The Transformer – October 29, 2007

Volume 4, Number 7

PacifiCorp’s dilemma


PacifiCorp is a large investor-owned utility serving portions of six states: Oregon, Utah, Wyoming, California, Washington and Idaho. In the not-to-distant past, NW Energy Coalition and other clean-energy advocacy groups commended PacifiCorp’s foresight in assessing the cost of carbon emissions while compiling its 2004 integrated resource plan (IRP) and including some 1,400 megawatts of new renewables.

Those commendations were tempered by the IRP’s call for several new coal plants. The Oregon Public Utility Commission refused to approve the coal plants’ inclusion in the 2004 IRP, while Utah’s Commission endorsed them. PacifiCorp delayed action on the controversial plants pending the outcome of its current IRP.

In the interim, PacifiCorp’s relationships with regulators and clean-energy forces have deteriorated, and more so since PacifiCorp’s recent sale to Warren Buffett’s Mid-American Energy Holdings, a coal-heavy Midwest utility with little experience in the Northwest’s eccentric regulatory culture. And the utility’s latest long-range power plan again proposes more new coal and other fossil-fuel generation than any other investor-owned utility in the Northwest.

This issue of The Transformer considers the saga of PacifiCorp’s fossil-fuel-heavy IRP.

Regulation 101: IRP basics

Most regulated (investor-owned) utilities in the Northwest must file integrated resource plans (IRPs) with their utility commissions every two to three years. An IRP provides the company’s analyses and rationales for its action plan — its blueprint of what kinds of resources it will acquire during particular time periods to meet its future electric demand.

States’ rules differ, but most require the IRP to evaluate energy efficiency and supply-side resources equally, to give some weight to environmental impacts, and to balance costs with risks. In recent years, utilities have begun to consider carbon dioxide emissions in addition to traditional impacts such as sulfur emissions, harm to fish and wildlife, water quality/quantity, etc.

Utility commission staff, stakeholders including utilities’ industrial customers, and members of the public review and comment upon the plan, and their feedback often leads the utility to modify certain elements or alter its analyses. Finally, the state regulatory body considers all the comments and issues its opinion. Depending on the state, this opinion carries varying consequences for the utility.

For example, the Oregon and Idaho public utility commissions “acknowledge” or refuse to acknowledge the plan. Acknowledgment doesn’t guarantee favorable rate treatment (the ability of the utility to recover the plan’s costs in its rates), but it does ease the utility’s burden of proof. Non-acknowledgment doesn’t prevent the utility from pursuing its plan, but it serves as fair warning that the utility will face a very high burden of proof to justify rate recovery should it go forward.

Utilities confront lower hurdles in Washington than in Oregon and Idaho.  They still must conduct extensive analyses, but need only prove to commissioners that the plans are procedurally sound (public meetings were held, alternatives were considered, etc.). It may seem counter-intuitive, but most utilities actually favor stronger regulation, since a commission’s low-level endorsement neither guarantees approval of future rate recovery nor reduces the risk that proposed rate recovery will be disallowed at a later date.

Most utilities would prefer to know that their commission thought the plan was at least reasonable before investing millions of dollars in the plan’s recommended resources, which is one reason they like doing business in Utah. The Utah Public Service Commission’s (UPSC) policy stands at the opposite extreme of Washington’s. UPSC approval essentially puts the costs in rates … unless the utility does something extraordinarily imprudent.

Two places at once, or nowhere at all?

PacifiCorp’s largest loads are in Oregon and Utah, so those states’ regulatory commissions generally take the lead in overseeing the utility’s resource planning process.

Until recently, the two states have had very different visions. To oversimplify, Oregon has been the “green” state, insisting on lots of renewables and conservation, showing a willingness to pay more for cleaner power and seeing global warming as a greater risk than higher prices. Utah has tended to consider coal plants good for economic development, to think “real utilities” don’t do conservation and to view global warming as a liberal con job. (This situation appears to be changing. In the past year, Utah’s governor has begun moving the state toward a more serious consideration of global warming.)

Differing IRP rules and conflicting resource-development philosophies can cause headaches for utilities that operate in multiple states. In PacifiCorp’s case, these pressures have combined to warp the utility’s latest IRP profoundly. NW Energy Coalition and many other intervenors conclude that the company took liberties with its assumptions in an apparent attempt to forge a compromise action plan that both Utah and Oregon could grudgingly accept.

