The Credit Impact of Climate Change

Swami Venkataraman, Director, Corporate and Government Ratings, Standard & Poor's

The energy industry is bracing for a carbon-constrained world, but the ways governments choose to control carbon around the globe may have vastly different impacts on company bottom lines.

INDUSTRY TODAY GENERALLY ACCEPTS THE inevitability of mandatory carbon controls and is now seeking to influence the final form of these controls and negotiate the future participation of developing nations in a global carbon regime. Credit consequences may result as restrictions on greenhouse gas (GHG) emissions cause significant increases in capital costs and/or reductions in profitability. Standard & Poor's sees carbon controls impacting power sector credit quality globally in four broad ways.

Sectoral Distribution

The distribution of emission reductions, and hence costs, among various sectors of the economy will vary substantially depending upon the mechanisms chosen to implement carbon legislation. A cap-and-trade approach, taken in isolation, may result in a disproportionate allocation of emission reductions to certain sectors while not meaningfully affecting others at all. The power and automobile sectors provide a prime example of this. Some US Senate bills propose that oil refiners be responsible not only for their own emissions, but also for auto tailpipe emissions. However, refiners control neither the fuel efficiency of cars nor the driving habits and model preferences of drivers. At best, refiners can indirectly affect such decisions by passing through to customers the cost of carbon allowances in the form of higher gasoline prices. But this is potentially a weak price signal. To take an extreme example, at a price of $100 per ton for CO₂ credits, the price increase to consumers would only be about $1 per gallon of gasoline (burning 100 gallons of gasoline produces one ton of CO₂), a sum drivers have absorbed in the recent past without switching en-masse to less-polluting vehicles.

If an economy-wide emission cap is adopted in the US, legislation that mandates higher fuel economy for autos, or greater use of ethanol, biofuels, etc. would be a key determinant of how much reduction is achieved from autos and how much is demanded from other sectors. Power generation may end up with a disproportionate share of emission reduction responsibility (and the attendant credit risks) because, even at $100 per ton for carbon credits, auto emission reductions will depend on the extent to which consumers and automakers view higher gas prices as permanent and change their behavior. By contrast, with carbon at $100/ton, the power sector could become almost entirely emissions-free, as mostif not allestimates of the cost of capturing and sequestering carbon are less than $100 per ton.

Impact on Existing Power Plants

The choice between auctioning and allocating carbon credits in a cap-and-trade regime has the greater impact on the value of existing generating assets. The free allocation of CO₂ emission allowances in Phase I of the European Union's Emission Trading Scheme (ETS) allowed gas and coal-fired generators to be more profitable than in the absence of the ETS. This profitability will decline in Phase II and beyond as more credits are auctioned rather than assigned and the absolute level of freely granted allowances to the power sector is likely to decline. Regional initiatives in the US, such as the Regional Greenhouse Gas Initiative (RGGI) in the Northeast, are looking at auctioning a majority of their credits.

We used a dispatch model licensed from EPIS by Platts to identify aspects of the power markets that drive compliance costs. We've estimated the economic cost of compliance as the change in EBITDA that a power plant (or portfolio of plants) earns under a base case with no carbon controls and under two GHG scenarios, each modelled after one piece of pending Senate legislation—the Carper/Feinstein bill (GHG1), and the more stringent Boxer/Sanders bill (GHG2). We ignore factors such as regulation and contractual arrangements since these will only influence how costs are allocated and not the economic cost itself.

Given any level of allowance auctioning, two factors have a key influence on the value of power generation assets in a carbon-constrained world.

1. The characteristics of the power markets in which the company operates, chiefly the fuel that sets the marginal price, and how that changes over the years: Whether a coal plant recovers closer to 40% or 100% of its emissions costs through power prices will depend on the number of hours gas is on the margin and the number of hours coal is on the margin. At the same time, power prices in gas-driven markets will reflect much more of the demand-driven increase in gas prices, which increases the marginal costs of gas-fired generators but not of coal-fired generators. This can partly offset the higher carbon cost for coal units. Whether coal-fired generators are better off in markets dominated by gas or by coal would depend on which of these drivers dominates, although the carbon price effect is more likely to dominate. Hours with coal on the margin increased substantially in many US markets under GHG2, with inefficient coal plants being the marginal units and the relationship between more and less efficient coal plants becoming similar to combined cycle and peaking gas plants today.

2. The nature of a company's generation portfolio, whether it is fossil-heavy, carbon-light or diversified. Table 1 below shows the impact on ten large generators, who we're not naming because the results are also applicable to others with similar portfolios, and it's not our intent to single out the companies analyzed.

1. Ratio of 2026 (in %) EBITDA to 2007 EBITDA (both nominal dollars).*

Fossil-heavy portfolios obviously suffer the most. Their EBITDA is basically flat 20 years from now and even 10% to 20% lower under GHG2. Such companies face three areas of concern:

*The lack of growth in EBITDA (even in nominal dollars) will be a concern for investor-owned utilities,

*Just maintaining this EBITDA will likely require significant ongoing capital spending not captured by our modeling, and

*A flat-to-declining nominal EBITDA means that existing assets contribute almost no cash flow toward meeting future load growth, implying greater reliance on external financing.

