The Power and the Money

Lessons from day care for President Trump

Bloomberg reports that President Trump is moving towards making good on at least one of his campaign promises: making a profit off American overseas deployments. 

The story is that he plans to introduce “Cost plus 50” as the standard: i.e., basing costs plus 50%. Now, this won’t earn the United States a whole lot of money. Basing costs just aren’t very high. The U.S. could conceivably gross about $15 billion from our allies, including a $5 billion markup. We would net less than $15 billion, however, considering that Japan currently kicks in about $1.8 billion and South Korea and Germany add another $800 million each. (South Korea just agreed to up that last to $1 billion.) But hey, $11.4 billion is still $11.4 billion. 

The thing is, the administration also plans to reduce the charge for countries that align their foreign policies with the United States. In other words, the plan would give governments concrete incentives to follow American foreign policy.

Before you get too excised about that, note that the United States already does exactly this. In 1983, the U.S. passed a law requiring the State Department to report how closely countries voted with the U.S. in the U.N. General Assembly. Since then, study after study has turned up the same pattern: countries that vote with the U.S. get either more U.S. financial support or more support from U.S.-supported institutions, like the IMF. 

Wait, the IMF is a tool of American foreign policy? Who would have imagined? Besides everybody, that is. (Specifically: Barro and Lee (2005) found that IMF loans are associated with similarity to U.S. voting patterns in the UN and economic ties with the United States. Stone (2008) and Eichengreen et al. (2006) found that countries that receive substantial amounts of U.S. foreign aid are more likely to be eligible for IMF financing.)

And of course the U.S. trades favors in even subtler ways. Germany and France support the U.S. at times when they otherwise wouldn’t because they know that the U.S. will support them even especially if the Russians invade. 

So who cares if the Trump administration turns subtext into text? Or even supertext, a sort of neon-green megawatt Bat Signal of our intentions, with a specific price tag attached? Heck, economists always say that certainty is good and prices are good. Trump will certainly bring more of both. Maybe we will get a big bang for our $11.4 billion bucks. 

Or maybe not. Sometimes, you see, it is better for quos and quids to remain implicit.

Consider what happened when an Israeli day-care chain introduced late charges. The number of late parents went up. By a factor of two. And if you are wondering, this was a controlled study, in which only some centers introduced the charges. 

The lesson? Before the centers introduced fines parents tried hard to arrive on time because it was the right thing to do. After the fine was introduced, arriving late became simply another commodity that you could buy, even if that price was called a “fine.” Be late, pay the fine, cleanse your conscience, free of any worries that other parents will disapprove. 

In other words, right not countries cooperate with the U.S. in part because they consider it the right thing to do. The U.S. might reward them, but only as part of an unspoken system of mutual favors between friends.

But in a Trumpian world, well, now U.S. support becomes just another commodity with a price tag on it. If the cost is too high, then a foreign government may as well just pay the price. Keep U.S. bases on your soil but follow whatever bloody foreign policy suits you, plus a small check to Washington.

In short, for an additional $11.4 billion per year (or less) the Trump administration will likely get a whole lot less cooperation while preserving most of our overseas commitments. Nice job, I suppose.

Of course, we would bet that this will not ultimately happen, the same way we predicted that NAFTA would not be abolished.

Reposted without comment from March 2016: How much could President Trump win from our s̶a̶t̶r̶a̶p̶i̶e̶s̶ allies?

 So, Donald Trump wants to turn America’s hegemony into a protection racket. Okay, then. How would those negotiations go?

First thing to do is estimate how much the U.S. could save by pulling out of Germany, Japan, and South Korea.

First take: billions and billions! The link goes to a 2009 study by the Commonwealth Institute. They get a cost of $127.1 billion for fiscal 2010: $132.1 billion in cost minus tribute host payments of $5 billion. (Page 9.) But that estimate is dumb. It simply pro-rates the defense budget by the number of personnel stationed overseas. Now, maybe some president might save all that money by simply downsizing the military. (Well, the Army would be “downsized” in that scenario; the Navy would be gutted.) 

