By: Steve Blankinship, Associate Editor
Few people expect a repeat of the natural gas plant construction boom that started about 1990 and came to a somewhat abrupt halt shortly after 2000. However, a combination of circumstances is spurring a different kind of gas-fired capacity boom in the United States. Natural gas units, almost the only kind of power plant built in the United States for a decade, remain the option of least resistance for many power providers trying to stay ahead of narrowing reserve margins and higher demand.
Natural gas combined-cycle (NGCC) plants designed for intermediate to base load service and simple-cycle (peaking) plants will soon sprout up from Florida (where regulators recently rejected requests to build new coal plants) to California (where public policy bans buying power from coal plants even when it is produced in other states).
A worker inspects the opened half shell of a GE 9FA gas turbine. Photo courtesy GE Energy.
The latest wave of gas-fired plant construction and operation will differ from the last construction push in several key ways. Among the differences: most builders this time won’t be merchant developers and independent power producers (IPPs) out to make what some thought would be a quick return on investment in newly deregulated power markets. Many of those players are now out of business and a few more are only now climbing back in after protracted financial restructurings. Those players failed in part due to a lack of discipline when it came to matching power plants with markets. Equally flawed were basic plant design and sizing decisions.
“Some big merchants affiliated with utilities were known to just plop plants down anywhere in the known universe even if they didn’t make economic sense,” said Jeff Schroeter, managing director of Genova Power Solutions. Schroeter has been involved with developing more than 8,000 MW of gas, coal and wind capacity in the U.S., most recently a 550 MW NGCC plant in Hobbs, N.M. scheduled for service next summer. Xcel Energy has contracted for all of the project’s output for 25 years.
This time around, most gas capacity builders will be traditional regulated utilities, many of which tried and failed in their pursuit of new coal-fired capacity and now have few options available to meet demand and reserve requirements. And unlike the IPP/merchant plants that typified the last natural gas boom, plant design and construction this time will be under the scrutiny of state utility regulators and their staffs.What’s more, a significant amount of construction will involve repowering brownfield coal sites with NGCCs.
“There is a tremendous potential for repowering in the U.S.,” said Ian Lutes, director of commercial development for Alstom’s heat recovery steam generator (HRSG) business. Two of Alstom’s most recent repowering projects are Tampa Electric’s Bayside coal plant (seven units totaling 1,700 MW) and We Energies’ Port Washington plant (four units that can achieve up to 1,200 MW in peaking mode). “Existing sites have cooling, infrastructure and community acceptance. We have lots of locations for power plants that don’t have to be greenfield,” Lutes said. “We have a lot of 40- and 50-year-old oil and coal-fired plants that really need to be repowered.”
Surge Won’t Be Modest
Although likely to be more measured and better focused than the previous boom, most signs suggest the coming gas plant building spree will be significant – driven by immediate and, in some cases, almost desperate need.
“There is certainly a lot of gas turbine activity outside the United States, but I think overall the U.S. market is getting very close to having quite a few projects move forward,” said Rick Antos, director of product line marketing – 60 Hz – for Siemens Power Generation. “A couple of years ago there were a lot of coal plants being planned and many are being delayed or cancelled. And even the delayed ones could drift out and not happen.” Uncertainties about costs related to prospective greenhouse gas regulations have stalled them.
Antos said he sees stalled coal projects creating a window of opportunity for gas turbines. “That’s really the only product that can be put in to meet a short-term need,” he said. “Eventually nuclear can be a factor, but probably not in the near-term. And IGCC (integrated gasification combined cycle) and coal plants with carbon capture are not moving ahead that quickly. So the only thing that can be put in to meet the need is combined- or simple-cycle gas turbines.”
Although he doesn’t think there will be anything like the overbuilding in the late 1990s and early 2000s, Antos said he foresees a significant increase in activity in the next couple of years. “It will not be modest,” he said. “And it will be more than that as you look toward the 2010 to 2011 period. I think it will be a pretty substantial market.”
Gray Market Gone
The first gas boom’s abrupt end created a lot of stranded assets – equipment ordered by developers who never took delivery, cancelled orders or scrambled to sell gas turbines, HRSGs, steam turbines and other components they could no longer use.
“There was a lot of surplus equipment available for a couple of years after gas projects got cancelled,” said Antos. But aside from a few remaining assets “all of that has now been taken.” In some instances, developers are buying unfinished projects from others rather than starting from scratch, but there are “no gas turbines sitting around anymore,” Antos said.
Schroeter agreed. “What little was left all went away in the last 12 months,” he said. When Genova Power bought its Mitsubishi 501 Fs for Hobbs, three had been available that were originally intended for a project in Tennessee. Tenaska bought them and looked for a buyer. “We bought two of them and a month later an IPP in Texas bought the third one,” he said.
All of Mitsubishi’s “gray market” inventory of gas turbines has been sold said John Adams, senior vice president for Mitsubishi Power Systems Americas (MPS). Mitsubishi recently sold nine G Class gas turbines. “Last year was a good one,” he said. “This year looks better and next year looks better yet.” Mitsubishi bought back a lot of F Class units, refurbished them to bring them up to current market standards and resold them. “In some cases, owners sold them directly and we got involved in bringing them up to current standards.”
Mitsubishi’s latest orders include combined-cycle plants for Florida Power & Light and a second large – and as yet unnamed – utility in the south. FPL will install nine G Class units near Palm Beach consisting of three 3 X 1 configurations for a total of 3,600 MW scheduled for service between 2009 to 2011. “Gas-oriented utilities are putting in more gas capacity because they have no other option and they are putting in the most efficient technology possible, such as 3 X 1 G applications,” said Adams. “They will have the most efficient plant around: the first to run and the last to shut down and will displace less efficient gas plants.” The unnamed southern U. S utility’s new combined-cycle plant will be of a similar design and size.
In response to the expected demand growth, Mitsubishi is building a new turbine parts manufacturing facility in Orlando to build combustion turbine blade and vane components. “Critical gas turbine components from this new facility will directly support our Western Hemisphere turbine fleet and augment Mitsubishi’s worldwide new parts manufacturing capacity,” said David Walsh, senior vice president of MPS Service and Manufacturing.
Flex and Stretch
Along with greater prudence and more rigorous due diligence in siting and justifying proposed gas-fired power plants will come more demanding performance expectations. The new standards will call for greater operational flexibility than many project developers sought in the 1990s. New gas capacity projects must be more adaptive to real market conditions than those built a decade ago when most combined-cycle plants were built as baseload units.
“A major difference is the new focus on operating flexibility,” said Adams. That means the total plant design, from the gas turbine to the HRSG and the steam turbine. “All of it will be considered to improve overall plant flexibility,” Adams said. By that, he means being able to start the plant quickly, cycle it, turn it down or shut it off entirely at night and then start it quickly the next day. “We’ve done a lot of work to make our products more capable to operate reliably in peak, intermediate and base power markets,” he said.
When developers bought combined-cycle plants in the past, their pro-forma typically was based on running 8,000 hours a year. Today, operators are cycling their gas plants, which is an option unavailable to coal or nuclear plants. Even large plants “like Mitsubishi’s 501Gs”are cycled either daily or a couple of times a week. “We went back and focused on the plant’s design to accommodate that operating regime,” Adams said. “Owners are going to cycle them because a gas turbine is the only plant that can.”
Alstom’s Chief Engineer Tom Mastronarde offered a different perspective. He doesn’t see much difference in the operating profiles between plants built in the 1990s and those being built today.
“From a technical standpoint, most of the work Alstom did during the gas bubble was for 2 X 1 (and some 3 X 1) plants designed to cycle,” he said. “It was intended that they would cycle daily, shut down on weekends, perhaps operate continuously during certain seasons and run overnight for weeks at a time if needed.”
