Through the use of plant audits and sophisticated modeling tools, power plant operators can enhance cooling system performance and reap huge rewards in terms of increased generation, reduced fuel costs and reduced emissions.
Today’s power plant operators face increasing pressure to lower fuel consumption and maximize total megawatts delivered to the grid, while continuing to meet increasingly stringent environmental regulations. One of the most undervalued and misunderstood components in modern generating stations is the cooling tower and related systems. While basic steam turbine theory teaches that steam turbine efficiency improves with lower backpressure, few plants have truly taken an aggressive approach to maintaining minimum condenser pressure throughout the year. Those plants that apply today’s sophisticated analysis and modeling tools can reap huge rewards in the form of increased generation, reduced fuel costs and reduced emissions.
Blaming Mother Nature
The steam plant’s ultimate heat sink is the atmosphere. Since plant operators can become accustomed to the expected fluctuations in cooling system temperatures associated with normal weather patterns, the tendency is to be lulled into a sense of acceptance. High condenser backpressures experienced on a hot August afternoon are attributed to the current heat wave – there’s nothing plant operators can do about it. The original system was probably designed to operate properly only 98 percent of the time, so momentary high backpressures are an expected condition and are not viewed as a material deficiency.
With today’s operators comfortably accepting this simple fact of life, abnormal system operations easily get overlooked for years before they are discovered, costing generating stations millions of dollars in the process. The effect is even more dramatic if the original system was designed based on 15 to 20 year old market assumptions or the antiquated methods of choosing and evaluating design points. Additionally, recent developments in heat transfer technology have rendered many original installations obsolete. Even when plant operators recognize this fact, and evaluate the benefit of system enhancements, the use of antiquated design analysis techniques cause many projects to get stuck on the drawing board, unable to meet stringent capital hurdle or return rates.
Discrete Calculations vs. Statistical Analysis
Traditional plant operation models utilize a discrete set of operating conditions and parameters to perform financial calculations. Analysts may choose a “best case,” “worst case” or even “average” set of atmospheric and operating conditions. From there, an impact on operations is calculated, which is then used as a basis for financial decisions. Eventually, financial risk managers may analyze the volatility of the results based on the variability and fluctuation of factors such as fuel costs, power prices and dispatch schedules. While this analytical method has been standard since the days of slide rules and pocket protectors, it can easily mask the true value of an improvement and lead to inappropriate delays in plant improvements. Since the operating temperatures of the cooling system are highly dependent on ambient weather conditions, plants that operate in environments that experience hourly, daily or seasonal weather fluctuations can only be properly modeled if the entire range of ambient conditions is modeled. Modern computers and standard software packages enable analysts to fully simulate a plant’s operations with a high-resolution model that is not encumbered by such shortcuts as weather data averaging and single design point analyses.
Mirant’s Chalk Point Generating Station used rigorous statistical analysis to justify the cross-flow to counter-flow conversion of its Unit 4 natural draft cooling tower.
Another all too common problem with discrete calculations is the inevitable and seemingly endless debate over the “validity” of the single set of parameters used in the calculations. Fact is, a plant may have an average annual wet bulb temperature of 70 F, but the plant may only actually experience a 70 F wet bulb 1 percent of the year, meaning that the other 99 percent of the operating hours, the plant is in a different condition, experiencing a different set of financial impacts. The natural response is to make decisions based on some worst case or near worst case, just to be conservative or, as they say, “Err on the side conservatism.” Certainly this was an adequate approach when more sophisticated analysis tools were unavailable, but now that the technology is there, the question becomes “why err at all?”
Statistical system modeling and simulations can provide a much more complete prediction of actual financial results. Instead of arguing over whether a particular improvement has a $350,000/year financial benefit or a $500,000/year financial benefit, and haggling about what will be used as the worst case, why not just determine that the project has a 90 percent probability of yielding a net value of between $375,000 and $515,000 each year? If the spread is unacceptably wide at this point, basic sensitivity analysis can quickly home in on the key variables and help determine which risk mitigation or hedge strategies are available to lower the risk, or unpredictability, of the project.
