Buying microturbines to generate combined heat and power (CHP) usually is done for sites where lots of natural gas is being burned and the relative impact of "free" electricity will be great. How, then, could someone possibly justify installing a 60-kW generator when local power is already super cheap and where fueling it would necessitate major increases in natural-gas consumption? Why would a plant drawing 1.8 MW even care about deriving a puny 3% from another source anyway?

Surprising answers were discovered recently at Stripco, a steel processing plant in Mishawaka, IN. The story illustrates how the often quirky rules for determining utility rates can create unexpected opportunities; a proposition that was seemingly nonsensical can be transformed into a big winner. Here's how managers at this heavy industrial plant learned they could save at least $50,000 a year by investing in CHP cogeneration—and gain several other very significant advantages in the bargain.

Natural Gas: Erratic Consumption and Volatile Pricing
Stripco, a specialty services firm just east of the Indiana steel belt, uses precision cold and hot rolling machines to provide ready-to-use steel for buyers nationwide. Its processing center can anneal, slit, roll to thickness, finish the edges of, cut to length, treat, and condition the steel to a customer's specifications. In one particular process, a rolling mill uses 100 kW of electricity to warm a protective oil. In another, annealing equipment burns about 18,000 therms of natural gas in summer months; gas consumption goes way up in wintertime. In a third process, a heated fat derivative is applied to finished steel as a protective coat and yields a shiny, crisp look. Still more natural gas is burned judiciously in the cool months to warm the work spaces in Stripco's two buildings, totaling 170,000 ft.² and housing 150 employees. All told, natural-gas consumption at Stripco in 2001 was around 255,000 therms, but customers' orders and production volume were on the rise.

2001 gas costs averaged out to about $0.76/therm, but the real story here is the market's wild volatility. In January, as Technical Services Manager Joe Eads remembers, prices soared to $1.15/therm. By November, the market had fallen, and he was paying just $0.49. In between, natural gas had jumped all over the map. Eads seemed to be buying, on average, about $15,000–$20,000 worth per month as a baseline—a figure that often doubled during cold weather.

Stripco's electricity costs were more stable, running about $20,000 monthly. This was based on charges of just more than $0.03/kWh, a figure which Eads found to be—based on utility price surveys—one of the lowest in the nation. Hence, in a typical month, Stripco might use 320,000 kWh at just more than $0.03, totaling $10,240 for its baseline billing. On top of this, Stripco would be surcharged for peak rate usage, calculated by metering the power use during the highest 15-min. increment (which might come in at, say, 1,350 kWh) and then multiplying it by $6.70. This adds another $9,000 monthly (in this example), nearly doubling the base rate, and Stripco is forking out about $20,000. Still Eads knows his electric bill was cheap in relative terms and probably impervious to much haggling or discounting.

On the other hand, cutting down on the cost of natural gas—or at least leveling the wild monthly swings—seemed more doable. With that in view, Eads called his local gas supplier, NiSource Inc. of Merrillville, IN, a diversified firm doing everything from exploration and production to storage and transmission. What, he asked, could Stripco do to save on its $200,000 annual natural-gas bill?

Eads wound up talking to Mike Zdyb, vice president of business development for a division called NiSource Energy Technologies (NET), which sells energy-related capital equipment. Zydb advised Stripco to consider buying a 60-kW microturbine for CHP cogeneration.

Hellooo? At first glance, the idea seems surreal. A heavy industrial plant wants to cut its natural-gas cost—not add to it. They're already buying megawatts of power at bargain rates. Where's the logic? Eads would be burning more fuel to make higher-priced power he didn't need.

Zdyb patiently explained. If Stripco could produce even a small amount of power to heat whatever processes currently were being gas-stoked, this would enable Stripco to be reclassified as having an alternative fuel, thereby qualifying it for "interruptible" gas rates. In this category, a customer may purchase a fixed block of gas for a year at a time, at a set price for considerable discount. The pricing is so good that it almost approaches what he'd pay to hire a commodity broker and buy gas at the wellhead. And this rate reclassification would instantly accomplish Stripco's primary goal, saving money on fuel.

Eads took the odd-sounding tradeoff to his boss, Stripco President Jack Hiler. Receptive but not fully convinced, Hiler basically favored anything that would lower his gas costs and that made at least some business sense.