Since Trump has made it clear that he has no intention of shrinking the military, then those personnel will simply be redeployed back to the United States. 

Second take: In 2013, the U.S. Senate took a look at basing costs. It estimated costs around $10 billion, 70% in Germany, Japan, and South Korea. (Page 1.) These costs do not include personnel costs or overhead: rather, they are the direct costs of keeping up the bases in those countries. German bases cost $4 billion. South Korean bases cost $1.9 billion. (Page 18.) Japanese bases cost $2.5 billion. (Page 40.)

But South Korea and Japan, unlike Germany, pay pretty large portions of the U.S. cost! (Hold that thought on Germany; this is based entirely on the Senate report.) In South Korea, U.S. taxpayers foot only $1.1 billion of the cost; South Koreans make up for $800 million. (Page 21.) Japan, meanwhile, covers more than 70% of U.S. costs, to the tune of $1.8 billion per year. Until 2010, in fact, Japan covered all U.S. basing costs. (Page 41.) So President Trump might wrap up the American empire-by-invitation in Asia for savings of ... $1.8 billion. Get out of Germany at the same time and you save $5.8 billion per year. 

Only you don’t! Consider the cost of transferring 7,000 Marines from Okinawa to Guam: $19.3 billion. (Page 49.) Scale that up by the other 43,000 American forces in Japan and you get $119 billion to withdraw without demobilization. Add in the 29,000 forces in South Korea and you get $197 billion. Throw in the 36,000 troops planned to still be in Germany in 2017 and you get $296 billion. 

Now, that is a one-time cost, not an annual flow. And it is likely exaggerated, since most of the Naval and Air Force personnel overseas will not be as expensive to relocate as the Marines. So lets arbitrarily cut the cost by 60%. 30-year treasury bonds are currently yielding around 2.5%. Call it an annual expense of $5.8 billion between 2017 and 2047. 

$5.8 billion ≈ $5.8 billion. No savings. With interest rates at a near all-time low and shaving 60% off the topline estimate. Huh!

And now lets return to the German issue. Germany pays a share of American construction costs for our bases, but those costs are lumpy. In  2009, Rand estimated total German support to total $831 million. (Page 154.) If that number is accurate for other years, then the U.S. pays only $5.0 billion for our three main overseas presences. And $5.8 billion > $5.0 billion.

Third take: This big Rand report. It is long and complicated and does not cover the complete withdrawal option. But their minimum posture saves around $2.9 billion per year ... assuming that foreign contributions remained unchanged.

In other words, the American defense perimeter does not cost all that much. 

Now we need to figure out what these cost estimates imply for the negotiations.

President Trump, in this world, cannot simply threaten to walk away, because that would cost him. He needs to make one of two additional threats:

1. He is willing to incur costs to the United States if negotiations break down. That is, he tells his counterparts in Berlin, Tokyo and Seoul that he is willing to pay money to bug out unless our satraps allies fork over even more money. Good luck with that one.

2. He is willing to bug out and downsize the American military budget. Now those cuts do not have to be large for us to come out ahead — we are talking maybe 1% of the total defense budget — but they would need to be made. 

So, yeah, I suppose a cost-cutting President Trump could threaten to cut American military spending unless those foreigners pay us more. But unless he is willing to resort to an outright protection racket, he would net us 1% of the military budget on the threat of making a 1% cut in the military budget. 

Yeah, that makes sense. Totally worth it. After all, it isn’t like you could just, you know, cut the military budget without threatening our allies ... 

Two of these places are not like the others

 So the intrepid data team at the Economist collated all internet mentions of the term “51st state” and came up with the following map:

Don’t ask me to explain the references to existing states. But for most of the foreign countries, I get it. Australia, Britain, Japan, very close treaty allies. Canada and Mexico, in NAFTA. (Canada’s also a very close treaty ally.) Taiwan, special relationship. Caribbean, Philippines, Ireland, lots of emigrants to the U.S. (Bora Bora and Brunei, that’s just weird, but whatever.) 

In fact, I’m surprised that more Latin American countries didn’t crack the top ten: Colombia, El Salvador and Panama have a right to feel slighted. 

