Lessons from Great East Japan Earthquake:Fukushima Nuclear Power Crisis

Makoto Saito, Hitotsubashi University

An outline of talk about Fukushima Crisis

2013/3/19-212

What happened? A visual approach A mechanical approach

Several factors responsible for the crisis A failure in the on-going crisis management Very old facilities Poor regulation

A corporate finance aspect of the crisis Restructuring Tokyo Electric Power Company

Some economic issues Disposal of nuclear wastes Generation costs Insurance for nuclear damage

Some proposals

What happened?

2013/3/19-213

Where is Fukushima No.1 Nuclear Plant?

2013/3/19-214

Neighboring nuclear power plants: Onagawa, Fukushima No.2, and Tokai No.1

2013/3/19-215

Six reactors at Fukushima No.1

2013/3/19-216

Plant formation at Fukushima No.1

2013/3/19-217

Key events in Fukushima Crisis

2013/3/19-218

March 11, 2011 — Magnitude 9.0 earthquake (14:46) and tsunami (15:41) cripple plant, cutting off power to the entire site. Government declares nuclear emergency, directing residents in a 3-km radius of the plant to evacuate.

March 12 — Prime Minister Naoto Kan inspects the plant. Hydrogen explosion occurs at reactor 1 building. Government expands evac zone to 10-km radius.

March 14 — Reactor 3 building suffers hydrogen explosion.

March 15 — Reactor 4 building hit by hydrogen explosion from gas from reactor 3.

Damage by tsunami

2013/3/19-219

Fukushima No.1 Plant hit by tsunami

2013/3/19-2110

Before and After

2013/3/19-2111

Severely damaged Reactors 1 through 4

2013/3/19-2112

Damaged Reactor 1

2013/3/19-2113

Damaged Reactor 3

2013/3/19-2114

Damaged Reactor 4

2013/3/19-2115

Damaged Reactors 1 through 4

2013/3/19-2116

Some mechanical aspects of the crisis

2013/3/19-2117

Serious damages on the primary cooling system : Damages by the earthquake : Damages by the tsunami

18

Sakashita Dam

Possible effects of the earthquake

Reactor Building

R/BTurbine Building

T/BSeawater

Pump Building

External Power Sources

Emergency

Generator

?

?

2013/3/19-21

Fact 1: The facilities were not back-fitted according to the new seismic safety standard.

19

Fukushima Dai-ichi was designed in the 1960s, and opened in 1971, prior to the introduction of the seismic safety standard in 1981.

Those facilities had not been retrofitted according to the most recent seismic safety standard, which was revised substantially in 2006. TEPCO reported to the regulator that they would be back-fitted by

2016. It cost a lot! (80 billion yen for old Fukushima Dai-ichi)

The nearby transformer substation (Shin-Fukushima Hendensho) and the cables connected from it to Fukushima Dai-ichi were much less earthquake-proof.

The conduits and pipes from the Sakashita Dam was not so robust as it should have been.

2013/3/19-21

Fact 2: The facilities were not robust with respect to tsunami risks.

20

The seawater pumps were not either water-sealed or protected by solid buildings at all.

As TEPCO did not modify the facility arrangement designed originally by GE, all emergency diesel generators were located at the lower floors of less robust turbine buildings.

2013/3/19-21

①

Fact 3: In principle, a primary cooling system (PCS) cannot be substituted for by an emergency core cooling system (ECCS), much less by an inefficient one.

21

ECCS

Water Tank

Pool in Suppressi

on Chamber

To T/B

From T/B

A reactor is isolated once severe accidents take place.

Containment Vessel

Pressure Vessel

②

Note: ECCS relies as water sources first on water tanks, and later on pools in a suppression chamber.

2013/3/19-21

Why did ECCSs not work effectively?

22

ECCS of Unit 1 (isolation condenser) was extremely old. An isolation condenser was expected to work for only half a day even it

was handled normally. Even HPCI and RCIC (installed in Units 2 and 3) are expected to work for

a few days under a normal condition.

Due to a loss of alternating and direct current power sources, HPCI (most efficient ECCS) did not work effectively.

Plant operators were not trained well for how to handle ECCSs. There were many human errors in operating ECCSs.

What is most important, without any definite prospective that the primary cooling system would be recovered quickly, TEPCO and the regulator relied in vain on ECCS, thereby delaying their decision of ventilation and seawater-injection.

2013/3/19-21

Fact 4: Ventilation facilities were considered as unnecessary.

23

Dry Well Vent

Wet Well Vent

Supposed to be filtered, but they did not install any filter.