However, unlike the typical customer of the 1990s, Mastronarde said today’s customer plans to own the plant long-term. “The last time, there were a number of IPPs that wanted to build them (the plants) and sell to someone else. We see more understanding, knowledge, concern and insight into how equipment is spec’d and more interest in different ways to do a number of things. Owners will give more attention to high cycling ability and extreme flexibility than we saw in the 1998 to 2000 timeframe.”
The typical gas plant Alstom builds today is a 500 MW 2 X1 with supplemental firing (for meeting various peak demand conditions) and usually with selective catalytic reduction (SCR) and CO catalyst. That translates into a flexible plant designed for daily cycling and weekend shutdown. What’s more, it can operate in bypass mode without the steam turbine so that the startup times can be tuned for emissions compliance quickly without respect to the steam turbine’s precondition.
Antos of Siemens said that such a holistic approach starts with the gas turbine. “We’ve introduced a 10-minute start to our SGT6-5000F and can provide up to 150 MW in about 12 minutes.” The company claims this fast start capability significantly reduces both the startup fuel requirements and startup emissions. Siemens is targeting this approach to peaking and intermediate markets.
The nominal 200 MW SGT6 – 5000F gas turbine during production. Photo copyright Siemens Power Generation 2007.
“We’ve also stretched inspection intervals so owners don’t have to take the unit down as often,” Antos said. “We have pretty much doubled the number of starts you can do before you need an inspection.”
Today, a broad range of capabilities exist that are largely standard prerequisites, according to Antos. For one thing, machines must run between 50 percent load and base load consistently or operators just won’t be competitive. This is particularly true on an F-class machine, which tends to run a lot more cycles as demand dictates, Antos said. The strategy on a G-class is different as those machines tend to run more of a baseload profile.
Schroeter said gas plants turn down very well and are therefore well suited for following load. “Combined-cycles are great as intermediate plants,” he said. “You can turn them down to about 20 percent of full output. A nominal 550 MW project’s minimum load is 125 MW. In terms of bringing gas turbines up fast, it’s not the gas turbine that’s hard, it’s the HRSG because of the thermal stresses inside. They have a lot of mass.”
HRSGs-incorporated with the steam turbine, condensers and bypass systems- introduce complex issues when called upon to cycle. That makes HRSGs the focus of improved performance for combined-cycle configurations. Siemens has moved to the once-through Benson technology, which allows the gas turbine to achieve full load rapidly rather than sitting at a low load while the HRSG heats up. The Benson once-through technology is licensed to several boiler and HRSG suppliers and allows more aggressive startup times by eliminating thick-wall components like the high-pressure drum.
Using the technology, Siemens offers its “Fast Start Plant” startup scheme. After an unrestricted gas turbine ramp-up to rated power, the steam turbine is also rolled. The steam turbine synchronizes while the hot reheat and low-pressure system bypass valves are closed and the steam turbine accepts all generated steam. At that point the plant achieves about 97 percent of its rated output.
The startup scheme is about twice as fast as what its predecessor could achieve following overnight and weekend shutdowns. The once-through HRSG can be applied with or without supplementary firing and SCR/CO-catalyst systems. “That allows us to let the gas turbine go up to a very high load very quickly and get it in emissions compliance much faster,” Antos said. “That also helps the permitting and siting.”
Once-through HRSGs are now the standard, said Alstom’s Mastronarde. “This is the only way people will want to build a high-efficiency plant that cycles daily and we expect that market will grow.” Alstom’s once-through HRSG is based on Siemens-licensed Benson technology with “certain technical improvements” developed and patented by Alstom and marketed as OCC OT (Optimized for Cycling and Constructability-Once-Through). Mastronarde said that nearly all water/steam cycle technology advances have come from outside the U.S., which has a greater focus on high efficiency combined cycle and new steam cycle technology.
Various enhancements to gas turbine performance have long been available for retrofit and some are becoming standard in newer machines. Front end add-on technologies such as inlet cooling and fogging were widely marketed in the late 1990s as ways to enhance gas turbine performance. “You might consider adding evaporative cooling or chillers for additional capacity on hot days with high peak electricity prices,” said Mas Fukumoto, gas turbine marketing executive for GE Energy.