But what are Russia and North Korea doing on this list?? Who refers to them as America’s 51st state? And why?

How much could President Trump win from our s̶a̶t̶r̶a̶p̶i̶e̶s̶ allies?

 So, Donald Trump wants to turn America’s hegemony into a protection racket. Okay, then. How would those negotiations go?

First thing to do is estimate how much the U.S. could save by pulling out of Germany, Japan, and South Korea.

First take: billions and billions! The link goes to a 2009 study by the Commonwealth Institute. They get a cost of $127.1 billion for fiscal 2010: $132.1 billion in cost minus tribute host payments of $5 billion. (Page 9.) But that estimate is dumb. It simply pro-rates the defense budget by the number of personnel stationed overseas. Now, maybe some president might save all that money by simply downsizing the military. (Well, the Army would be “downsized” in that scenario; the Navy would be gutted.) 

Since Trump has made it clear that he has no intention of shrinking the military, then those personnel will simply be redeployed back to the United States. 

Second take: In 2013, the U.S. Senate took a look at basing costs. It estimated costs around $10 billion, 70% in Germany, Japan, and South Korea. (Page 1.) These costs do not include personnel costs or overhead: rather, they are the direct costs of keeping up the bases in those countries. German bases cost $4 billion. South Korean bases cost $1.9 billion. (Page 18.) Japanese bases cost $2.5 billion. (Page 40.)

But South Korea and Japan, unlike Germnay, pay pretty large portions of the U.S. cost! (Hold that thought on Germany; this is based entirely on the Senate report.) In South Korea, U.S. taxpayers foot only $1.1 billion of the cost; South Koreans make up for $800 million. (Page 21.) Japan, meanwhile, covers more than 70% of U.S. costs, to the tune of $1.8 billion per year. Until 2010, in fact, Japan covered all U.S. basing costs. (Page 41.) So President Trump might wrap up the American empire-by-invitation in Asia for savings of ... $1.8 billion. Get out of Germany at the same time and you save $5.8 billion per year. 

Only you don’t! Consider the cost of transferring 7,000 Marines from Okinawa to Guam: $19.3 billion. (Page 49.) Scale that up by the other 43,000 American forces in Japan and you get $119 billion to withdraw without demobilization. Add in the 29,000 forces in South Korea and you get $197 billion. Throw in the 36,000 troops planned to still be in Germany in 2017 and you get $296 billion. 

Now, that is a one-time cost, not an annual flow. And it is likely exaggerated, since most of the Naval and Air Force personnel overseas will not be as expensive to relocate as the Marines. So lets arbitrarily cut the cost by 60%. 30-year treasury bonds are currently yielding around 2.5%. Call it an annual expense of $5.8 billion between 2017 and 2047. 

$5.8 billion ≈ $5.8 billion. No savings. With interest rates at a near all-time low and shaving 60% off the topline estimate. Huh!

And now lets return to the German issue. Germany pays a share of American construction costs for our bases, but those costs are lumpy. In  2009, Rand estimated total German support to total $831 million. (Page 154.) If that number is accurate for other years, then the U.S. pays only $5.0 billion for our three main overseas presences. And $5.8 billion > $5.0 billion.

Third take: This big Rand report. It is long and complicated and does not cover the complete withdrawal option. But their minimum posture saves around $2.9 billion per year ... assuming that foreign contributions remained unchanged.

In other words, the American defense perimeter does not cost all that much. 

Now we need to figure out what these cost estimates imply for the negotiations.

President Trump, in this world, cannot simply threaten to walk away, because that would cost him. He needs to make one of two additional threats:

1. He is willing to incur costs to the United States if negotiations break down. That is, he tells his counterparts in Berlin, Tokyo and Seoul that he is willing to pay money to bug out unless our satraps allies fork over even more money. Good luck with that one.

2. He is willing to bug out and downsize the American military budget. Now those cuts do not have to be large for us to come out ahead — we are talking maybe 1% of the total defense budget — but they would need to be made. 

