtration system called ALPS. This removes most of the radioactive contamination, but leaves tritium, a radioactive isotope of hydrogen which can not be chemically removed from water. So it ends up in the storage tanks.

Proponents of the release argue that tritium poses little hazard in small quantities. Radiation from tritium is so weak, it only travels for a couple of mm through air and it is stopped by the dead cells on the outside of human skin. Even if ingested it does not accumulate in the human body.

The water released will be diluted to levels so low it would meet drinking water standards in Japan and in other countries. Opponents fear an economic backlash against local fisheries or argue in principle that Japan has no right to contaminate the Pacific ocean, which is not just their territorial waters but shared by many other countries.

Proponents call such criticism hypocritical, given that many other countries, including the Republic of Korea, routinely release tritium into the ocean from their own nuclear facilities.

The issue is complicated. First of all, whether the danger from the water release is real or exaggerated, fishermen will suffer economically because consumers will end up avoiding fish from Fukushima more than they already do, even if it was safe to eat. If the release is unavoidable, the fishermen should receive compensation for their economic losses. That is only fair.

The truth about the water is not black or white. The 1.2 million t of water that has accumulated over the past decade was treated in different ways at different times. Some may indeed contain only those low levels of tritium as a contaminant, but other tanks will hold water that still has significant amounts of caesium, strontium and other dangerous isotopes that unlike tritium can accumulate in organisms and pose long term hazards. More purification and testing will definitely be needed before a release can take place. As Motoko Rich and Makiko Inoue reported for the New York Times in 2019:

Until last year, Tepco indicated that with the vast majority of the water, all but one type of radioactive material — tritium, an isotope of hydrogen that experts say poses a relatively low risk to human health — had been removed to levels deemed safe for discharge under Japanese government standards.

But last summer, the power company acknowledged that only about a fifth of the stored water had been effectively treated.

Last month, the Ministry of Economy, Trade and Industry briefed reporters and diplomats about the water stored in Fukushima. More than three-quarters of it, the ministry said, still contains radioactive material other than tritium — and at higher levels than the government considers safe for human health.

The authorities say that in the early years of processing the deluge of water flowing through the reactors, Tepco did not change filters in the decontamination system frequently enough. The company said it would re-treat the water to filter out the bulk of the nuclear particles, making it safe to release into the ocean.
(New York Times, 2019-12-23)

Long term there is no real alternative to releasing the water. Once its radioactivity has been reduced to only tritium, dilution and disposal at sea should pose little risk.

The challenge however is that Tepco and the government have a public trust problem, at home and abroad. How do we know the water released will be as clean as claimed?

Any release process needs to be transparent and independently verified to make sure there are no shortcuts or other shenanigans.

See also:

My team “Maillot 24Tokyo” ride of AR Nihonbashi Flèche 2021

I survived my second Flèche ride from Toyohashi in Aichi prefecture back to Tokyo (on Strava) and my third Flèche overall.



Although we officially did not finish again, I rode 401 km altogether from Saturday morning to Sunday afternoon, including the entire 368 km route as planned, just not within the set hours. A Flèche is a randonneuring event where teams of 3 to 5 machines (tandems only count once) ride at least 360 km in 24 hours towards a central location / meeting point. At least 25 km have to be covered after hour 22 of the 24 hour ride. It was organised by AR Nihonbashi.

We used almost the same course again as last year, only the part close to Tokyo was different. The biggest difference overall was that it didn’t rain all day on Saturday as it had last year. Therefore I rode the whole day in shorts instead of in rain gear and the temperature was much more pleasant too.

To get to the start, I drove to Aichi by car the day before (I can’t rinko my Elephant Bikes NFE). I was joined by my wife and my son. Together we visited Cape Irago (Iragomisaki) on the Atsumi peninsula of southern Aichi. After dropping me off they drove back to Tokyo. The peninsula is beautiful. I was impressed by the natural forests that are a sprinkle of different colors, unlike around Tokyo where much of the current forests are regrown mono-cultures planted after post war clearcutting.

I had dinner with two other team members, then went to bed at 21:00.

The alarm went off at 05:15 and we assembled at 06:00 to get the bikes ready.

It was a 20 minute ride to the official start at a 7-11 on the outskirts, where we set off at 07:00. We head a very pleasant tailwind on our ride through farm country out to Iragomisaki, where we uploaded a group picture in front of a road sign to prove passage.

The view from the road next to the Irako View Hotel (伊良湖ビューホテル) was breathtaking. You could see the coast of Mie prefecture on the other side of the entrance to Ise Bay and various islands in the sea. I took in the view but we didn’t stop for a picture. Here’s a picture from Wikipedia (By Bariston – Own work, CC BY-SA 4.0):

We headed into the headwind that would be blowing in our faces for the next 120 km. Sometimes we took turns leading the ride. Many of the farmhouses had a storehouse between it and the coastal side, probably to block the wind.

There were also many greenhouses. Regardless of shape and size, glass or plastic they all seemed to have fuel oil tanks with the JA logo (Japan Agricultural Cooperatives), so it’s a safe bet that JA sells most of the fuel oil consumed to help grow crops in the cold season. Lots of signs advertising melons which are currently out of season but we came across many kei trucks loaded with cabbages.

There were many wind turbines in Aichi and also Shizuoka, as well as many photovoltaic installations. Their ubiquity there highlighted for me how few of them we have in Tokyo and Kanagawa. Perhaps Chubu Power is easier to deal with for feed ins than Tepco is, especially for wind power.

At noon we stopped for lunch at a ramen and gyoza place about halfway between Cape Irago and Omaezaki.

As we passed the former Hamaoka nuclear power station (it is permanently shut down) we were passed by a group of three cyclists on mamachari. Actually, one was a hybrid bike with flat bars while the other two were bona-fide mamachari. It was team ”マチャリはロング向き!” (“Mamachari is suitable for long rides!”) running in the AR Nihonbashi event and they were steaming ahead of us.

We got to Omaezaki a little after 16:00. By then it was a Century ride (160.9 km / 100 mi), but not even half of what we had set out to do.

As the course turned north here, the headwind ceased and became more of a tailwind again. It got dark near Shizuoka City.

I had felt a bit sleepy after lunch but then felt OK again. Over the next couple of hours others became sleepy as we were riding through the dark and it became more and more of a problem.

I wasn’t able to see Mt Fuji on the drive on Tomei expressway on Friday because of low clouds and now I couldn’t see it because it was night time. After crossing Fuji city and Numazu we started our climb in Izu towards Atami toge. When we finally got to the top, we had to take another power nap break at the tunnel entrance. We put on all our extra clothes for the steep descent down to Atami (13 percent). After that my rear disk brake, which recently had been very noisy and not very effective (maybe due to oil contamination from the chain) has been working perfectly again, as the heat and wear effectively decontaminated it.

Dawn approached as we headed from Atami to Yugawara and Manazuru.

We had burnt up most of our time buffer for the sleep break planned at the 22 hour stop by then, but the sleepiness in the team only got worse. So after another long break at Manazuru we sent in our DNF-notification to the event organiser. We headed to Odawara and had breakfast at the station.

After that, my friends rinko’ed their bikes for the train home while I continued on the planned route to Yamato, then another 26 km to my home. I also needed a few naps to get me home safely.

With this ride, I now have 104 contiguous months of Century a Month.

I may join a 400 km brevet later this spring and a 200 km brevet or two again after the summer.

As for the Flèche that we DNF’ed twice now, let’s see what we can come up with next year. We may just try it again a third time 🙂