Avoiding Risks to Power Sector Investments From An Unstable Carbon Market

Under a carbon reduction program, there would be strong economic interactions between existing pulverized coal (PC) plants that adopt CCS and those that do not. Also there would be strong economic interactions between non-retrofitted units which could potentially be displaced by new NGCC investments. Here we claim that a simple “text-book” unconstrained carbon market would introduce dynamic inconsistencies, and hence instabilities in gas and coal markets.

A solution is to limit the carbon charge to a level just sufficient to induce a dispatch order that corresponds to CO₂ emissions rates from lowest to highest. Incentivizing this dispatch order would significantly reduce CO₂ emissions without incurring capital expenditures. Then the revenue from the carbon charge could be available to fund a capacity market for CCS investments.

Limiting a carbon charge to a modest level also avoids what is known in the optimization literature as a “bang-bang” solution, in which existing capacity is retired rapidly and replaced with a new technology. The M.I.T. model seems to have this property, as IGCC units with CCS rapidly displace existing PC plants, and no existing PC plants are retrofitted with CCS. Slower changes for a carbon charge, natural gas prices, and coal prices allow for a dynamic, least-cost transition period path.

In this paper we show some simple simulations with dynamic inconsistencies. This is a well-known phenomenon in the economics literature with many important applications. The problem is as follows: First a least cost solution is computed. However, once the associated capital investments are made, new incentives arise which undermine the original solution (causing regret). For example, a least cost solution for CCS penetration, NGCC investment, and a gradual retirement path for the existing PC capacity that does not retrofit CCS needs to be checked to see if after the NGCC investments have been made that there is still an incentive to operate them as planned in the computer simulation. It could be that once the NGCC capital is sunk that there would be an incentive to run NGCC units as base-load rather than shoulder load, displacing and retiring existing PC plants without CCS. However, if this happened, the price of gas would “explode” leading to unintended, extremely volatile gas, coal, and power markets. This instability happens at high carbon prices, but can be avoided by keeping the carbon price low.

Finding the market clearing gas price is an important model calculation. Our gas supply model provides the supply side representing gas production from shale gas, offshore gas, Alaskan gas, other lower 48 states conventional gas, and net gas imports. Using data from the sensitivity cases in the AEO 2012, we employ identification theory in economics to estimate the price responsiveness of these gas supply functions over time (price movements “along the supply curve”). 

In summary, we show:

• there exists stable, dynamically consistent, transition paths with relatively small adjustment costs over which CCS is adopted systematically by existing PC plants over the period
• carbon, gas, and coal prices adjust slowly and smoothly during the transition
• those PC units which choose not to retrofit with CCS, gradually phase out, but play an important role of providing substantial electricity during the transition period
• least cost CO₂ reduction targets can be achieved.

There is a lot of important specific unit variation among PC unit characteristics. Therefore it is best left to the market to decide which existing PC units are most suitable for CCS retrofits (rather than command-and-control environmental regulations).

REFERENCES

1. Paltsev S, Reilly JM, Jacoby HD, and Morris JF. The cost of climate policy in the United States. MIT Joint Program on the Science and Policy of Global Change. 173 (2009).

1. Hanson DA, Kryukov Y, Leyffer S, and Munson T. Optimal control model of technology transition. Int J Global Energy Issues 33(3-4): 154-175 (2010).

1. Shelby MG, Fawcett A, Smith E, Hanson DA, and Sands R. Representing technology in CGE models: a comparison of SGM and AMIGA for electricity sector CO₂ mitigation. Int J Energy Technol and Policy 6(4): 323-342 (2008).

1. Hanson DA. Efficient Transitions from a Resource to a Substitute Technology in an Economic Growth Context. J. of Economic Theory, 17:99-113 (Feb. 1978).