So, yeah, I suppose a cost-cutting President Trump could threaten to cut American military spending unless those foreigners pay us more. But unless he is willing to resort to an outright protection racket, he would net us 1% of the military budget on the threat of making a 1% cut in the military budget. 

Yeah, that makes sense. Totally worth it. After all, it isn’t like you could just, you know, cut the military budget without threatening our allies ... 

Methane hydrates are the fuel of the future and will be for a long time

 The Financial Times had a piece last year about Japanese attempts to develop methane hydrates. I am not convinced that these will ever be commercialized.

Vox fails World War 2 history

 Vox is a very useful website. I admire what they are trying to do. But sometimes they just completely screw things up. This time, they did so in their grandiosely-named “40 maps that explain World War II.”

The big problems start with Map 10: “The amphibious invasion of the United Kingdom that never happened.” All you need to know about the problem is in the last line of the caption: “As the British advantage in the air grew, Hitler was forced to shelve his invasion plans.”

Christ almighty, no. No. No no no. Nazi Germany had no way of successfully invading the British isles regardless of the outcome of the Battle of Britain. First, the Royal Navy was in the way, not just the Royal Air Force. Second, on four occasions the British military waved away the entire Royal Navy and Royal Air Force in order to game out an invasion ... and in all four exercises the invasion collapsed with days. Click the link for Alison Brooks’ classic essay.

After that, the other problems are minor. Map 13 is about Operation Barbarossa. It is true that the Germans failed to take Moscow before winter set in, which was a serious problem ... but the order of the sentences in the caption implies that the Germans failed to take Moscow because winter set in, which is not what happened: 

The Nazis were able to drive hundreds of miles east and reach the outskirts of Moscow by October. But then Stalin was saved by the bitterly cold winter. The Soviets had more experience operating in cold weather and were better prepared than the Nazis. German equipment was not designed for below-zero temperatures, German soldiers were under-dressed, and they lacked essentials such as antifreeze. Germany never took Moscow.

And while Map 1 points out the Allied economic advantage (meaning really the U.S., U.K., and USSR) the point gets lost in all the later discussions. Here are two examples. First, Map 19 recounts Free French victories in Africa and provides a much-needed antidote to the image of French fecklessness. The problem is that the caption says that the victories “deprived Hitler of much-needed resources,” which is not true. The Axis did not draft African soldiers and the  area did not provide any strategic raw materials. (Frex, Gabonese oil production did not start until 1956.)  

Second, Map 28 describes the Battle of Midway as a turning point, which it was in the sense of beginning the U.S. offensive ... but was not in terms of being key to a U.S. victory. By the summer of 1943 the U.S. had a wave of new carriers. Japanese-occupied Midway would be right on the doorstep of the main U.S. naval base. It would be thousands of miles away from Japan’s. The U.S. would be able to bomb it at will (as we did Wake Island) and rather rapidly starve out the survivors. The Marianas campaign might be pushed back to 1945, but by then the U.S. would have overwhelming superiority. And then the atomic bombs are ready by August.

A later Marianas campaign might give the Soviets the time and inclination to occupy all of Korea instead of just the northern half ... which would make the Battle of Midway the turning point of a war that hadn’t even started. You could call the South Korea the big winner of that battle. 

But honestly, the implication that Operation Sealion could have worked is the real error. Not to take anything away from the Battle of Britain, which accomplished much, but keeping the Nazis out of London was not one of them.

Inexplicable Japanese scandal

 Germany has a lot of inexplicable political scandals, simply because standards are so high. Japan, not so much. 

There is one recent Japanese scandal which does on its face indeed seem scandalous, even by the lax standards of this glorious hemisphere ... but which upon further investigation seems to be the opposite of a scandal.

Yoichi Miyazawa, the trade minister, got in some trouble last week because his people spent public money to frequent an S&M club. Meh. The amount was only $170. Of course it should not have been expensed, but neither should $170 in tickets for the Equalizer. 