It appears that well-diversified companies may be indifferent to whether carbon legislation is stringent or lenient, as the change in EBITDA is the same under both GHG1 and GHG2. Of course, these factors determine only the total magnitude of compliance costs; the actual credit impact will be determined by the ability of any company to pass these costs through to customers.

Impact on Future Power Plants

The nature of future power plants promises to be substantially different from that of the existing fleet. If a global consensus develops around the need for post-Kyoto legislation that achieves a long-term CO₂ concentration target anywhere close to the 500 +/- 50 parts per million indicated by many scientific models, no technological silver bullet exists to achieve this reduction, and a portfolio of strategies would be required. A paper that Stephen Pacala and Robert Socolow of Princeton University published in the journal Science in August 2004 estimated that emission reductions of about 7 gigatons of carbon-equivalent (GtC) per year would be required by mid-century compared to a business-as-usual scenario to stabilize CO₂ levels at this target. The number is perhaps higher now given that economic growth generally and growth in Asia in particular has been stronger than assumed in many models. With developing countries, which accounted for 49% of global 2006 carbon emissions, expected to grow emissions 60-80% by 2050, OECD countries need to reduce their emissions by about the same amount to keep world emissions constant. Socolow and Pacala also calculate the magnitude of effort that will be required under 15 different GHG reduction options in power, autos, afforestation, agriculture etc. Power-related options are shown in table 2, which illustrates the magnitude of the emissions reduction task.

2. Power-related Options. Source: S. Pacala and R. Socolow

So Which Technologies Will Get Built?

The power sector's future fuel mix can vary significantly, as demonstrated by table 3.

3. Power sector fuel mix changes.

Energy efficiency represents the amount by which consumption from the grid is lower in 2026 under GHG legislation compared with the base case. In our model, energy efficiency represents a 0.35% reduction in annual demand growth in GHG1 and 0.6% under GHG2. Based on historical econometric regressions, Platts estimates that about 25% of this reduction represents a price elasticity response, with the remainder attributable to proactive energy efficiency. The model assumed that all state renewable mandates will be met under both GHG1 and GHG2. Economics did not drive further renewable builds under either scenario, although this could vary if gas prices are higher or if utilities were to build wind for regulatory or public relations considerations. In GHG1 and GHG2, a tremendous decline in coal-fired power occurs, with about a 6% decline in absolute megawatt-hours under GHG1 and 16% under GHG2 from 2007 levels, the result of fuel-switching and retiring old units. Coal generation is actually flat to moderately rising until around 2020, when GHG1 and GHG2 require significant reductions in emissions. This suggests that any legislation that provides for a fairly long initial period of modest emission abatement (perhaps to allow for R&D and for an international regime to develop) will provide significant time for coal-based generation owners to adjust.

Nuclear would have grown much more substantially in our scenarios if not for our restriction on the number of new nuclear plants that utilities might build. IGCC with carbon capture and sequestration could contribute 3% of the nation's energy by 2026 under legislation such as GHG2 and more if IGCC costs decline. Carbon credit prices aren't high enough to justify building IGCC units based on the economic merits alone under GHG1. However, GHG1 results are sensitive to the availability of about 200 million tons of offsets, which provide the bulk of emission reductions. The split between IGCC, nuclear, and gas-fired plants will depend on advances in IGCC technology, costs of nuclear power and the price response of natural gas to increased demand. We assumed gas prices 5% higher than the base case in GHG1 and 10% higher in GHG2.

The credit impact varies by the choice of technology. Energy efficiency reduces utility revenue growth, and thus margins, in the absence of regulatory "decoupling mechanisms" that separate utility profits from sales. Coal gasification and CCS suffer from higher costs, lower reliability and a lack of commercial track record, all credit negatives. CCS also suffers from substantial legal and regulatory risks if utilities are tasked with monitoring CO₂ storage sites over periods potentially lasting thousands of years. Wind energy, which currently has the greatest potential among renewable sources, is an intermittent resource and requires additional investment in back-up generation, while nuclear power involves extremely large capital expenditures and suffers from construction, waste disposal and terrorism risks.

Conclusion

Ultimately, any system will be effective only when it makes emissions costly. If the goal is to stop and reverse climate change, policy decisions will only determine the timing of the associated costs. However, policymakers will have to sort out several issues, primarily those pertaining to cap and trade versus a carbon tax; allocations versus auctions of credits; acceptance of global offsets, and economy-wide versus sectoral approaches. We believe the U.S. eventually will adopt an approach that will attempt to avoid the spectacle of coal power plants making surplus profits from emissions allowances, as has happened in Europe, while also minimizing costs to the most affected units (vintage coal) at the program onset. In theory, this should provide these units time to deploy the most optimal abatement measures.

For a more detailed discussion of our results, please consult Standard & Poor's CreditWeek special issue of May 23, 2007 titled "The Credit Impact of Climate Change."