George Apostolakis

Ep 177: George Apostolakis - Former Commissioner, U.S. NRC
00:00 / 01:04


From student to nuclear professor (0:17)
0:17 - 5:23 (George discusses his first introduction to nuclear and the classes he taught at UCLA and MIT.)

Q. How did you first get into nuclear energy?
A. George was born in Greece and obtained a degree in electrical engineering. George then attended CalTech for both his Master’s and PhD. In 1974, he went on to become an associate professor at UCLA in the department of mechanical and nuclear engineering and moved to MIT in 1995 to teach in the department of nuclear science and engineering.

George became interested in nuclear risk assessment as an associate professor, becoming a pioneer in this field. He helped establish courses on nuclear reactor safety and probabilistic risk assessment, which is still available at UCLA. At MIT, George taught a popular course on reliability, decision analysis, and risk assessment. In 2010, George became an NRC commissioner and is now the Head of Research for the Nuclear Risk Research Center in Japan.

Using probability to safeguard against risk (5:24)
5:24 - 13:51 (George discusses what probability means for nuclear risk and the first methods used to determine risk and safeguard against it.)

Q. What are the main methodologies that are used to understand risk, safety and technological systems?
A. The basic rules of probability create the backbone for risk assessments. This includes the probability of the union of events, intersection of events, elementary probability theory and statistics. Most importantly, understanding what probability means is key for risk assessments. Unlike looking at probability as relative frequency, nuclear reactor events are very rare, meaning conceptual problems arise when taking a relative frequency approach. Fortunately, probability can also be seen as a measure of confidence. This relies on evidence, which is established in repeated trials or from new information, such as new statistics and expert opinions. Statistics and expert judgements come together to form state of knowledge, which is expressed through probabilities. These probabilities are put into a probabilistic risk assessment (PRA), which is a distribution of probabilities that express uncertainty.

In the early days of nuclear energy, physicists realized that nuclear accidents were a possibility. In the 1950’s-60’s, scientists were unable to quantify this risk, so they created the concepts of Defense In Depth and Safety Margins. Defense in Depth are the multiple barriers put in place to prevent release of radioactivity in the event of an accent. Safety Margins make sure there is a margin between the operational and failure points, such as temperature. Together, these decrease the probability of an accident. These deterministic and prescriptive principles created the basis of the regulatory system. This bottom up, primarily technical approach focuses on the idea that designing a plant to withstand severe accidents will ensure smaller accidents will also be covered.

Quantifying risk to influence regulation (13:52)
13:52-23:56 (George discusses PRAs and how they came to be used in regulatory matters.)

Q. The RSS was the first PRA?
A. PRAs, on the other hand, take a top down approach and are not focused solely on technologically caused accidents. The first PRA was the Reactor Safety Study (RSS) published in 1974 by the Atomic Energy Commission. It looked at the entire plant and brainstormed every possible accident, including those caused by human error. PRAs focus on 3 questions: what can go wrong, how likely is it, and what are the consequences? Millions of potential accident sequences are generated using computers by looking only at the plant level and not taking regulations into consideration. These accident sequences are then ranked by likelihood of occurrence and the consequences are determined. In general, the dominant nuclear accident sequences total 15 to 20, which is far less than those determined for space shuttles (about 1,000 single failure dominant accident sequences). This difference is because nuclear plants include Defense in Depth measures which protect against single failure accident sequences. This is often not possible for aircrafts that must consider weight for flight.

During the late 1970’s, the NRC received criticism and pressure from engineers who did not study probability and from society who thought the RSS underestimated risks. Because of this, the NRC told staff not to use the RSS in regulatory matters. Unfortunately, Three Mile Island occurred in 1979, which had been mentioned as a possible sequence of events in the RSS. This accident moved the NRC towards using the RSS in regulations and setting acceptable probability levels. In 1986, the NRC issued the Quantitative Health Objectives or Safety Goals, which state that the accident probability should be less than 1/10th of 1% of all other risks. While the NRC submitted regulations to decrease risk, they were unpopular within the industry who did not see significant safety improvements. To appease the industry, the NRC issued a regulatory guide known as 1.174, stating how risk information can be used in regulatory affairs.

Becoming an NRC commissioner (23:57)
23:57 - 30:38 (George discusses his role in the development of 1.174 and his position with the NRC as a commissioner.)

Q. What was your role in 1.174?
A. George was appointed to the Advisory Committee on Reactor Safeguards (ACRS) in 1995 and undertook an advisory role in the creation of 1.174 in 1997. This guide focused on how to use both traditional methods of Defense in Depth and Safety Margins alongside PRA to create a risk informed approach to regulation. George considers 1.174 to be one of the major achievements of NRC staff.

