Common Tragedies

Thoughts on Environmental Economics

CCS and the long haul

Posted by Evan Herrnstadt on February 29, 2008

When carbon capture and sequestration is discussed, the phrase “in perpetuity” tends to come up. This cannot be literal, but what we conceptualize as forever is at least a very long time, and someone needs to be liable (or at least responsible) for potential damages related to geological sequestration (GS).

I recently attended an EPA public workshop on Underground Injection Control (UIC) regulations for carbon sequestration. The Agency is in the midst of writing regs pertaining to GS of carbon, and they were looking for input. It was made perfectly clear that no comments were to be attributed to participants during the main session, so there’s going to be a lot of anonymity in this post.

I went to a breakout session on financial assurance of long-term site care and monitoring (LTCM). Although the discussion was meant to focus on EPA regulations under the Safe Drinking Water Act, it spread to broader themes of long-term liability and responsibility. Specifically, the conversation quickly turned to state indemnification (Note: EPA has no authority under the UIC regime to transfer liability away from the owner/operator of a site; it would take an act of Congress to do so).

The main argument for indemnification in GS is that uncertainty reigns at this point, and the damages associated with a major release or leak are sufficiently large as to deter investment in sites. One of the speakers countered this idea, stating that there are surely some firms that would be willing to take on this liability — it all depends on whom you want in the GS business. In response, a participant memorably noted that to store carbon “in perpetuity”, we need an institution to last as long. He suggested the Catholic Church, as it has significantly split only once in the past 2000 years or so. Joking aside, his point was that corporations are too short-lived to take on liability stretching far into the future. However, he claimed that a nation would be the next-best option.

Indemnification is by no means new to the U.S. government. The Price-Anderson Nuclear Industries Indemnity Act is a prominent example. After purchasing the maximum available $300 million of private insurance, licensees have to be prepared to contribute an additional $100 million or so to an industry insurance pool. However, the licensees are not liable for damages beyond that point. Other examples are the Terrorism Risk Insurance Act and the Oil Protection Act (OPA).

The OPA is perhaps the most applicable model. It governs the Oil Spill Liability Trust Fund, which is maintained with a 5-cent/bbl tax on oil. It is capped at $2.7 billion and may be drawn upon if a member absolves itself of negligence and illegal activities. A similar GS trust fund could be established with a small tax on each ton of carbon stored. However, the heterogeneity of sites could be a barrier to such an arrangement. Different types of GS carry different levels of risk. Former enhanced oil recovery sites are generally well-explored and understood, versus relatively mysterious saline formations. EOR sites also are sure to have horizontal caps which held the oil in the first place. Saline formations have potential for horizontal leakage.

This leads me to the major themes of the workshop at large: flexibility and heterogeneity. First off, the demarcation between the post-closure and LTCM stages should be somewhat flexible. If there is a strict rule that LTCM kicks in 30 years after closure and the government provides indemnity, then there is massive incentive for poor practices. Clearly, a firm would be tempted to minimize its closure costs, subject to the 30 year minimum constraint.

Second, the rules must be a bit amorphous to allow for future technologies and policy. As we develop better models and surveying technology, some of the uncertainty will melt away. Also, there is the possibility that carbon could become something of a commodity, within a carbon pricing scheme, or perhaps as some sort of industrial input, a la EOR. The regulatory structure should be sufficiently pliable such that a transition from a pure waste management situation toward a resource situation is possible. One speaker put it succinctly: “Failures come from easy, arbitrary decisions; it’s hard to guess what we’ll have in the future.”

Third, it was suggested that any indemnity arrangement must have room for repricing. I found the point solid, but the argument a bit odd. The claim is that if the fund is getting too big (small), then the tax should be reduced (increased). I don’t quite understand the argument for cutting the tax if the fund is too big. The underlying reason for the bloating purse could simply be a matter of probability: the more extreme parts of the underlying damage distribution (for which indemnity exists) have not happened to play out. Still, it makes sense that if an indemnity program is underfunded (as they almost always are when actually tested), that the regulation should allow for some wiggle room.

Regulating GS is an extremely tortuous, intricate process, and there are significant institutional boundaries. For instance, the EPA is attempting to regulate GS under the SDWA, but this only covers leaks of CO2 or leaching of metals and saline solution into groundwater. There are many other long-term liability issues, such as climate liability should the U.S. establish a carbon pricing regime (who buys up permits in the case of a release?). We have a long way to go, but it’s good to see initial forward-looking policymaking which helps reduce the regulatory risk associated with CCS investment.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: