By Brian K. Schimmoller,
As noted in this magazine in several recent issues, a groundswell of interest in new coal-fired facilities is emerging. The dramatic rise in natural gas prices has sent many utilities and developers scurrying to wipe the dust from the coal plant blueprints that never made it off the drawing board (or CAD screen). If the numerous rumors swirling around the industry about new coal plants materialize into some firm commitments to proceed, developers will then have an interesting technology choice to make: Should I select a conventional, conservative subcritical steam cycle, or should I take a little risk and opt for a more efficient supercritical design?
Supercritical boilers have had an interesting history in the U.S., dating to the mid-1950s. The popularity and growth of supercritical plants paralleled the vitality and growth of the U.S. economy and the power sector during the 1960s and 1970s. Many utilities viewed supercritical (SC) technology as part of the natural evolution toward larger, more efficient generating units. As economic conditions changed in the early and mid 1970s, however, and as more baseload nuclear plants came on-line, the industry reverted to the subcritical steam cycle due to the lower installed costs and proven operating experience. While installed SC capacity in the U.S. amounts to a sizeable 80+ GW, most of that came on-line prior to 1980.
The real and anecdotal stories about the difficulties associated with SC technology point to several causes:
- Limited unit cycling capability due to boiler valve wear, turbine thermal stress and turbine blade solid particle erosion;
- A complex and time-consuming start-up process;
- Lower availability compared to subcritical units; and
- Higher maintenance costs compared to subcritical units.
In hindsight, the typical SC unit appears to have suffered more from the rapid increase in unit size than from the high-pressure technology itself. During the development of SC technology in the 1960s, the average fossil unit increased in size from 247 MW to 500 MW; more than half had pressurized furnaces and more than one-fourth used double reheat cycles. While not unique to the SC steam cycle, the complications and lack of experience with these systems hampered unit performance.
Recent studies and long-term operational performance have largely erased concerns about SC plant reliability. In an internal report evaluating SC technology, ALSTOM Power cites three studies-from NERC in the U.S., VGB in Germany, and KEMA in the Netherlands-that report no significant difference in availability due to subcritical/supercritical steam parameters for modern plant designs (Figure 1). EPRI studies have reached the same conclusion. As another indication of the confidence plant operators have in SC technology, American Electric Power plans to operate its supercritical plants at a relatively higher capacity factor than currently operating, according to Manoj Guha, AEP manager, corporate technology development. Finally, the continued, sustained level of commercial activity in overseas markets lends validity to the claim that SC units can be operated as reliably as subcritical units.
Oak Creek is one of the power plant locations where Wisconsin Electric is considering the addition of a new coal unit. Photo courtesy of Wisconsin Electric.
Of course, SC units still have several inherent operational complexities. Because of their startup system, where boiler circulating water pumps are used to maintain mass flow in the furnace wall tubes to prevent overheating, once-through supercritical units have a greater potential for turbine water induction through the main steam system than do drum-type subcritical units. Similarly, because of the once-through design, SC units are more sensitive to feedwater quality. Full-flow condensate polishing, therefore, is required to protect the turbine from stress corrosion cracking.
Solid particle erosion (SPE) of turbine blading and vanes can be a significant maintenance and heat rate concern for SC units. While typically not causing forced outages, the presence of SPE is likely to result in extended maintenance outages. Some believe the start-up system design of older SC units may account for the SPE; newer units with integral separators included in the start-up system design are not expected to experience the same level of SPE.
These operational complexities point to a non-technical factor that must weigh heavily in a developer’s decision to pursue SC steam technology – operator experience. Experienced SC unit plant operators are not exactly a dime a dozen, and those that do have experience are likely reaching the end of their careers. The average power plant employee is about 48 years old, which means that plant owners would likely have to train a new generation of operators.
Despite the lack of domestic interest in the last two decades, boiler manufacturers/licensees have retained their expertise in supercritical steam technology by focusing on overseas markets. ALSTOM Power, Babcock & Wilcox, Babcock Borsig Power, Foster Wheeler, Mitsui Babcock, and Siemens/KWU have been able to promote and improve supercritical technology via the slow but steady stream of plant orders from Europe and Asia-Pacific. Now the wheel might be turning back toward North America again.
Although the lack of generating assets in certain parts of the U.S. appears to signal a call for new coal-fired capacity, it is important to understand the nature of this demand and the supply-side reaction to it. “In most areas of the country, there is not any current need for baseload power, and probably won’t be for another 10 years,” says AEP’s Guha. “Even in those isolated pockets where some baseload shortage may exist, electricity can be supplied from gas-fired combined-cycle plants at lower cost.”