PacifiCorp’s approach can be described as “computers on steroids.” The utility uses an enormous, sophisticated black box to analyze hundreds of possible resource portfolios against thousands of alternative futures. This approach is supposed to make decisions more objective, but in fact has the opposite effect. That’s because the policy choices are hidden in the assumptions that drive the results, and in the weighting and averaging of separate results.

But before analyzing the assumptions and modeling flaws that tilted the output, let’s quickly summarize the results. PacifiCorp’s “preferred” plan between now and 2016 calls for a 4-percent rate increase based on acquisition of the following new resources:

  • 867 megawatts of conventional pulverized coal plants
  • 1,507 megawatts of gas-fired combustion turbines
  • 2,000 (or about 700 average) megawatts of wind
  • 250-450 average megawatts (depending on cost-effectiveness) of energy efficiency

This array represents the company’s attempt to please rival masters. To its significant credit, PacifiCorp is calling for a lot more wind —about 600 megawatts more — than it did in its last IRP. It should, since wind is the utility’s cheapest resource. Unfortunately, PacifiCorp apparently placed limits on the amount of wind power its computer could choose. In the single scenario in which the computer was allowed to choose more, it called for 3,100 megawatts of wind.

Instead, the plan is top-heavy with fossil fuels, including two conventional coal plants. Such a choice flies in the face of existing and imminent legislative restrictions on greenhouse-gas emissions. Renewable energy standards are already in place in Oregon, Washington and California and have been proposed by Utah’s governor and by several U.S. Congress members. Meanwhile, Washington and California have passed emissions standards that outlaw (or make prohibitively expensive) power from coal plants that don’t store (sequester) their carbon emissions, and Oregon is poised to follow suit.

PacifiCorp claims that its modeling did consider existing and future legislation, and included carbon “adders” raising the cost of coal generation 3.5 cents per kilowatt-hour – almost doubling conventional coal’s current cost! And still the company’s computer found the fossil-heavy plan the least costly of those it analyzed.

GI (garbage in):

Carbon laundering and other misdemeanors

The utility armed its computer with some powerful assumptions and novel modeling methodology. Here’s how PacifiCorp got it wrong:

Fantasy cap-and-trade. The company did indeed model high carbon adders, but under some funny assumptions. It began by assuming that, under a carbon cap-and-trade system, it will be given free “credits” to emit carbon dioxide from the many dirty coal plants it now operates. Then, as it adds new gas plants (which produce about half the greenhouse emissions as coal) and more modern coal plants (which operate a bit more efficiently than the older ones), it will run the oldest, dirtiest ones less, allowing the company to sell excess carbon credits into a cap-and-trade market. Thus, the higher the CO2 adder, the more money the company will make!

This scenario has three major problems:

  • All anticipated cap-and-trade regimes feature declining caps – fewer emissions allowances over time rather than the constant number of credits assumed in the company’s analysis.
  • Free initial allocation of credits is far from guaranteed.
  • The strategy for reducing the output of its older coal plants will require more new expensive resources.

How can this be a low-cost strategy?  Why would PacifiCorp invest in two more coal plants when it’s planning to build enough natural gas-fired facilities to reduce production at (and, thus. total emissions from) its existing coal plants, despite years of projected load growth? The answer is a “carbon laundering” scheme, described next.

Two-for-one sale.
The assumption that leads to the counter-intuitive result above is what one must call carbon laundering. PacifiCorp proposes to build base-load coal plants, even though it needs additional power mainly to meet just a few hours a day of peak summer demand from growing air-conditioning use in Utah. During all the other hours in the year, according to the computer model, the utility would be selling the coal plants’ output into the market.

Here’s the trick. The model assumes that when the company sells power, the buyer takes on PacifiCorp’s emissions. The high carbon intensity of PacifiCorp’s mix makes this power quite dirty – about a ton of CO2 per megawatt-hour. But despite these huge emissions, the model assumes no sales-price penalty. Rather, it assumes buyers will pay the same for Pacific’s dirty power as they will pay for someone else’s cleaner power. In reality, this greenwashing couldn’t occur: After considering carbon costs, no one would buy dirty power from Pacific at $30-$60 per ton if it could buy cleaner power from other sellers at the same price.