After 15 years on the ACRS, George was appointed by the White House to join the NRC as a commissioner. George moved to Washington DC in 2010 and managed a staff of 6, including a technical advisor on nuclear materials, an advisor on nuclear reactors and a legal advisor. In this role, George spoke with many stakeholders, receiving visits from industry who voiced complaints or advocated for particular things. George was one of five commissioners who are subject to special rules, including one stating that no more than three commissioners could be in the same room at the same time.

The NRC’s reaction to Fukushima (30:39)
30:39 - 37:42 (George describes the aftermath of Fukushima)

Q. A year after joining the NRC as commissioner, Fukushima happens. What was this experience like?
A. People were scared, especially on the West Coast. In the beginning, Japan was reluctant to accept foreign aid, but changed their position once the scope of the accident became more apparent. In Washington, the NRC chairman was called to the White House and a member of the NRC staff was sent to Fukushima as an adviser. The NRC also formed the Near Term Task Force (NTTF) to make comprehensive recommendations on what actions the US could take.

The first 2-3 recommendations were under Adequate Protection, meaning the NRC held no discussion with industry and implementation was mandatory. Adequate Protection is a concept that ensures adequate protection of public health and safety without the consideration of cost. These regulations are infrequent and only created when absolutely necessary. The other recommendations made by the NTTF were varying degrees of urgency and some recommendations are still under consideration today.

Japan’s nuclear industry after Fukushima (37:43)
37:43 - 46:03 (George discusses the extreme measures Japan undertook to restart their nuclear industry post Fukushima. He also describes the differences between the US’s and Japan’s approach to nuclear regulation and the creation of the NRRC.)

Q. Did your work on the Fukushima recommendations lead to your involvement with the new research center in Japan?
A. After Fukushima, Japan’s plants were shut down and their regulatory system was reinvisioned. The Nuclear Regulation Authority (NRA) was established, which took the opposite approach to the prior regime of being intertwined with industry. This system was more extreme than the US, becoming completely isolated from industry. The stringent new regulations cost plants hundreds of millions of dollars to restart operations. Additionally, local governments in Japan were given the power to disagree with the NRA’s safety decisions, meaning local mayors could stop a plant from regaining operation. George sees this as a problem because it is unclear on what basis a local government is using to determine plant safety.

PRAs were also taken more seriously in Japan post Fukushima. In 2014, the Central Research Institute for the Electric Power Industry (CRIEPI), which was established in the early 1950s, created the Nuclear Risk Research Center (NRRC). The NRRC focuses on PRAs, nuclear risk and risk management. After George’s tenure on the NRC ended in 2014, he became the NRRC’s Head of Research.

The challenges of communicating nuclear risk (46:04)
46:04 - 55:10 (George describes his role in the NRRC and the challenges involved in communicating nuclear risk.)

Q. 46:04 - 55:10 What is your role as Head of Research in the NRRC?
A: George works to develop probabilistic models for natural disasters, including tsunamis and earthquakes. He also creates strategic plans (how to use risk information in decision making) and action plans (how to implement strategic plans). George hopes to move Japan away from a deterministic process and towards one that adopts the concept of risk management.

George also works on communicating Japan’s new nuclear culture to various stakeholders through newspapers articles and meetings with senior people, such as Chief Nuclear Officers (CNOs). This has its challenges because the nuclear establishment has been educated and operated in a traditional engineering environment, which is resistant to change.

Communicating with the public is also challenging. Many Japanese communities have a negative attitude towards nuclear energy because they feel they were lied to regarding plant safety. Public attitudes are slowly changing, which is critical in restarting plants in local communities. Similar to public views in the US after Three Mile Island, Japanese public confidence is shaken. It may take many more years before Japan widely supports nuclear energy.

A future with cheaper, safer nuclear reactors (55:11)
55:11- 1:00:15 (George states his view on the future of nuclear energy, focusing on the economic and safety advantages of SMRs.)

Q. Where do you see the industry going and what do we need to think about in the future?
A. In the US, the state of the nuclear industry is not good. Because of Defense in Depth and other safety measures, plants cost billions of dollars to build. Additionally, fracking has caused natural gas to become inexpensive. The cost difference between nuclear and natural gas is significant, greatly influencing investment funds and threatening the future of nuclear.

To counter this economic problem, the development of SMRs must be supported. Because of the small, modular design, SMRs relieve the financial burden of new construction. Electricity can be generated immediately after one unit is built, unlike large reactors that require spending billions of dollars prior to producing a single megawatt. SMRs are also safer than large reactors because they require smaller radioactive material storage.

  • Spotify
  • iTunes
  • YouTube
  • Twitter
  • LinkedIn

© 2020 Production of the Energy Impact Center