But that’s not the non-scandal. The non-scandal is that Minister Miyazawa turns out to own 600 shares in Tepco. Tepco, you might recall, is the operator of the Fukushima Daiichi nuclear power plant. And Miyazawa’s ministry would  ultimately be responsible for regulating Tepco. A clear scandal!

Well, maybe. Miyazawa owned those shares before the 2011 tsunami. 

In other words, after taking an 80% loss the Minister Miyazawa has held on to some flatlining shares. It does look bad, I will admit, and he should sell them. But a serious scandal this really is not. Or am I missing something?

Fukushima keeps getting more expensive

Tepco just announced that they are going to construct two 500-MW coal plants at Fukushima at an eye-popping capital cost of $2.96 billion.

This adds to the cost of the disaster in two ways. First, coal! Carbon dioxide. Ugh. The plants are supposed to be integrated gasification combined-cycle (IGCC) stations, which are more efficient ... but it is still coal. (Wikipedia has a nice summary of the technology.)

Second, actual cost. At $3.0 million per megawatt, the plants will be expensive. Japan finished three nuclear power plants in 2004-06. Their capital cost came to $2.0, $2.8, and $2.3 million per megawatt respectively ... rather less than the proposed cost of the coal plants. (Figures fromTable 3 in Yangbo Du and John Parsons, 2009, “Update on the Cost of Nuclear Power,” Center for Energy and Environmental Policy Research WP-09-004.)

Moreover, coal is rather more expensive than uranium. The average landed price in Japan was $106.28 per ton in August 2013. Even if we assume a 50% heat rate (the average for the Japanese coal fleet is 45%) that would translate into a fuel cost of $30.61 per megawatt-hour, against $12 for a nuclear plant.

Tepco says that it uses a 3% discount rate to evaluate projects, which seems rather low (even for Japan) but let’s use it anyway. (It won’t make a difference, since the straight-up capital costs for coal are the same.) Let’s also assume that both a coal and a nuclear plant will last only 25 years. (Nuclear plants will make it to 60, and Japan’s BWR designs are longer-lived than the PWRs common elsewhere.) Add in a cost of $1.2 million per MW for a CCGT plant (they are more expensive in Japan) and a natural gas cost of $16 per MMBTU and you get the following levelized cost of electricity per megawatt-hour:

$52 is a lot higher than the $37 it would cost to get the electricity from nuclear! And that doesn’t take into account that the cost is even lower from Japan’s existing reactor fleet, which is mostly depreciated.

And so, the cost of Fukushima keeps growing, as one of the few countries where nuclear power is economically competitive turns to burning more expensive coal. I find it sad.

China claims Okinawa!

OK, not really. But this story in the New York Times was a bit jaw-dropping.

Me, I’m still ambivalent about the whole giving-it-back-to-Japan thing. It would have made a great state. (The honorable Senator Keiko Itokazu, one of the few Democrats from this usually Republican state and vital legislative ally of her fellow Mormon, Harry Reid. C’mon, you can see it.) At the very least, a compact of free association with an Okinawa Republic. But it was not meant to be.

So I can’t call the Chinese who want the Ryukyus to become an independent state under Chinese suzerainty completely crazy. (After all, an Okinawa Republic with a CFA is nothing more than “an independent state under American suzerainty.”) And it would have the additional benefit of pushing the American strategic perimeter away from China.

But I find it hard to take random non-official Chinese claims seriously enough to warrant a page 4 story in the what is still America’s paper of record. Slow news day? Or am I missing something about the seriousness of what looks from the outside to be random noise?

Brazil imposes a tax on atomic energy to promote atomic energy

Brazil is about to place a 10% tax on the gross revenue generated by the sale of nuclear electricity.

This is a strange idea. As a revenue-raising device, it is improbable. Nuclear energy is not cheap: there are no ex ante rents to seize. Post hoc, it is possible to grab value from the financiers — capital expenditures are what drive up nuclear’s costs — but that would serve only to disincentivize reactor construction.

As a disincentive, however, a 10% revenue tax is a lousy one. First, most of the costs can be passed on to consumers: electricity demand is not very elastic. Second, regulators can prevent those costs from being passed on, but regulators can also stop nuclear power plants from being built. Heck, almost any policy uncertainty can stop a nuclear power plant from being built. In short, using a tax to disincentivize reactors is either ineffective or unnecessary.