While this may be true, the run-up in gas prices has initiated, or at least accelerated, planning efforts to develop coal-fired power plants-both subcritical and supercritical-in areas anticipating either baseload or load-following demand growth. Advanced planning is crucial. Siting, permitting and equipment sourcing for coal-fired power plants have long lead times that can’t be significantly shortened.
“We’re aware of a number of developers considering SC systems,” says Ian Lutes, Foster Wheeler senior vice president. “Washington Group is evaluating an 800 MW SC boiler for a Midwest client, Black & Veatch has conducted a technical review for the Lignite Energy Council in North Dakota of both drum and once-through boilers for a 750 MW unit, and a utility in Alberta is evaluating subcritical and SC options for a new 440 MW plant. These are not idle analyses, but represent serious, detailed budget price estimates.” Burns & McDonnell is currently working on 11 new coal-fired projects. These efforts include design, cost estimating, siting studies or permitting activities for projects across the U.S. Most of these are for additional units at existing facilities, and the majority of these are considering SC boilers.
Wisconsin Energy Corp. (WEC) has explicitly stated its interest in new coal-fired generation. As part of its “Power the Future” plan to provide reasonably priced electricity for the next two decades to Wisconsin residents, over the next 10 years WEC is poised to commit more than $2 billion to new power generation, a combination of coal, gas and renewables. A large portion of the investment will be for two new 600-700 MW coal-fired units. “Our decision to pursue coal-fired generation was not just a case where we saw natural gas prices increasing and identified an opportunity,” says Al Mihm, manager of fossil engineering with WEC. “To meet our state’s energy growth of 3 percent per year, reliably and cost-effectively, we believe we need to maintain a diverse generation mix.” While WEC is still evaluating various coal-fired technology options, pulverized coal supercritical technology is “floating to the top,” according to Mihm.
It is interesting to note how developers are using the CO2 issue to their advantage in promoting SC coal-fired plant designs. While obviously not as efficient as a comparably sized gas-fired combined cycle, a SC plant does reduce carbon emissions relative to the same size subcritical coal unit. WEC has affirmed its commitment to building coal-fired facilities that will be as “environmentally friendly as ever built.” Driving WEC’s interest in SC technology, therefore, are the benefits it provides in terms of efficiency and environmental performance. A SC plant on the WEC system would have an efficiency about three percentage points higher than a comparable subcritical design, thereby reducing carbon emissions, while also meeting all new source performance standards.
In Canada, EPCOR (Edmonton Power) is pursuing the addition of a third unit at its Genesee plant southwest of Edmonton to provide baseload power to a growing market. Using SC pulverized coal combustion technology, the 400 MW (net) unit would have CO2 emissions 8-10 percent lower than those from the existing units at Genesee, despite having a 6 percent higher capacity. Moreover, EPCOR’s goal is that net incremental greenhouse gas emissions from the new unit will be equal to or lower than those from a comparable natural gas-fired combined-cycle plant. How EPCOR will achieve this goal is unclear, particularly since modern combined-cycle power plants achieve efficiencies in the mid-50 percent range, substantially higher than even advanced coal-fired plants (Editor’s note: several attempts were made to contact EPCOR about this issue).
Concerns about noise, emissions, water use, visual aesthetics and many other tangible and intangible factors make power plant siting difficult, even for the new crop of gas-fired facilities. Greenfield siting, particularly for a coal-fired plant, is that much more difficult, prompting North American developers that are considering new coal-fired generation to look almost exclusively at existing stations with room to grow. WEC is evaluating two sites: Oak Creek, which had eight units for many years, but currently houses only four, and Pleasant Prairie, which currently operates two 600 MW units, but was originally designed to add one or two more. Physical space, therefore, is not a problem, and transmission, rail access, and water availability are also more than adequate. EPCOR followed similar logic in selecting Genesee. The existing cooling pond and coal-handling systems are already sized to accommodate the capacity addition, and the plant is adjacent to a coal mine with ample reserves.
Greenfield siting of SC power plants is not impossible, however, and actually affords some design flexibility. China and Korea’s new SC units are primarily at greenfield locations. And in Australia, InterGen is developing an 840 MW SC power plant that will serve as a merchant generator to feed power into the Queensland grid. The Milmerran Power Project will rely on two 420 MW Babcock & Wilcox universal pressure boilers to provide the operational flexibility and reliability necessary to succeed in an open market.
Spiral or Vertical?
The furnace wall enclosure design of SC units has gone through several iterations over the years, from the use of multiple pass designs to intermediate mixing headers. Modern designs are essentially limited to the spiral tube arrangement and the vertical tube arrangement. The spiral tube design has more than 30 years’ experience. The basic concept is to reduce the number of tubes required to envelop the furnace wall without increasing the spacing between the tubes. The primary drawback to the spiral tube design is the hardware needed to support the spiral tubes. The vertical tube design has a much shorter history, but is garnering much interest because of its simpler configuration and reduced pressure drop.
Since reliability is paramount in today’s energy market, fleet experience counts for something, and it’s particularly important to operators like WEC, which does not currently operate any SC units. “Spiral wound designs are a more proven technology than the newer vertical tube technology,” says Mihm. “While there are some vertical tube supercritical units in operation, few are on coal, so I’d need a lot more information before I’d take that step. The simplicity that a vertical design offers does warrant further consideration, however.” According to Mihm, the 600-700 MW units it is considering fit nicely in the generation mix at WEC, as well as being an off-the-shelf, proven size for recently constructed SC plants in Germany and Japan.
In China’s Henan province, the Yaomeng Power Station is taking the proverbial leap into vertical tube technology. The 300 MW Unit 1 is being retrofit with a once-through, low mass flux, vertical tubing design that Mitsui Babcock has been developing with Siemens/KWU for more than 15 years. The low mass flux design minimizes thermal transients by virtue of its positive flow characteristic, wherein a furnace tube absorbing greater than average heat will have an increase in flow. The lower overall pressure loss results in life-cycle savings in feed pump power compared with a spiral wound design, which requires an auxiliary support system.
Yaomeng opted for the SC retrofit because it represented a more cost-effective solution than replacing the 25-year old facility with a new plant. Post-retrofit, Unit 1 will see steam output increased from 1,030 ton/hr to 1,045 ton/hr, while maintaining steam pressure and temperature. Sliding pressure operation will provide improved part-load cycle efficiency, and a new firing system will improve combustion efficiency and reduce NOx emissions. The retrofit will be completed in 2002.
The efficiency advantage of the SC cycle over the subcritical cycle usually dictates that it remains baseloaded. The energy savings for a 600 MW unit, assuming an 80 percent capacity factor, is about 735,000 million kJ/year. If the fuel cost is $1.50/million kJ, this represents a fuel savings of $1.1 million per year. While SC units are capable of higher ramp rates than subcritical units (7-8 percent per minute vs. about 5 percent per minute between 50 and 100 percent load), most new SC boilers are expected to operate as baseloaded units due to the economic advantage of the SC cycle. Still, FW’s Lutes notes that the SC Taishan plant in China is designed to cycle down to 25-30 percent load.
SC units can even accommodate daily start-stop cycling in certain instances. Wisconsin Electric’s Mihm points to one plant trip that dramatically changed his outlook on supercritical technology. The Rostock plant in eastern Germany, a 500 MW SC unit that entered service in 1994, has been relegated primarily to peaking service because of cheap import power, but regularly cycles on and off each day. Remarkably, the unit has achieved an availability of more than 90 percent. One of ALSTOM Power’s SC boilers, Bexbach I in Germany, has averaged more than 100 start-ups per year since coming on-line in 1983.
Pay the Piper
Materials and equipment costs for single reheat SC boilers at 3500 psi are 3-5 percent higher than their subcritical boiler island counterparts, for a heat rate reduction of 5.0-7.5 percent. On an EPC basis, however, the costs are much closer. The supercritical boilers (2×660 MW) designed by Foster Wheeler for the Taishan plant in China had EPC costs only 1 percent higher than a similar subcritical design, and Lutes is confident the same small delta could be maintained in the U.S. Mihm agrees that the percentage difference in capital cost between SC and subcritical units is in the single digits, but maintains that the lower life-cycle costs for the SC units provide a strong attraction. This assumes, of course, that the SC unit achieves comparable availability and reliability records.
Double reheat, ultrasupercritical (USC) steam cycles at 4,500 psi/1000 F/1025 F/1050 F provide a larger efficiency hike, from about 37 percent to 44 percent, but the cost increase is not clearly defined because of uncertainties with the cost, performance and availability of the necessary materials. The next step is to USC steam cycles with pressures of 4,500-5,000 psi and temperatures of 1,150-1,200 F. The key to the economic viability of such plants lies in the availability of ferritic steels with high strength at 1,200 F so that the use of more expensive austenitic steels can be minimized. Current superalloys used for this type of application are too expensive and/or have not been demonstrated successfully at these elevated conditions for a sustained period, according to AEP’s Guha. With annual R&D investments of $18-20 million, however, it is estimated commercial materials could be available in 5-7 years.
