Monday, May 16, 2011

Post # 44: The Carbon Footprint Of Electricity Production - Time Value of Carbon Offsets

Several days ago I had the pleasure lunching with Gregg Marland, a colleague of mine at ORNL.  Well, actually Gregg recently retired from ORNL and is now with Appalachian State University's Institute for Environment, Energy, and Economics ( ).  Gregg is a long-time protege of the late Dr. Alvin Weinberg and is well know in the climate research and carbon cycle research community.  See for instance, , and ).  He has researched and written extensively on the topic of carbon offsets and the dynamics of carbon offsets.

The idea of carbon offset dynamics probably should have, but really never had crossed my mind until I met with Gregg.  Imagine a carbon management framework in which one can purchase or barter carbon emissions offsets (say from planting a forest as a carbon sink) to counter carbon emissions stemming from an industrial activity (say manufacturing, energy production, etc.).   The issue stems from the fact that the time-dependent nature of the carbon emissions from the industrial activity is not the same as the carbon sink from the forest.  The carbon emissions come in the short-term and continue at a rate and for a period of time dependent upon the nature of the industrial activity.  However, the "off-setting" carbon sink builds through time and, (at least theoretically) could continue indefinitely - long after the industrial activity terminates.  This is, of course, unless the terms of the offset agreement allows the forest owner to harvest or "cash-in" their carbon credits at some point in time.

Anyway, this scenario raises the question, "What is the time-value or "discount rate" of time-dependent carbon offsets?  The question touches upon inter-generational equity issues similar to those I've discussed here before.  It raises the question of appropriate time-frame for the analysis.  How does one determine the appropriate time frame for the analysis?  Is it simply the period of a business agreement?  A human generation?  A human lifetime?  Several lifetimes?  What about the geological timeframe?

And what if there's a tipping point somewhere out there in the future such that the value of carbon offsets prior to the tipping point is very high, but very low after the tipping point occurs?   (I'm thinking here of scenarios such as ice cap melting, burst release of carbon dioxide due to the melting of frozen tundra, ocean turnover, etc.)

And then it struck me that the burning of fossil fuels for energy production is an interesting example of a very-longer term release or "redemption" of carbon credits or offsets.  Obviously, a hugh inventory of carbon was sequestered as the world's coal reserves were formed over geological time.  Now, we're redeeming those carbon credits in in exchange for the energy we derive from burning the fossil fuels from which they derived.

Why does all of this matter?  Because it is issues such as this that impact the dialog about the value of carbon credits (or alternatively, the rate set for carbon "taxes"), and the development of regulatory frameworks in which credits and taxes are utilized as management devices.  These issues, in turn, have the ability to significantly impact the "break-even" economics of various energy generation technologies.

Enough said for now about the topic, but it's one all of us who care about sustainable energy would do well to understand better.

Thanks to Gregg for alerting me to the topic.  I find it very interesting and hope you do as well.

Just thinking....

Monday, May 2, 2011

Post # 43: Fukushima, Defense-In-Depth, and Risk-Based-Regulation

I'll begin this post by posing a provocative question:

Does the Fukushima Dai-Ichi accident demonstrate the supremacy of "defense-in-depth" over "risk-based regulation" ?  Or put differently, "Is Fukushima an indictment of risk-based reactor safety regulation?"

Five minutes before the devastating earthquake shook Japan on March 11, neither a richter 9 earthquake or a 15 meter tsunami were considered to be "credible" events. Yet they both occurred.  Whether due to bad data, mis-interpretation of data, or poor risk methodology, almost everyone was caught "flat-footed" by the historic event.

The devastation unleashed upon Japan and the Japanese people by the quake and the resulting tsunami was horrific.

Fukushima Dai-Ichi was designed to withstand neither event.  The damage to four of the six units was extreme.  Yet, to this point in time, the reactors and primary containments in units 1-3 have apparently avoided gross energetic failures (not so the refueling pools and secondary containments, but that's another story).  I'm not minimizing the severity of the accident or it's off-site consequences – just acknowledging that matters could have been even worse.  Why aren't they?

The answer is due in large part to good old-fashioned "defense-in-depth".

Defense-in-depth is a traditional reactor safety design philosophy that integrates multiple engineered barriers and redundant layers of defense to compensate for mechanical and human failures so that no single barrier is relied upon to protect against an accident.  An example is that radioactive fission products are contained in the fuel, which is inside the fuel cladding, which is inside the reactor vessel, which is inside the primary containment, which is inside the secondary containment.   For those unfamiliar with the principle, see NRC's discussion of the topic here and here.

Many things went wrong at Fukushima, yet the engineers who designed the plant (prior to the era of risk-based regulation) incorporated a number of conservative assumptions, defense-in-depth design strategies, and robust design margins that have, until now, prevented the accident at Fukushima Dai-Ichi from evolving to a much more dire situation.

So back to my question...The truth is that risk-based regulation, properly applied, should result in designs more capable of withstanding the threats we expect them to face during their operating lifetimes.  This is good – but not sufficient.  Fukushima confirms our need for a healthy dose of humility when it comes to quantifying "credible" events.

Defense-in-depth and risk-based regulatory approaches complement each other.  Call it the "belt and suspenders" approach.  Fukushima confirms the importance of this approach.

So, when it comes to reactor safety, I'm a "belt and suspenders" man.  And so should we all be who are serious about nuclear energy and a sustainable energy future.

Just thinking...

Sunday, May 1, 2011

Post # 42: Early BWR Station Blackout Severe Accident Analyses

I've noted back in Post # 31 that unmitigated BWR severe accident sequences similar to the recent Fukushima Dai-Ichi event were analyzed as early as 1981 at Oak Ridge National Laboratory.  Several folks have asked me if the original NUREG/CR reports for the two original Oak Ridge analyses were available.  I've located them and uploaded them here.

The first report is the original 1981 report in which the unmitigated long-term station blackout sequence for Browns Ferry nuclear plant (a BWR-4 / Mark-I) was analyzed:


Recall the "long-term" qualifier implies that, while the accident sequence assumptions assumed both off-site power and on-site station diesel generators were unavailable from the on-set of the event, the station batteries were assumed available until they were exhausted four hours or longer after accident initiation.

In 1982, ORNL released a second companion report in which the fission product transport phenomenology for the Browns Ferry long-term station blackout sequence was evaluated:


I caution everyone that these were the original analyses, performed with a suite of computer codes and models (accident progression and fission product transport) that were primitive by today's standards.  Additionally, Browns Ferry is a larger plant than Fukushima, and had a shorter station battery life.  

In subsequent years, ORNL analyzed other station blackout and loss-of-decay heat removal sequences in BWRs, developed more advanced tools for invessel accident sequence analysis (the BWRSAR code, developed by  Larry Ott is particularly notable), and worked closely with Sandia National Laboratory to develop BWR modeling approaches for the MELCOR code.

In 1994, one of our colleagues, I. K. Madni, then at Brookhaven National Laboratory, performed a similar long-term station blackout analysis with MELCOR for the Peach Bottom plant (another BWR-4/Mk-I plant):


While I cannot go into all of the details here, a close read of these reports will aptly demonstrate the significant influence station battery lifetime, automatic safety system operating logic protocols, and manual operator actions can have on key accident event timings.

Just thinking...