So, why?

The key feature to note is that the revenue from the tax will go to local governments, not the federal authorities. It is actually a way to build local support for reactors.With the tax, localities will have some skin in the game.

In fact, that was how they did things in Japan.

In 1974, Japan created a special purpose tax on electricity. The government funneled most of the revenue into Electric Power Sites Account, which went to local governments that allowed power plants to be built in their jurisdictions. Most of those plants, not surprisingly, turned out to be nuclear. Between 1997 and 2007, 68.4% of Electric Power Sites Account grants went to nuclear power plants, against 27.3% for thermal power and 3.9% for hydroelectric.

To give an example, seven Tepco-owned nuclear re­ac­tors in Kashiwazaki City in Niigata Prefecture in western Honshu generated cumulative amounting to ¥113.3 billion from 1978 to 2009; the combination of central grants and sub­sidies (plus direct taxes on Tepco)], accounted for 34.5% of Kashiwazaki City’s revenue.

The Japanese designed the system to reduce local opposition to new plants. The government front-loaded the subsidies so that most of them went to the local government between the start of construction and five years after beginning operation. That gave localities strong incentives to approve new plants in order to prevent a decline in public revenue.

By 2011, parts of the Japanese business press were calling the subsidy system a “drug” for local governments and plant neighbors. Of course, that was the intent! It took the Fukushima disaster to change public opinion.

It remains to be seen if Brazil will be as successful in creating a constituency for nuclear power. There are a lot of differences between the two nations, not least of which that Brazil plans five new reactors over the next decade (at best!) while Japan rolled out 35 over the ten years following the 1974 law. But it could work. As long as regulators let utilities pass on the cost of the tax — not a certainty in Brazil — then a tax on nuclear power designed to promote nuclear will be an ingenious step in the never-ending battle against NIMBY.

Japan’s methane hydrates should not be developed, and likely won’t be, huzzah

I am in Panama! It is an awesome place. (Including the traffic at 5pm on a Monday. And the lack of sidewalks.) So, of course, I will talk about Japan.

A Japanese drilling ship, the Chikyu, recently extracted methane from hydrates 1,000 feet down in the Eastern Nankai Trough. Methane hydrates, aka clathrates, are deposits of natural gas trapped in ice. (OK, it’s not really ice, but what else would you call it?) They are not easy to extract. A good primer is here at Wonkblog.

How not easy to extract? Seriously not easy to extract. This blog likes cost numbers, and fortunately the Japanese consortium running the process provided them in 2008:

	¥ per m3	$ per MCF
Present construction costs	¥ 92	$ 33.98
Assumed conditions	¥ 46	$ 16.99
Volume 25% lower than expected	¥ 174	$ 64.27

“Assumed conditions” means a threefold drop in construction costs. Worse yet, costs are extremely sensitive to volumes; produce less than expected, and profits go out the window. The low price of $16.99, meanwhile, is more than Japan currently pays for liquified natural gas (LNG).

Still, breakthroughs happen. Maybe Japanese engineers can move beyond proof-of-concept tests to cost-effective methane extraction. What then?

Well, do not worry about the amount of methane that can be extracted and burned. That is entirely a red herring. The world has plenty of cheap conventional natural gas to burn between now and 2100. Unless the hydrates drop in price way more than anyone expects — say, down below $6 per MCF — then commercial exploitation will not cause the world to burn natural gas any faster than it would have otherwise. I very much fear global warming, and once natural gas has killed coal dead then natural gas will become the problem, but the hydrates won’t make a damned bit of difference unless you believe that we will be burning methane for power into the 22nd and 23rd centuries.

OK, so no worries in burning methane from hydrates. (More specifically: no additional worries than we currently have.) But do worry! I have been talking about this with my friend Will Baird out in California. The problem is that not all of that methane will be burned. Maybe not even most. Maybe even catastrophically little.

