Mark P. Mills
Public Utilities Reports – Fortnightly Magazine
PCs didn't kill mainframes. Distributed generation won't erase large central stations.
Distributed generation is the latest "killer app," creating excitement at energy conferences and seminars. A growing constituency in the environmental community says DG is to electricity what the personal computer has been to the computer industry. They see DG as a vital link in the "new" energy universe, offering a solution to climate change.Their favorite analogy: Just as PCs took down mainframes and the likes of IBM, they say, so too will DG erase central, fossil-fueled power plants and big utilities.
Don't bet on it.
DG enthusiasts believe the day soon will come when consumers can head over to Home Depot and buy a little "appliance" to take home, plug in and supply all the power needed, grid-free. Commercial units are in the pipeline from a half-dozen vendors, with market leader AlliedSignal making news and waves. These refrigerator-sized generators work like the "auxiliary power units" that airplanes use to make electricity while sitting at the gate. Target markets include tens of millions of small and mid-sized businesses. The trade press is filled with DG hype. Independence (from those "evil" utility giants that have provided us with cheap power for 75 years) looms near, they say.
Even otherwise realistic vendors of DG technology are playing the climate change card in hopes of benefitting from imminent federal largess.
Energy technology forecasters and believers in catastrophic global climate change share an ally. The anti-fossil-fuel lobby for 25 years has been predicting the imminent demise of fossil fuels, the planet's primary energy sources, and the imperative to shift to something else. The climate change threat only increases the urgency of an ostensibly inevitable transition to a post-fossil fuel world.
What's more, DG kills two birds with one technology: fossil fuels and utilities, each a perennial political target. DG, it has been claimed, will free us from central coal-fired power plants.
For those advocating a post-coal world, a viable and major alternative is needed. After all, coal supplies 55 percent of what goes into the power grid. And that percentage is unlikely to diminish significantly under any current forecast. Even with the dominance of natural gas for new generation, the Energy Information Administration predicts about 40 percent of demand growth from coal, with coal's national market share remaining at nearly half of all generation in 2015.
It is true that there are exciting developments on the DG horizon. But they will not have the transformative effect DG advocates would have you believe. In fact, DG will not replace coal plants, but will complement them and almost certainly increase the use of fossil fuels. On-site power may well even pit oil against natural gas – not exactly the outcome some enthusiasts hope for.
Efficiency? A False Analogy
The PC analogy, while seductive, completely fails. Regardless of the astronomical growth in PC use, the mother of computing's "heavy iron," IBM, is far from out of the picture, as its stellar stock performance this decade attests. IBM and its mainframe-making cousins are benefitting from, rather than being eviscerated by, the information revolution in all its forms. Indeed, IBM's latest mainframe was such a hit that the company sold a thousand of the million-dollar beasts in the first 100 days. The data traffic that PCs and the Internet create, and the data appetites that expanding applications for computing create, are driving the market toward so-called "super servers" – the 21st-century version of mainframes. There are 60,000 more mainframes installed in the United States compared to a decade ago.
But those using the DG-PC analogy usually mean to imply that DG stands on the threshold of rapid cost reductions, and will emulate the collapsing price and rising performance of PCs during the past 10 years. They warn utilities that central station power plants will follow the fate of slide rules.
The PC price-performance trend arose from advances in the technology used to fabricate integrated circuits. Declining scale and increasing speed equal lower costs. It's "Moore's Law." Still, though today's desktop is more powerful than yesterday's mainframe (though today's mainframes are awesome), Moore's Law just doesn't apply to DG and electricity. Sorry.
Power plants enjoy the distinct disadvantage of being constrained by a much longer-standing law, from the realm of physics – the Carnot limit for thermodynamic systems, which is the same for all power plants, big and small. According to Carnot's limit, the temperature of combustion sets the limit for the energy efficiency of burning a fuel. Size doesn't matter, and small actually may be worse. Technologies to tweak efficiency are not only applicable to all sizes, but many of the tweaks are easier and more cost-effective for big iron. This basic tenet holds true for all DG technologies based on burning fuels, which are the systems most likely to emerge in the near-term.
But what of solar, wind and fuel cells, the environmentalists' DG darlings? After all, photovoltaics are made from the same basic stuff as microprocessors. Sorry, the analogy still fails.
Sure, PVs are made from silicon (or similar materials) just like microprocessors. But the similarity ends here. To gain greater PC power, engineers make ever-smaller components of increasing density, thus expanding the total number of microscopic electronic devices per square inch.