The water may start to boil due to extremely hot steam. Then, boiling water disables ECCS functions and the ability to condense radioactive vapor.

Used to be a rather fragile

one

Pool in Suppressi

on Chamber

2013/3/19-21

Why did they not exercise effective ventilation for Units 2 and 3 reactors?

24

GE Mark I engineers considered ventilation as unnecessary. They presumed that the water at suppression chamber pools was

good enough to condense steam and deflate the internal pressure. They did not even install solid (hardened) ventilation in their original

design.

While hardened ventilation was forced to be installed by the U.S. regulation after the Chernobyl disaster, they still considered ventilation as unnecessary. Consequently, they did not attach filters to ventilation. The easiness of operations was not considered seriously.

Accordingly, they were reluctant to release highly radioactive vapor without any filtering, while they delayed in ventilation due to rather complicated operation procedures.

2013/3/19-21

Fact 5: TEPCO and the government were reluctant to inject seawater into Units 2 and 3 reactors.

25

Portable Pump

2013/3/19-21

Why did they delay in injecting seawater into Units 2 and 3 reactors?

26

They seemed to have a dim hope that the primary cooling system would manage to be recovered. They relied in vain on ECCS with such a faint hope.

It is said that they feared that seawater injection would immediately lead to reactor decommissioning. They had had their intention to use these reactors

for another 20 years. However, it sounds a bit odd because once a

meltdown proceeds, a reactor has to be decommissioned anyway. They might not have understood the consequence of a

meltdown, or informed of its occurrence. 2013/3/19-21

Consequence: not only meltdowns (the case of the Three Mile Island nuclear accident)

ECCS

27 2013/3/19-21

But also, melt-throughs (serious damages to pressure vessels)

ECCS

28 2013/3/19-21

But also, hydrogen explosion (serious damages to reactor buildings)

29

A top lid might have moved up due to high

pressure.

Note: Many researchers failed to recognize any possibility that hydrogen explosion happens outside containment vessels. But, a few papers pointed out that the capacity of BWR’s containment vessel may be too small to survive extremely high pressure.

2013/3/19-21

But also, serious damages to containment vessels, then, we had had a tragic situation

Leaks

Leaks

Leaks

30 2013/3/19-21

Necessary to develop a large-scale mission to decommission severely damaged reactors:1. Need to build another container to contain the damaged containment vessels.2. Need to cool down the melted nuclear fuels by fresh water for at least three years.3. It might take one century to complete the entire process.

31 2013/3/19-21

A corporate finance aspect of the crisis

2013/3/19-2132

The aftermath: A window-dressing settlement of TEPCO account

33

The Nuclear Damage Liability Facilitation Fund (NDF) facilitates TEPCO to finance expenses for nuclear damage compensation (2.4 trillion yen). NDF was founded by a special measures law in August, 2011,

and backed financially by the government.

But, TEPCO’s borrowing from NDF is regarded as not a loan, but a grant. (What a generosity!) 2.4 tril. yen grant from NDF was appropriated as extraordinary

income (revenue) in the TEPCO account! With such a window-dressing treatment, the current net loss

reduces substantially from 3.2 tril. to 0.8 tril. yen, and such a loan from NDF was off TEPCO B/S.

NDF’s grant to TEPCO is supposed to be repaid jointly by TEPCO and the other power companies for the next ten years.

2013/3/19-21

Without such window-dressing, TEPCO is immediately insolvent.

34

Without such window-dressing, TEPCO is immediately insolvent at a negative capital of -1.6 tril. yen, instead of a positive capital of 0.8 tril. yen.

In addition, TEPCO’s B/S has so far appropriated only the followings: The reserves for scrapping damaged reactors: 0.8 tril.

yen It may reach10 tril. yen.

The reserves for damage compensation: 2.1 tril. yen It is likely to double with decontamination, and would triple with

class actions against TEPCO.

2013/3/19-21

The asset side of Tepco (book value, unit: trillion yen)

35

Total Assets: 15.5 Water-power plants 0.6 Thermal power plants 0.9 Nuclear power plants 0.7

Power transmission facilities 2.0 Power transformer facilities 0.8 Power distribution facilities 2.1

Long-term investments 3.7 Liquid assets 2.3

2013/3/19-21

The liability side of TEPCO (book value, unit: trillion yen)

36

Total Liabilities: 15.5 Corporate bonds 3.7 Long-term bonds 3.3

Capital 0.8 Capital without window-dressing -1.6

2013/3/19-21

Some brief comments on economic aspects of the crisis

2013/3/19-2137

Economics of disposal of nuclear wastes

38

Which way is more economical?

How long will it take to complete disposal?

Where should we dispose of nuclear waste?