The Department of Energy reports that the Nankai Trough has 16 to 27 trillion cubic meters of methane hydrates. Methane is approximately 21 times more potent as a green house gas as CO2. If you were to dump all of the Nakai Trough's hydrates into the air at once, it would the equivalent 20 years worth of U.S. methane emissions all at once.

The good thing is that methane falls out of the atmosphere quickly. The bad thing is in the middle sentence of the above paragraph. A catastrophic emission of the Nankai hydrates would cause a short painful spike in temperatures ... and given how close the Earth is to climatological tipping points, it could push us over.

That probably won’t happen, but you’d be crazy not to think about it. What most likely would happen is that commercial exploitation would destabilize the deposits. The Japanese want to remove ice from under sediment. In a geologically inactive zone, that is mildly annoying, but, well, Japan is not in a geologically inactive zone. Even without the effect of quakes, if the Japanese use warm(er) water to melt and separate the gas from the ice, then surrounding ice could be destabilized. A catastrophic release may be unlikely, but the probability of large-scale methane releases is too high for comfort.

And some degree of methane release is inevitable. So far, we do not know what the capture ratio will be ... and we probably won’t until there is a commercial test.

But let me finish with two pieces of good news. First, the process is not likely to be economically competitive; we will probably therefore never have to worry about the environmental effects. Second, this is a government-funded project, so even if it is competitive there will be fewer vested interests insisting on development.

In short, the hydrates are scary if you want to leave this world to your grandchildren ... but the obstacles to development are high enough that they will likely not be developed.

Can LNG solve Japan’s (or at least Tepco’s) energy troubles?

What are the implications of Japan’s infrastructure on the future? Let’s start with LNG. Can Tepco use LNG to replace its nuclear fleet?

Tepco certainly can build CCGT. It is true that CCGT plants can be built in 27 months. It is also true that CCGT plants are not expensive. (My students estimated a capital cost of $1.25 billion per GW.) And it is true that they do not need that much fuel: a gigawatt CCGT plant requires only 46 BCF of fuel per year — the additional 782 BCF needed to replace Tepco’s nuclear plants is well within the capacity of its existing import terminals.

The rub is that Japan has little ability to move natural gas within the country, so the new plants would have to be built near the terminals. The region is not exactly unpopulated. (See map.) Siting and building plants on the requisite scale is likely to take a lot longer than 27 months unless the country builds more remote terminals or a better high-pressure distribution network ... which will take a lot longer than 27 months. Catch-22.

The second problem is that LNG is an expensive fuel. In February, LNG in Japan went for about $16.66 per MMBTU. One of the reasons LNG is so expensive it that sellers prefer contracts linked to the price of oil. Japanese consumers, of course, would prefer contracts linked to the American wholesale price. The problem, of course, is that high transport costs segment the world market. An exporter in Qatar or Australia has little reason to accept a contract linked to the North American price.

But how much cheaper is North American gas? Cheniere Energy contracts price LNG (at the liquification plant) at Henry Hub × 1.15 + $3.00, or $7.19 at a Henry Hub price of $3.64. It costs maybe another $3 to ship it to Asia. (This report is invaluable.) Call it $10.20. Now, this blog has argued that natural gas prices are going to rise. Not immediately, but in the next few years. If they go back up to $5.50, then the price to Japan will be about $12.33. (The Development Bank of Japan believes prices in this range are possible if import prices are linked to Henry Hub.) That is less than $16.66, but hardly the millenium.

According to my students, CCGT in Japan is only marginally competitive when LNG goes for of $17 per MMBTU: 7.5¢ per kWh, against a nuclear cost in the 6¢-7¢ range. Rates therefore would have to rise. If the cost goes under $13, however, then CCGT electricity would go below 7¢. In short, replacing nuclear power with gas without price hikes depends on access to North American gas.

Thus, on February 22, 2013, Shinzo Abe asked Barack Obama to approve LNG exports to Japan. Tepco has already signed an agreement to purchase gas from the Cameron project, but it still needs U.S. export approval and might not happen. (If the Europeans are smart, they will make free trade in natural gas a plank of the U.S.-E.U. FTA.) A less controversal project is going ahead in Kitimat, British Columbia (this would be better for America, since the B.C. deposits are new and hard to move south and thus have no effect on U.S. prices) but Chevron is now a partner and Chevron insists that prices be linked to oil prices.