Such enhancements are site-dependant. One downside of dry cooling is that during very hot weather, air-cooled units can’t produce as much power as water-cooled units. Dry cooling is not a solution for hot areas such as the Texas Gulf Coast. Adding evaporate cooling has largely become standard on units expected to perform on hot days, although inlet cooling tends to get a bit more complicated still. “It’s costly and uses additional water,” said Schroeter. He said most turbines will have some form of cooling – mechanical chillers or wet cooling – depending on what’s appropriate for the climate.
“How much you want to enhance your peaking capabilities depends on the price of peak power in your region,” said Fukumoto. If it’s high, it may make sense to put on an inlet system. “You also have to consider the expected ambient conditions to decide which cooling system will work best,” he said.
Plants being built today would probably be cycling applications except in markets such as Florida, Texas or California where natural gas is on the margin and gas-fired plants run more hours than in the rest of the country. “But in the Midwest, where coal sets the market, a combined-cycle plant would be more expensive to dispatch than coal, so those plants are likely to cycle,” he said. Most customers ask for greater operating flexibility on the combined-cycle plants to allow for faster start-times and to allow for lower part-load operation while staying in emission compliance.
New gas plants will also have to run cleaner than their predecessors. “Our air permit at Hobbs was ratcheted down to 2 ppm NOX,” said Schroeter. Back in 1999 a few plants were still getting permitted at 9 ppm NOX and some at five. Today, 2 ppm is considered best available control technology, he said.
But the need to cycle runs at odds with lower emission level standards. That’s because emission control systems do not operate at maximum efficiency as units are started up or brought down. In the past, many air permits included waivers for startup emissions. New air permits, however, typically count emissions during startup and shutdown “so we have to reduce emissions as much as possible,” Antos said. “We want you to get to emissions compliance as soon as possible.”
Power plant control systems have also advanced significantly over the past 10 years. Many systems installed in the ‘90s were DOS-based. Today, all are Windows-based and far more networked. Instruments are “smarter” and many plants, for example, now have valves that provide instantaneous information on their position.
Another contrast to the conditions of a decade ago is the composition of pipeline fuel, primarily due to greater use of liquefied natural gas (LNG). “It hasn’t happened as fast as had been talked about,” said Antos, “but there will clearly be more LNG dumped into the pipelines as we move forward. LNG imported to the U.S. tends to have more ethane, butane and propane than typical natural gas. There are also significant variances in natural gas depending on the point of origin. Manufacturers have done considerable combustor testing to probe the possible impacts of fuel variations.
Siemens has expanded fuel spec ranges, said Antos. “Testing has shown we can accommodate a wider range than what our previous specs allowed. We believe we can accommodate just about any LNG that would be put in the U.S. delivery system. There will be some additional instrumentation required, but not a drastic change in the combustor system.”
Antos is referring to technology such as Siemens’ Integrated Fuel Gas Characterization (IFGC) system, which integrates the readings from dynamic tuning systems that monitor gas turbine operating conditions, fuel quality and emissions, then adjusts air-to-fuel ratios and operating parameters. The system allows the gas turbine to “self-tune” and adapt instantaneously while operating.
GE Energy offers its OpFlex Wide Wobbe technology, which provides customers the ability to operate across a range of natural gas-derived fuels of various compositions while maintaining performance, reliability and operating flexibility. Recent OpFlex offerings can be used where natural gas heating values fluctuate. Wide Wobbe automatically adjusts control parameters based on gas turbine performance and environmental conditions, allowing the unit to continue operating within emissions constraints.
Harry Morehead, manager of Siemens’ North American IGCC business development said that combustion testing on high hydrogen syngas for the carbon capture versions of IGCC have yielded retrofit options for existing 5000 Fs operating in places like Houston where combined-cycle plants are located near large supplies of petcoke. “There are some opportunities there,” he says.