Selecting the Real “design point”
By evaluating cooling system improvements through statistical plant analysis over full-year operating conditions, a whole new level of resolution and clarity is available. But the question remains: what is the “best” installation? This decision is ultimately driven by the financial metrics. For example, in a recent plant audit, four different system design changes were evaluated, along with multiple transaction structures for funding the projects (capital purchase, leasing, cold water outsourcing). These projects were evaluated based on a number of different standard financial hurdle metrics: simple payback, internal rate of return (7, 10, and 15 years), project net present value (7 and 10 years) and economic value added (EVA) of 7 and 10 years. In practice, the project that gave the shortest payback underperformed the others in terms of EVA and long-term IRR (Figure 1). Likewise, the project with the highest IRR did not yield the best simple payback. The design and size of a particular cooling tower can only be truly optimized if the financial metrics, weather variations and market variations are fully incorporated into the design and selection models. Unfortunately, tradition and the old way of thinking usually prevail, and project design and evaluation methodologies seldom incorporate this level of statistical discipline and rigor. While the traditional and statistically based evaluations may sometimes yield the same solutions, there are a number of common situations where an order of magnitude difference in calculated results exists.
Consider, for example, the correlation between ambient temperature, turbine output capacity and market power prices. Many generating stations experience situations where maximum rated megawatts cannot be produced due to high condenser backpressure. Assume that the original design engineers selected a 98 percent design point for the cooling tower. This means that, in an average year, the system will be exposed to excessive temperature conditions 2 percent of the time. Based on an average power price of $35/MWh, the expected financial impact of a 1 MW derate due to this design compromise would be approximately $6,000 a year. However, if one considers the concurrent weather impact on power prices, this number can be significantly higher. Since peak power prices are highest during the hottest hours of the year, the real value of the derated MWh could be closer to $500/MWh, resulting in an annual impact of more than $87,000 for each derated MW. By ignoring the simultaneous variance of weather, plant temperatures and market prices, many potential projects become grossly undervalued and fail to clear capital hurdle rates. As a result, the projects are delayed, or cancelled altogether, costing operators millions in lost revenues and excessive costs.
Chalk Point Station
Mirant Mid-Atlantic’s Chalk Point Generating Station recently utilized the full power of statistical project modeling and assessment. Standard payback analysis readily justified the Unit 3 tower conversion from a cross-flow to a counter-flow configuration in 2002. This conversion, from splash fill to higher efficiency low clog film fill, resulted in an average temperature reduction of better than 7 F. The productivity impact for the 640 MW generating station was more than adequate to justify conversion of the identical unit 4 tower. However, tightened capital budgets brought about by Mirant’s 2004 Chapter 11 reorganization delayed the planned Unit 4 conversion.
For Chalk Point, the operating variables surrounding the plant were enormous. As a peaking plant in the Washington DC area, the predicted dispatch schedule is heavily dependent on the spread between market power prices and the price of the two fuels that the plant can consume (natural gas and fuel oil). In addition to the complex financial dynamics, the installed natural draft towers display more operating temperature variances based on relative humidity as well as wet bulb.
By building a sophisticated hourly operations model, Mirant was able to complete a detailed risk assessment that demonstrated the probabilistic financial impact on the plant and the business. Additionally, this approach allowed the customer to evaluate multiple project finance options designed to conserve short-term capital and distribute some operating risks. The rigor and discipline of this exercise resulted in project approval, avoiding a potential costly delay in upgrading the cooling tower.
Side of the Fence
Although the power industry has experienced significant change over the last 10 years, many plant operators embrace old-fashioned methodologies and operating cultures. As a result, some improvements in operating efficiencies that were expected to evolve from wholesale market deregulation have been sluggish in developing. The plant operating companies that embrace new technology and advance modeling methods will soon edge forward from the pack, with more consistent improvements in revenues, operating profits and environmental compliance. While the Enron debacle of the early 21st century taught the industry the true benefits of solid business and financial fundamentals, those who fail to evolve in their approach to plant efficiency, optimization and design, may soon find themselves slipping behind the industry. p
Tom Dendy currently serves as Director, Market Solutions for SPX Cooling Technologies, a division of SPX Thermal. In this role, he oversees Marley’s customer productivity and marketing team, coordinating these efforts with product development and research activities. Dendy holds a degree in chemical engineering from Tulane University in New Orleans and spent 11 years in the United States Navy as a Submarine Officer.