But ... you knew there was one ... even if gas exports happen, Cameron won’t be ready until 2018. Kitimat was supposed to open in 2017, but now will be later. In short, Japan will be at the mercy of oil-price-linked contracts for some time ... and even after that will face a substantial import bill.

Which is why there is so much interest in the methane hydrates. Japan has been the first to demonstrate that it is possible to extract natural gas from undersea hydrates. Game changer, they say.

Or not. In a future post, I will say why I do not think that methane hydrates will be developed soon. Or at least why I hope not, since they pose a catastrophic risk to the climate.

The frequency frontier in Japan

After the Fukushima nuclear disaster, the utility company Tepco faced a problem: how to keep the lights on with the reactors off? Four of my students (Dan Bodley, Andrew Boudreau, Allison Dee, and Nick Phoutrides) found the answer. It boils down to “they didn’t.” But why they didn’t turns out to be very interesting. Preview: it was not because the nation of Japan lacked the excess capacity to replace nuclear electricity.

Japan is divided into 10 different utility areas.Tepco enjoyed 65 GW of capacity (17 from nuclear) and imported up to 13 GW. Typical summer peak demand ran around 61 GW. The quake took 17 GW of nuclear off-line, leaving 48 GW of local capacity under the heroic assumption that the utility could run all its non-nuclear plants at peak capacity.

OK ... so just import! Problem solved, no blackouts, all good. And if you just looked at the headline numbers, that would have been feasible, even with the national reactor shut-down.

Except there were two problems. First, the quake bady hit Tohoku — which normally provided all of Tepco’s imports — even worse than Tepco.

The second problem was the “frequency frontier” that ran roughly along the border between the Tepco-Tohoku service areas and the Hokuriku-Chubu service areas. Back when Japan first electrified, in the 19th century, General Electric supplied the Osaka utilities whereas the precursors to Tepco bought from Allgemeine Elektricitäts-Gesellschaft. AEG ran its equipment at 50Hz, whereas G.E. used 60Hz. The end result was the “frequency frontier” that bisects the country. There were three converter stations on the frontier, but their combined capacity came to only 1.2 GW. The end result was that Tepco’s maximum available peak import capacity declined from 13GW before the quake to only 7.3GW ... and that 7.3GW number, remember, was a maximum.

In short, Tepco faced a peak shortfall of 5.4 GW. Here is how that worked out: 47.7GW non-nuclear supply + 7.3GW imports − 60.4GW peak demand = (5.4GW). Tepco filled the gap by “conservation.” Why the scare quotes? Well, we normally think of conservation as turning off the AC, keeping lights low, taking the stairs, that sort of thing. We do not think of rolling blackouts as conservation. But rolling blackouts are indeed how Tepco reduced demand. In fact, Tepco exported 1.4GW to Tohoku in August 2011.

Running the company’s thermal plants at full speed meant importing a lot of liquified natural gas (LNG): Tepco’s imports jumped from 959 BCF in 2010 to 1,127 BCF in 2011. Fortunately for Tepco, its LNG facilities had a capacity of 2,785 BCF.(Of that, Tepco wholly owned two terminals capable of producing 1,287 BCF, with another 1,377 BCF of capacity in two terminals co-owned with Tokyo Gas. Tokyo Gas is the sole owner of the fifth; the gas company imports about 560 BCF per year in total.)

2012 appears to have been little different from 2011. Demand peaked at 50.8GW; Tepco had a cushion of 3.8GW on imports peaking at 6.9GW. But demand was held to 50.8 only by dint of blackouts. LNG imports remained about the same, at 1,120 BCF.

And so there you have it. On paper, Japan could handle taking nuclear off-line almost as easily as Germany. In practice, not so much, because of decisions made a century ago about sourcing electrical equipment from Germany and the United States. Economic history, it is an awesome field.