“To the best of my knowledge, no one has precisely estimated the costs of an USC coal-fired power plant of this type simply because no one really knows the price of materials for this purpose,” says AEP’s Guha. “Based on a natural gas price of $3.00/MMBtu with 2 percent real inflation, and a coal price of $1.10/MMBtu with 0 percent inflation, a new coal-fired plant must cost less than $800/kW in today’s dollars. However, since natural gas prices have hit record highs in recent months, the economics change. In my opinion, though, it’s not the capital costs, but the uncertainty of environmental requirements that will dictate if anyone builds a new coal-fired plant. If EPA can come up with a systematic, integrated policy for addressing the emission requirements associated with NOx, mercury, hazardous air pollutants, CO2 and other criteria pollutants, it is possible to add new coal-fired capacity at competitive prices.” Guha added that AEP currently has no plans to add new coal-fired units at least until 2010.
Patience and Perseverance
No one is placing commercial orders yet for SC pressure boiler parts. Optimistically, the earliest WEC would put a new SC coal unit in operation is 2007, and EPCOR is looking at a 2005 commercial operation date for the new Genesee unit. Still, the time for planning is now. Siting, even at existing locations, is a process littered with potholes. Technology selection could make the difference between a unit that makes money in a carbon-constrained world and one that gets stuck on the drawing board. And operator training is taking on added significance as the workforce ages and retires.
Still, the prospects for SC technology, and coal in general, are good, and looking better with every uptick in natural gas prices. WEC is looking even further into the future to maintain its diverse generation mix. Mihm expects WEC to develop more coal (and gas) plants between 2010 and 2020, using an array of advanced technologies from combustion turbines to supercritical fluidized bed combustion (see sidebar) and IGCC. “We are anxiously monitoring the development of new coal technologies.”
The advantages of fluidized bed combustion technology are well known: fuel flexibility, low NOx emissions and readily controlled SO2 emissions. Similarly, the advantages of supercritical, once-through steam cycles are well-known: higher efficiency, lower carbon emissions, and reduced fuel costs. Integrating the two concepts has not been a high priority, particularly in the U.S., because the added efficiency doesn’t result in significant fuel savings since the fuel is typically low in cost anyway.
With a growing interest in using waste fuels, however, and the threat of tighter emission regulations, supercritical CFB units might have a future. Foster Wheeler and Siemens have developed a boiler configuration that integrates Siemens’ Benson once-through technology with Foster Wheeler’s compact CFB unit.
Critical to the CFB once-through unit (CFB OTU) design is the ability to accommodate heat absorption variations without overheating the furnace enclosure tubes. Historically, this has been done by designing for high fluid mass flow rates, using spiral tube configurations. The high mass flow rate, however, results in high pressure losses and increased auxiliary power consumption, and the spiral tube configuration is susceptible to erosion in a CFB environment. The Benson Vertical technology relies on a single vertical pass configuration (using optimized rifled tubes) to provide low mass flow rates and accommodate full variable pressure operation. In turn, because of its minimal use of refractory, the compact CFB can accommodate the rapid load changes associated with variable pressure operation.
Foster Wheeler and Siemens have developed a 350 MW conceptual CFB OTU that uses two double vortex solids separators to collect solids from the furnace, and four INTREX heat exchangers to cool the collected solids and maintain the desired furnace operating temperature. Start-up requires a minimum mass flow rate of 35-40 percent through the furnace walls. To accomplish this, a recirculation pump is needed to superimpose a recirculating flow onto the flow provided by the boiler feed pump.
Unique to the CFB OTU is the ability to accommodate process disturbances (fuel rate and quality) because of the stabilizing effect of the inventory and flywheel of circulating solids. Superheat versus evaporative duty distribution can be controlled by varying furnace enclosure, internal furnace heat transfer surface, and INTREX heat absorption characteristics.
- Burns & McDonnell internal report.
- Alstom Power internal report.
- Goidich, Stephen J., “Integration of the Benson Vertical OUT Technology and the Compact CFB Boiler,” Proceedings of POWER-GEN International 2000, Orlando, Fla., Nov. 14-16, 2000.
- Smith, J.W., “Babcock & Wilcox Company Supercritical (Once Through) Boiler Technology,” Technical paper published by Babcock & Wilcox, May 1998.
- Gorokhov, V. et al., “Supercritical Power Plants Hike Efficiency, Gain World Market Share,” Power Engineering, Vol. 103, No. 10, October 1999, pp. 36-42.
- “Genesee Generating Station Phase 3 Public Disclosure Document,” Submitted to Alberta Environment, Dec. 2000, available at www.epcor-group.com