But you just can't make a smaller, more efficient PV. Rather, you need more (lots more) square inches – nay, square acres – of silicon devices to gather the fuel, which in this case is the sun's energy. True, the sun is limitless, but it's just too darn far away to produce high-density power, hence the need for lots of acreage to gather the dilute power.
Wind power suffers from the same problem. Greater economy and power don't arise from making windmills smaller – you need bigger ones and more of them, lots more, to power a nation.
Then what of fuel cells, those intriguing devices that use electrochemical magic to make electricity without combustion? In brief, they're too expensive and still need fuel. The materials that make the electrochemical magic happen are expensive. Lower costs face basic, almost intractable (but probably eventually solvable) materials issues.
Fuel cells run on fuel, ideally hydrogen. Virtually all of the solar system's hydrogen is in the sun – a rather inconvenient location. So we can make hydrogen here (through an expensive, energy-intensive process) or use the hydrogen inherent in conventional fuels such as methanol and even gasoline (also a costly exercise). We will, to be sure, see real advances in fuel cells eventually, but they're no threat today to the gigawatts of conventional generation.
Growth? Don't Discount Diesel
That brings us to the last category of DG: microturbines and diesel engines. Most of the market hype surrounds microturbines, which really just are very small jet engines tied to an electric generator. They do work, but they need fuel – usually natural gas, although oil works too. They still cost too much, and despite the hype, you still can't buy one. Worse yet for efficiency mavens, they are less thermally efficient than central power plants.
That said, it is clear that practical and useful microturbines will emerge soon, and almost certainly in advance of any other new form of DG. The most likely near-term applications for microturbines will be in four areas: where reliability supercedes cost; where power is very expensive, capital scarce and incremental power needs modest (Costa Rica and Long Island, for example); in meeting costly peak demands; and in providing backup power in applications where reliability is paramount.
Remember last summer's astronomical price spike for peak power during the heat wave? Just a few of those go a long way toward covering the higher costs for DG peaking. In all likelihood, the folks installing microturbines to shave peaks will be the same as those operating or selling coal-fired base load power to create a seamless, blended, reliable and economical power source.
Ironically, the only immediately cost-effective DG technology is the venerable diesel engine. So-called diesel-gen sets exist by the tens of thousands, powering oil fields, small villages and military bases. Recent advances in materials and controls have made diesels even cheaper and more efficient (better than microturbines) and exceptionally reliable. And you can buy them right now. One suspects that credible providers of real-world DG like Caterpillar and Wartsilla are salivating on the legislative sidelines as markets open.
Deregulation likely will generate a boom in the use of diesel gen-sets. They can burn oil or natural gas, and in many applications use the former. That clearly is not what many environmentalists intend deregulation of utilities to effect.
Perhaps the worst nightmare for the inveterate anti-coal lobby is the potential of new technologies to achieve cheap off-peak kilowatt-hour storage – distributed storage. Small, high-tech flywheels appear to be promising (just park them outside next to your central air-conditioning unit). You spin them up at night with an electric motor powered by otherwise "wasted," ultra-economical off-peak power (probably 1.5 cents per kilowatt-hour). The motor works as a generator in the daytime, drawing kinetic energy off the flywheel: easy, reliable, no new fuels, one moving part. Slick. It uses the cheapest off-peak power too – hydro (and nuclear) in a few places, coal everywhere else.
The capital costs for diesel gen-sets already are a lot lower than for central power plants. Given that and the low cost of fuel, why isn't every business making its own power? Few end-users want the operational and maintenance hassles. Electricity coming off the grid is awfully low-maintenance. The collective cost of tending to millions of distributed (quirky) products remains the showstopper.
Nonetheless, significant and viable niche markets for DG are inevitable, probably up to 10 percent of total U.S. demand. Once momentum starts building and reliability grows, emerging technologies can make a noticeable dent in new supply. A critical leap for fuel-based DG will be cost-effective, network-based remote maintenance and monitoring of distributed equipment through advanced sensors, information technology and neural networks.
Distributed generation is coming, of course, inevitably in a much bigger way than experienced to date. And the computer analogy does work, but in a slightly different manner than is commonly understood. Just as PCs drive demand for mainframes (now re-named "super servers"), so too will DG drive demand for more large, efficient and low-cost central power sources.
© 2007 PUR Inc.