2013/3/19-21

A basic knowledge about nuclear wastes

39

New nuclear fuels (per ton) Uranium 238 (not fissile) 　　 970kg Uranium 235 (fissile) 　　 30kg

Spent nuclear fuels (per ton) 1. Uranium 238 (not fissile) 　　　　　　 950kg 　 2. Uranium 235 (fissile) 　　　　　　 10kg 3. Plutonium 239 (fissile) 　　 10kg 4. Fission products (not fissile) 　　　 30kg

A current plan by the Japanese Government: Extracting Uranium and Plutonium from spent nuclear fuels, Using the reprocessed fuels for nuclear power generation again, and Disposing of fission products, which are extremely radioactive.

2013/3/19-21

Several fundamental problems

2013/3/19-2140

It costs a lot to reprocess spent nuclear fuels. Disposal without reprocessing (called a once-

through cycle) may be much more economical than disposal with reprocessing (called a fuel recycle).

It is necessary to store radioactive fission products for a long time (much longer than a few thousand years!).

It is hard to find a place to dispose of nuclear wastes deeply under the ground.

Power generation costs

41

Not including pumped storage costs in nuclear power Nuclear power 8.9 yen/kWh Thermal power 9.0 yen/kWh Water power 7.5 yen/kWh

Including pumped storage costs in nuclear power Nuclear power plus pumped storage 10.1

yen/kWh Thermal power 9.0

yen/kWh Water power minus pumped storage 3.6

yen/kWh2013/3/19-21

International comparison of mandatory insurance for nuclear damage

42

US: Under Price-Anderson Act, the private insurance up to $375 million/site the industry self-insurance (ex-post financing) up to $11.6 billion/site the combined insurance capacity up to $12.2 billion/site

Japan: private insurance up to $1.5 billion/site (120 billion yen)

UK: In 2012, required insurance coverage has been lifted from EUR 140

million/site to EUR 1.2 billion/site. Germany:

unlimited operator liability and requiring EUR 2.5 billion/site security France:

requiring EUR 91 million/site insurance security Switzerland:

planning to increase EUR 600 million/site to EUR 1.1 billion/site Finland:

unlimited operator liability and requiring EUR 300 million/site

2013/3/19-21

Some proposals

2013/3/19-2143

In sum,

44

Fukushima Dai-ichi facilities were: Not back-fitted at all according to the most recent safety standard.

Not so earthquake-proof or tsunami-proof as it was supposed to be.

Neighborhood essential facilities (e.g. transformer substations and pipelines from the dam) were not so earthquake-proof either.

GE Mark 1 might have had a fundamental design flaw. The location of emergency diesel generators. The underassessment of the potential role of ventilation in severe accidents. The capacity of a containment vessel may be too small to survive extremely high

pressure

The plant managers and operators were not equipped with a thoughtful severe accident management at all. They relies in vain on ECCS with a dim hope of early recovery of PCS. They delayed in ventilation and seawater injection.

2013/3/19-21

Fukushima Dai-ichi and old nuclear plants

45

Any nuclear plant which was built prior to the introduction of seismic standards in 1981 (though it was mild) is more or less subject to the same problems Fukushima Dai-ichi had carried. Such plants will be 40 years old very soon.

Nevertheless, the regulator was planning to allow the power companies to extend the operation period of old nuclear plants from 40 years to 60 years. Another ten year extension of Unit 1 of Fukushima Dai-

ichi had been allowed just one month before the earthquake occurred in March 2011.

2013/3/19-21

Proposal 1: Rearranging existing nuclear plants

46

Impacts of decommissioning old nuclear plants From plants existing in 2010 49.1 gWh From plants built in 1970s 13.4 gWh (27.3%) From newly planned plants as of 2010 16.5 gWh

The impacts of decommissioning old plants may not be so large as it seems, because Smaller capacity. Fully depreciated (for initial 40 years). Completed reserves for decommissioning (for initial 40

years). Costly to retrofit old plants to meet new seismic

standards.

2013/3/19-21

Proposal 2: Reorganizing TEPCO

47

The current public support for TEPCO is not sustainable at all. It assumes that regional monopoly will continue. It will be given up sooner or later.

It would be much better to immediately make TEPCO bankrupt.

After sharing losses among creditors in a reorganization process, the government would take over the following huge liabilities: Unpaid damage compensation, which was determined officially by the

Dispute Reconciliation Committee. Long-run expenses on scrapping the damaged reactors.

For liquidity reasons, a new-born TEPCO might be forced to sell power plants or power transmission facilities. A spontaneous separation between power producers and power

suppliers.

2013/3/19-21