Ignoring carbon-content penalties in market sales makes coal the most profitable choice. Indeed, when allowed to choose freely, the computer chose seven coal plants! The company’s preferred portfolio has only two coal plants, because it arbitrarily restricted the computer to that number to mollify Oregon regulators.

Color-coded electrons.
Next comes PacifiCorp’s treatment of individual states’ renewable-energy standards. California, Oregon and Washington have laws requiring that 15-20 percent of a utility’s power come from renewables by 2020 or earlier.  Pacific’s plan results in renewables making up only 8.5 percent of its systemwide generation.

How does PacifiCorp justify this? The company argues that because Utah, Wyoming and Idaho are yet to enact renewables standards, it can ship all of its “green” power to the states that have standards and give the “brown” power to the standardless states. (Some money will change hands. The states that accept brown power will get paid a premium by the states that get the green power.) The company is assuming that Utah, Wyoming and Idaho will be placated with slightly lower rates and never want any green power of their own, and that the federal government will never pass a nationwide standard. It will be interesting to see if those states’ regulators agree. In a response to our comments on this issue, PacifiCorp says it is assuming Congressional passage of a very modest federal standard, resulting in a minimal “renewables deficit” for the company.

Live long and prosper?
Another coal-serving assumption is that new conventional coal plants will have productive lifetimes of 40 years, resulting in lower costs per year.

But the long-term greenhouse-gas reduction targets in Oregon, Washington and California (not to mention those pending in Utah and the Western Climate Initiative’s regionwide goals) cannot be met as long as coal plants that don’t sequester their climate emissions are allowed to operate. They will be forced to shut down or invest in costly retrofits (if available) to capture and store their CO2 emissions. The model fails to capture the high risk that new coal plants won’t operate long enough to pay their very high capital construction costs.

Three other questionable assumptions and modeling choices warrant at least brief mention. First, the company never modeled a portfolio optimized for low carbon emissions. Second, its model “discounts” future CO2 emissions, making them less important for future generations than for us. That is, it applies a lower inflation rate to the cost of CO2 than it does to other costs. And most importantly, the model ignores potential environmental damage, including rising sea levels, lost species, fiercer storms, etc., associated with fossil fuel-based generation, counting only direct financial impacts on ratepayers.

Don’t ask, don’t tell

If all goes according to plan, the company expects its chosen portfolio – which it deems “least-cost” –– to raise overall rates about 4 percent. This is important, because at the end of the process, NW Energy Coalition and its allies persuaded PacifiCorp’s modelers to analyze one portfolio that had no coal plants. (All previously analyzed portfolios featured between two and seven coal plants.)

So how did this non-coal, mostly gas and wind portfolio perform? The computer found that the cleaner plan – which merely substituted two gas plants for the two coal plants –– would raise rates just 1 percent more than the preferred portfolio. Imagine the outcome if the computer had been programmed to value low-carbon emissions rather than to favor fossil fuels!

The reality is that PacifiCorp seems bound and determined to build coal-fired generating plants. The company suggests that the two conventional coal plants in the IRP are just proxies for whatever large base-load plants it actually acquires, so environmentalists shouldn’t worry. On the other hand, it is issuing an all-source request for generation proposals, rather than seeking separate bids for coal, gas and renewable resources.

Given how the company has programmed its computer, manipulated the models and low-balled the dangers and costs of adding two more huge carbon emitters, the all-source RFP virtually guarantees a recommendation for the riskiest and the most polluting possible resources.

Going forward

State regulators must see through the morass of charts and numbers and direct the company to take a cleaner path.

Oregon’s commissioners refused to approve new coal plants in PacifiCorp’s last plan, and they’re unlikely to change their minds now. Utah’s commission wanted even more coal last time, but now that Utah’s governor has acknowledged the reality of global warming, the UPSC might not be so friendly.

Since the other states have few Pacific customers, Oregon and Utah are expected to cast the decisive votes.  Final decisions are expected in early 2008.

Stay tuned for results. In the meantime, read the NW Energy Coalition’s official comments to the Oregon PUC on PacifiCorp’s 2007 IRP.

What do you think?

We are interested in your reactions to these articles. We will print as many responses as possible in future editions of The Transformer. Please email comments to