The news out of Dynegy on Monday came in three parts: The big splash (selling a bunch of power plants for $1 billion plus); the mandatory, closely watched bad news (quarterly loss); the mundane bad news ($450 million in cuts over four years).
While the “big splash” was no doubt the result of a lot of hard work (the company did its earnings from New York City, where buyer LS Power is based, perhaps a sign they were working the deal up to the last minute?) the mundane stuff can actually be a lot harder. Tasks like cutting costs without cutting muscle or improving performance without simply getting bigger lack the glamour but often it’s what keeps a company afloat.
Dynegy’s Baldwin, Illinois power plant, where they’ve deployed software to help take a bite out of emissions.
Dynegy’s cost cutting starts this week with about 50 Houston workers — 10 percent of the corporate workforce — expected to get pink slips. One gets the impression Dynegy was a lean organization since its near-death experience following Enron’s collapse, so the cuts will hurt.
But the personnel are just a small part of the $450 million the company plans to save over four year. Other cuts will include capital cost reductions of $25 million to $30 million per year (i.e. not buying certain equipment) and reduced operational costs of $30 million to $40 million per year.
One might think power plant operators already squeeze as much as they can out of their facilities, but there’s actually a steady stream of small, incremental improvements that, over time, can add up. In the oil refining business it’s called “capacity creep,” the 1 percent-per-year increase in refining capacity that happens despite the fact no new facilities have been permitted in years. Operators simply get more out of existing facilities by upgrading equipment, tweaking processes. Power plants do the same thing but through different means.
For example, during a 2007 refueling of one of the reactors at the South Texas Project nuclear plant near Bay City the replacement of three older turbines boosted that unit’s output by 70 megawatts, or about 5 percent. In Texas that’s enough power for about 14,000 more homes.
Early this year Dynegy completed a successful project that involved optimizing how coal is burned in its massive Baldwin, Ill. power plant using software that learns to improve its performance over time.
The South Texas Project nuclear plant recently expanded its generation capacity.
Funded by a Department of Energy “Clean Coal Power Initiative” grant (and yes, we’re aware some readers believe that’s an oxymoron), Dynegy and software maker NeuCo installed software to control the combustion, soot blowing and catalytic reduction systems, which clean the emissions before they goes up the smokestack.
The software is able to do what human power plant operators cannot: monitor the changing character of the coal coming into a boiler and constantly adjust the equipment to make sure it burns as cleanly as possible, said Stan Sander, managing director of the Baldwin complex in an interview earlier this year.
“The fuel quality changes all the time, and there’s no way you can change to accommodate that with the normal control functions,” Sander said. “The software has a sort of artificial intelligence that is always looking for ways to do it better, to burn more efficiently.”
The end result: “reduced NOx emissions (a key ingredient in the formation of ground-level ozone, acid rain and other environment harms) by 12 to 14 percent; improved average heat rate (fuel efficiency) by 0.7 percent; increased available megawatt hours by an estimated 1.5 percent; reduced ammonia (NH3) consumption by 15 to 20 percent on one unit; commensurate reductions in greenhouse gases, mercury and particulates…”
You think that’s not so much? Given that the Baldwin power plant has been able to reduce NOx by some 90 percent in the past decade by switching to a lower sulfur coal (from the Powder River Basin in Wyoming) and adding all sorts of post-combustion scrubbers, getting a few percentage points more is no small task.