As Eads recalls, the question was this: What price-point for a year's supply of gas at these knock-down rates would yield at least breaking even "or [going] maybe a little better"? The answer was slightly less than $0.50/therm or 30% less than the $0.76 Stripco was paying when the concept first arose in 2001.

You might be wondering why on Earth a commodity supplier would consider agreeing to—let alone propose and facilitate—such a price slash. It will help if you know that Eads now was negotiating with two NiSource divisions simultaneously. One, the Northern Indiana Public Service Company (NIPSCO), was selling Stripco its gas to the tune of about $200,000 a year. The other, NET, was attempting to sell Stripco one or more natural gas–burning turbines or possibly other capital equipment. Normally these two subsidiaries would complement each other perfectly: A 60-kW microturbine might burn about 100,000 therms of NIPSCO product each year. In this particular case, though, purchasing a turbine couldn't readily be cost-justified, and a deal would only work if NIPSCO would sacrifice some of its short-term revenues by lowering unit pricing. Eads remembers days when he would talk to both vendors at once, almost playing one against the other. "One part of the gas company kept asking us, 'Hey, how soon are you going to give us a go-ahead on this turbine project?' And then we would call the other part of the gas company and ask, 'When are you going to hit our target price so we can cost-justify this for next year? Let us know,'" he said, "'and we'll sign the contract.'"

Also, the would-be turbine seller, NET, really wanted this particular sale, envisioning that it would serve as a groundbreaking departure from the usual microturbine applications found everywhere else in this market (i.e., in commercial, residential, and light industrial uses). As a sort of pioneering heavy industrial project, a Stripco microturbine might well become the model for other big power-gobbling industries to learn from and emulate. A whole new niche might blossom.

For its part, gas-seller NIPSCO wouldn't gain as much for its price-cutting, but several advantages might accrue to make this a long-term win/win deal. First, by dropping its per-therm pricing 30% compared with 2001 pricing, NIPSCO would be able to sell Stripco about 20–25% more natural gas a year to fuel each microturbine. The net effect would come out to rather comparable revenues before and after. Second, NIPSCO would gain a bit of volume leverage in its own commodity purchasing. Third, by helping Stripco lower its costs of production, NIPSCO would be making this customer more competitive and thus better able to increase its market share. "In the big picture," observes Eads, "NIPSCO knew that as we were continuing to expand our business, they were in essence ending up improving their revenues overall. They call this an 'alliance partnership,' where we are all working together rather than working in an adversarial relationship." Finally by cooperating on this deal, both Stripco and NIPSCO would benefit by flattening the huge swings between winter and summer gas deliveries, which were negatively impacting both firms. In signing the deal, Stripco's baseline gas usage would rise to at least 25,000 therms a month, up from 18,000, providing a steadier, more predictable volume.

Eads liked another aspect of this leveling. "It's basically a month-by-month and consistent bill now," he says. "Nobody comes by anymore and says, 'I can't believe how much this bill is!' Now, we basically know at the beginning of the year how much we need to budget for." In the new arrangement, Stripco would be able to buy about 85% of its total natural-gas demand at a negotiated, fixed annual price; the remaining 15% would be billed at seasonally impacted rates. Instead of the winter months' bills often doubling, the increase would only hit about 30%.

Next: Power Pricing
Such advantages, taken together, would almost—but not quite—cost-justify a turbine investment. More sharpening of the pencils was needed. Eads now had to return to his electric billings from American Electric Power, where—again—he was buying very low-cost power. Stripco's baseline demand was about 1.8 MW; what would be the impact of only 60 kW of microturbine? Eads now realized that the seemingly modest 60-kW output from a microturbine could be applied against the peak charges, which—again—were about double the basic charge of just more than $0.03/kWh. In fact, the turbine power could be turned on any time the plant was running up the light bill—specifically to mitigate against the peak demand rate—and then turned off at any other time, when its output might not be fully needed.

This arithmetic helped the cause; now the question involved precisely where to use the microturbine's cogenerated power and heat output. Here in one or more productive applications, Eads would need to calculate the potential savings and gains more specifically.