Water Issues Everywhere
Even in traditionally non-arid areas, water has become a big issue. And in regions where water has long been precious, air-cooled condensers have become the norm instead of wet cooling towers. That’s because dry cooling uses as little as 10 percent of the power cycle water compared with wet cooling.
“We still need water for our steam cycle and for our chillers when we use them,” said Schroeter. “But people are willing to accept a slight reduction in power output in order to conserve water.” Had wet cooling been used at Hobbs the developer would have faced procuring 4,000 acre-feet of water. “We ended up needing about 450 acre-feet annually because we went with dry cooling,” Schroeter said.
The downside of dry cooling is that during very hot weather air-cooled units can’t produce as much power as they could if they were water-cooled.
Genova Power’s Hobbs plant generates 604 MW in the winter and 504 MW in the summer largely because the air is thicker in the winter, Schroeter said. As a result, he said a bit of a summertime tradeoff exists from a systemwide perspective in terms of greater emissions and more fuel consumption. “Still, dry cooling is more acceptable to generation planners, who will accept a summertime power output reduction,” he said. He added that dry cooling can boost capital costs by about 8 percent.
I’ll Give You 10 Minutes
Tighter reserve margins also lead to greater need for peaking capacity. As a result, regional transmission authorities and independent system operators now pay on a continuing basis for guaranteed capacity reserves. The standard is to be able to provide additional capacity within 10 minutes. Guaranteeing such spinning reserve has become a critical niche for gas-fired generation.
“Specifying the design of a new gas plant really gets interesting these days,” said Antos, “because you have to get really specific with what the output, heat rate, emissions and start-up time will be. And if you do it right, you can qualify a machine as a quick-start unit, which is an interesting new market.”
Siemens for one has spent a lot of time in the last year improving startup speed, both for the F and G units. “If you can get a 10-minute start, you are in the running to get ancillary payment that you wouldn’t otherwise get,” Antos said. But there’s a bit of a tradeoff and Siemens has used its test bed in Germany to validate the effect of fast starts on the SGT6-5000F to address improving cycling speed while maintaining equipment durability.
One machine built specifically to produce 100 MW in about the time it takes to brew a pot of coffee if GE’s LMS 100, an aeroderivitive combustion turbine that can go from cold to 100 MW in 10 minutes. Since its introduction two years ago, GE has received 18 orders for the LMS 100 – 13 of which are in the U.S. Firms in Canada have ordered two, Latin America two and one is headed for Europe. One of the U.S. buyers, Basin Electric (which serves power suppliers in the Midwest), had logged 500 hours and 150 starts with its machine.
Renewable energy mandates are also driving the need for quick startup from gas turbines. California’s requirement to achieve a 33 percent renewables standard with large amounts of wind and solar power will require an additional 1,700 MW of quick-start generation, according to a recent report by Exeter Associates. The report said California can achieve its goal, provided it spends $5.7 billion for additional transmission line segments and installs the needed additional quick-start peaking capacity. The report added that fast-start, load-following resources such as simple-cycle gas turbines could be needed for strategic dispatch to respond to varying wind conditions.
No Escaping Higher Costs
Although the next round of natural gas-fired plants will be better suited to their markets, they will also be costlier to build than those of a few years ago. A combined-cycle plant could have been built in 1998 for $400/kW and in 2005 for $550/kW, said Schroeter. Today that number is around $750 to $850/ kW. “That’s all due to labor, steel, commodities and the fact the OEMs are starting to get some price support again for their turbines,” Schroeter said. But nothing like in the boom days. “People paid a lot more for combustion turbines in 2001 when the boom was still going than you could buy a new one for today from the OEMs.”
Schroeter recalled paying about $27 million for F-class machines in 1998. By 2001, some people were paying as much as $42 million. Today, the price would be about $33 million. Another telling data point: Schroeter said he has been paying $130 a cubic yard for concrete and another $900 a cubic yard to place it. “That comes to $1,030 per yard to set concrete. Twenty-four months ago, the budget to place concrete was between $400 and $500 per cubic yard. Everybody is just busy.”
Busy, that is, building the latest wave of gas-fired generation.