JUICE: Catch-22/MMBtu

By Published On: May 19, 2006

I get an earful about using computer models to figure out future supply and demand. Many dinner conversations chez moi revolve around the assumptions, data input, and algorithms calculating relationships (try saying that fast five times in a row) that my husband grapples with when building and reviewing models for groundwater resources. While I remind him to pass the salad, he points out the challenges arising from changing hydrologic systems – primarily changes resulting from global warming. Global warming alters fundamental assumptions – historic flows – plugged into his supply and demand models. At the same time, his method to find out whether his models work correctly is to compare them with existing models. Those rely on historic, and sometimes outdated, data. It creates a Catch-22. The dilemma my hydrologist hubby faces is the same one natural gas modelers confront. The models are based on past assumptions used to predict what the future may bring. Those assumptions may no longer apply, or should be given less weight to improve guesstimates of what is ahead. It’s like spinning a 33 rpm record in a land of iPods. Take the California Public Utilities Commission’s natural gas price projections for 2006-30. The table summarizing the findings shows a significant drop in gas prices over the next 25 years. The line on the chart that is used to set the benchmark price for renewable power supply contracts between the investor-owned utilities and power producers starts at about $11/MMBtu for 2006. The line drops steeply to reach a little above $6/MMBtu in 2011. From then on, the line curves slightly above and below the $6 mark through 2030. (Yesterday’s price just dropped to the $6 range, according to Federal Energy Regulatory Commission staff.) The graph speaks a thousand words. Those words might as well be Greek, however, as we don’t seem to be in a world of lessening gas prices but in one of diminishing gas supplies and escalating demand. While I didn’t get an insider’s view of the CPUC model over dinner, like the groundwater models I’ve heard so much about, I know it relies on assumptions and data that do not reflect the new global reality. It’s the nature of the modeling beast. The CPUC gas forecast at issue is fairly consistent with other models, including those used by the Energy Information Administration and NYMEX and ones estimating the value of the Tehachapi wind projects. They rely on similar assumptions about the future marginal cost of power production. But the models, EIA’s in particular, are not known for their accuracy. They are even less reliable in our quickly changing world. For one thing, price volatility isn’t factored into the mix. Mercurial gas prices have become the norm in the past few years, and many say they are here to stay. Costs reached unprecedented heights at the end of last year – more than $15/MMBtu. The historic range was $2/MMBtu (Circuit, Dec. 9, 2005). Why are we seeing rate increases if prices are dropping per the model? Earlier this year, San Diego Gas & Electric, Southern California Edison, and SoCal Gas all sought and were granted rate increases because of higher natural gas prices. Munis, including the cities of Roseville and Palo Alto and the Sacramento Municipal Utility District, also raised their ratepayers’ rates to reflect costlier gas supplies. The Los Angeles Department of Water & Power, the state’s biggest muni, is planning a similar increase. Apples and oranges, modelers may say. If so, why are we ratepayers not getting the benefits of lower future rates? “Any model can only deal with what you put into it,” noted David Maul, president of Maul Energy Associates and former natural gas office manager for the Energy Commission. Or as Rich Ferguson, Circuit’s weekly gas columnist, noted, “Garbage in, garbage out.” It is not just modeling’s inherent flaws that bother me. It is the inconsistency in comparative price predictions. The renewables market price referent’s (MPR) long-term projections, for example, were limited to that matter. (The MPR for renewable energy sets the average renewables price. Supplies that exceed that amount can tap into public-goods funds.) The CPUC’s energy division analyst Paul Douglas pointed out that the MPR graph is a “very specialized forecast,” one that tries “to calculate what a long-term contract would include for gas prices going out 20 years.” Why is it so specialized? It’s because models are built to answer specific questions. Thus, it boils down to the questions a model is asked to answer. While I don’t know the underlying policy question, I know that using a lower gas cost projection reduces the amount of renewable supplies that will be built, because financing becomes more difficult. It also goes against the grain of laws, policies, and goals intended to increase the supply of fossil-fuel-free and fossil-lite electricity. Moreover, consider the CPUC’s decision approving investments in new steam generators at Edison’s San Onofre nuke plant. The commission’s proposed decision found the investments to be “marginally economic” based on one gas price scenario. The final decision – based on a higher gas price scenario – found the investments to be “economic.” In reaching its final ruling, the commission doubled the projected future cost of replacement gas-fired megawatts (Circuit, Dec. 16, 2005). Why are separate multiyear gas price models and calculations used for the renewables benchmark, the San Onofre upgrade, and the price paid to qualifying facilities? Please pass the salad dressing and clarify the policy question. Once we know and agree on the underlying questions that models are built to answer, the next step should be to revamp the computer modeling to better reflect our changing world. Although it is no easy task, California, which is a world leader on many issues, from greenhouse gas emissions reductions to other cutting-edge environmental measures, should lead the way. Beyond technical modeling factors there are the unknowns of politics and the geostrategy of nations. Will we drop bombs on Iran? How much liquefied natural gas will China and India seek? Will other fossil-fuel-rich nations nationalize their markets? Although there is no crystal ball, a better forecasting job with the known variables can be done. Now my husband has had to remind me to pass the dessert as I grapple with the changing world paradigm. Elizabeth McCarthy

Share this story

Not a member yet?

Subscribe Now