A likely-looking target was a 6,000-gal. tank of oil used in the steel rolling, tempering, and edging process. This required a modest but fairly constant heat input, especially in the winter. Currently heat was provided by 100-kW electric resistance coils—largely billable at peak demand rates. If 60 kW of microturbine-supplied power could be spliced into these coils, Stripco might succeed in shaving that amount from its grid demand.

But what about that remaining 40 kW? Eads wondered if the turbine's cogenerated exhaust heat might somehow be channeled in to provide this portion in a kind of hybrid combination of sources. Zdyb informed Eads that indeed the Capstone 60-kW microturbine came equipped with unusually high-efficiency heat exchangers able to capture and use about 80% of the exhaust. This could warm a water loop to circulate into the oil tanks, yielding plenty of continuous heat. In sum, said Zdyb, both the 60 kW of electric resistance heating and the remaining 40 kW of thermal heat could be supplied from one 60-kW CHP source.

In reality, reaching that conclusion required Eads and Zdyb to do some in-depth analysis of precisely how much heat the turbine could convert into thermal energy for the water loop. A secondary refinement to be answered was this: What portion of the turbine heat was actually needed to warm the oil in summer versus in winter?

Determining these and other such calculations also proved a valuable discipline because the exercise enabled Eads to do more precise heat-balancing and fine-tuning plantwide.

After coming up with these numbers, it turned out the 6,000-gal. oil tanks didn't quite need all of the turbine heat. Perhaps 10% of the exhaust might be usable for warming employee workspaces in the winter and heating their tap water.

Thus, adding up all of the CHP energy gains, the net reduction from Stripco's 1,300-kW baseline demand was projected to be about 12%, translating into a couple thousand dollars lopped off of each monthly electric bill.

Lastly, what is the cost of the turbine itself? The Capstone C-60 Microturbine model carries a sticker price of about $90,000. Installation—consisting of pipes, circuitry, breakers, fuel compressors, meters, temperature gauges, and concrete slab—would bring the total to $180,000. Also, in anticipation that a second turbine might be added someday, fixtures and a pad were prepositioned for it. Underwriting the total cost, too, was a $30,000 grant from the Indiana Department of Commerce. Stripco's net outlay was $150,000.

In order to maximize on first-year benefits, which begin in the colder months, installation was set for late autumn 2002 and was completed smoothly; the system became fully operational in January 2003.

Year-One Results
As the figures in 2003 began coming in, Eads found that, just as had been anticipated, electric bills have been consistently down about 10–15% each month. They shot up again mid-year, though, due to production demands; business was booming, and Stripco scheduled an additional 40 hours of weekly production. In August, the power bill shot up 15%. "The boss asked me, 'Hey, what happened?''' Eads recalls. A quick analysis shows that the plant simply used electricity more proportionate to its increased production.

Annual maintenance expenses have turned out to be negligible so far. Eads was pleased to find that, when the turbine reached its scheduled 8,000-hour inspection in late summer 2003, the bill came to just $300 for some new filters and a valve.

Combining all of the first-year advantages—including lower fuel rates and cogen power savings—Stripco is envisioning a first-year savings of about $55,000, according to Project Engineer Richard Smith.

As 2003 wound down, Stripco's price guarantee with NIPSCO was expiring. Pricing for the coming 12 months was upped to $0.60/therm. That's 20% higher but still about $100,000 less than Eads thinks Stripco would be paying next year if it remained under the former customer classification.

Stripco's natural-gas consumption also is increasing rapidly. In 2001, it stood at 255,000 therms; in 2002, prior to the turbine's installation, this soared to 290,000 therms (per-therm prices collapsed that year to an average of $0.46, a one-time fluke); and in 2003—reflecting the first full year of turbine operation—gas use was projected to reach 340,000 therms, 65,000–75,000 of which will be for the turbine.

Besides benefiting from the deal's original advantages, Stripco has gained from several other incidental facets that emerged rather unexpectedly. These are helping to "re-cost-justify" the project on a new set of grounds. Among these gains were the following:

• The discovery that turbine exhaust heat alone could provide 80–90% of the energy needed to replace that 100-kW resistance heater, about double what Eads had hoped
• The ability to apply CHP to preheat oil on its way to waste disposal, thus saving on the energy needed to pump it and making the cleanup of disposal tanks require only one hour instead of four, with the CHP system also heating the water needed to clean the tanks
• Improvement of oil-temperature consistency throughout the rolling process, even on cold days

"The finished product comes out looking cleaner and shinier," Zdyb notes. "The quality of their delivered product is better."

About five months into 2003 came the discovery of a malfunctioning oil thermometer, which had been causing systematic quality problems even before the CHP installation. As Eads was calibrating the CHP output, he detected the broken device, replaced it, and then upgraded the microturbine's heat exchanger to a brand-new design. (Capstone Microturbine began offering this in mid-2003). This resulted in a nearly 150% improvement in heat-capture efficiency.

Most critical of all, during the turbine's first year, local storms knocked out the neighborhood grid power several times. The "Great Blackout" of August 2003 also impacted some nearby Indiana communities. These experiences graphically brought to the fore the potential value of a turbine as a backup power source for several critical plant functions. Turbulent storms and outages are common in the region. Although one small turbine can't power very much in a steel plant, by running a line to the administration building 200 ft. away, Eads discovered that 60 kW easily can keep several key functions going there, such as sales, accounting, engineering, and quality. Stripco customers keep inbound sales lines busy every day; just a few hours' blackout easily could add up to thousands of dollars in lost revenues. So the next time the grid goes down for an hour or more, Eads plans to switch the administration building to turbine power. Meanwhile there still will be enough heat output to keep the production oil from cooling until the grid power returns.

None of these advantages was even considered when the turbine cogen concept first arose, Eads point out. During the first year, however, "especially the backup power factor has evolved into a major consideration," he says.

Operationally the first year helped Eads determine more precisely when shutting down the turbine is more cost-effective than running it (as, for instance, it is in summer). The critical threshold turns out to be the temperature in that 6,000-gal. oil heater. If ambient heat is enough to keep it warm, the turbine stays off or runs at a reduced 40-kW output. In winter, the turbine usually runs full throttle. "We can use all the extra heat we have," he reports.

Future Power and Projects
Currently on the energy horizon, Eads is exploring ways to capture spent hydrogen, which Stripco uses as atmosphere inside its steel furnace. He wants to burn hydrogen for heat or even use it to fuel an external-combustion engine. This reportedly can burn waste or "dirty" hydrogen to generate electricity. Eads also recently learned that Cummins Diesel offers models doing the same, and other engines can even reuse heat from a building's cooling evaporation system. "In the summertime," he predicts optimistically, although this still is being explored, "eventually I'll probably have enough hydrogen to get all the heat I need. And I won't have to buy any natural gas." If gas prices keep rising as Eads expects them to, he might consider generating his own hydrogen as a supplemental or replacement fuel.

As for the second turbine, it's on indefinite hold until an identifiable need arises. The trigger point will probably come from customer demand for some new processing capacity that requires adding a subsystem. As always, the constraining factor for cogen power at Stripco is that low, low grid-base rate.

In about four years' time, at the 40,000-hour milestone, the first turbine will be due for a $20,000 rebuilt main drive replacement. This adds up to $0.50/hr. of running time, Eads notes. Any number of factors might come along in the meantime, however, to postpone or hasten that expenditure. In any case, he says, "we're factoring that into our long-range energy management" and utility-costing studies.

Zdyb offers a final thought about the Stripco project, noting that because the company is an industrial customer, "they're relatively sophisticated about purchasing their power, and they also have a good engineering staff." Both factors made it much easier for Zdyb simply to hand Stripco the turbine specifications and performance numbers and let the company make its own cost-benefit calculations and utilization plans. Such competency tends to contrast with more typical CHP applications in retail marts and commercial establishments, which need more handholding.

"We had a lot of conversation with Stripco," Zdyb recalls, "but what I liked best about this project is that as distributed generation is emerging into the industrial market, this project demonstrates that a customer—if it's got the right resources—is capable of determining whether or not distributed generation is a good fit. Stripco was an early adopter. They were open-minded and creative—and [distributed generation] worked."

La Mesa, CA–based writer DAVID ENGLE specializes in construction-related topics.

DE - Jan/Feb 2004