By: Brian K. Schimmoller, Managing Editor
The list of issues that developers face in siting new power plants is quite long. Among the more significant siting issues are transmission access, fuel supply, environmental impacts, local support/opposition, tax treatment, and regional market strength. Joining this list in recent years – and at a prominent level – is water availability. Local and state governments, public agencies and various citizen groups across the country are getting increasingly protective of water rights, and as the second largest U.S. consumer of water (behind only agriculture), the power industry faces significant challenges in the years to come.
The challenge is not just in the future, however. Many plant owners are already coping with significant problems related to water availability. Low water levels on the Missouri River have forced several power plants to curtail production in recent years. One Kansas City area power plant even had to convert from a once-through cooling system to a closed-loop system when river levels began dropping below the level of the plant intake suctions. The Mohave power plant in Arizona faces imminent shutdown, primarily due to a dispute over water availability from a local aquifer. Even “wet” states like Florida and Illinois are predicting water shortages in coming years due to growing demand. In Florida, population growth is stressing a reservoir system that is badly depleted and becoming briny with saltwater seepage. In the Chicago metropolitan area, a report by a regional planning commission determined that parts of six counties bordering one of the world’s largest freshwater sources, Lake Michigan, could face serious water shortages in 20 years. As stories such as these permeate public awareness, pressure will mount on the power industry to slash water consumption.
Dry cooling is certainly one option, and air-cooled condensers (ACC) have dramatically increased in numbers in the United States. But air-cooled condensers are significantly more expensive, require substantially more space, and result in less efficient facility operation than a wet cooling system designed for a comparable cooling load. Many power companies, in fact, have commissioned studies of air-cooled condensers to quantify the impacts on busbar power generation costs relative to wet cooling. On the other hand, conventional wet cooling systems continuously evaporate water, consuming huge amounts of water each day and requiring extensive water treatment.
To head off a potential wet/dry showdown at the OK Water Corral, cooling companies are pursuing other options that can reduce power plant water consumption. Parallel condensing, for example, pairs two proven technologies: a conventional surface condenser/cooling tower system and an ACC. The amount of steam condensed in each system depends on the overall heat rejection requirements, the amount of available makeup water, and the economic conditions for the specific installation. When insufficient water is available, or when ambient conditions make plume formation likely, the ACC can assume the plant’s entire cooling load. On hot days, the wet condenser/cooling tower system can assume more of the load to reduce the performance penalties associated with ACC operation at high temperatures.
Figure 2 illustrates the middle ground that parallel condensing can occupy between wet and dry condensing. Plant output can be maintained at near capacity with only a minor heat rate penalty. More importantly, water consumption can be reduced by up to 50 percent, according to Steve Rottinghaus, Associate Development Engineer with Burns & McDonnell, which has seen a steady increase in the number of client requests for parallel condensing studies in the past five to six years.
Several plants under development are considering parallel condensing systems to reduce water consumption, including the 750 MW supercritical coal unit Xcel Energy is adding at its Comanche Station in Colorado. “The main issue in designing a parallel condensing system is how big to make each component,” said Rottinghaus. “A facility can accomplish the same annual water consumption with a large ACC and small cooling tower or a small ACC and large cooling tower, depending on how efficient the plant is and how the plant is operated.” Selecting the optimal size for each component, therefore, depends on many factors, including water availability, desired water consumption, delivered water costs, water treatment costs, regional energy prices, fuel costs, and the owner’s cost of money.
Parallel condensing offers plant owners considerable operating flexibility. Heat duty can be shifted from the cooling tower to the ACC units by reducing cooling tower fan speed while sacrificing steam turbine backpressure, or the ACC fans can be slowed down or shut off to conserve auxiliary load and increase plant output, but at the expense of additional water consumption. Burns & McDonnell recently performed an economic comparison of several cooling configurations for a new 600 MW brownfield coal plant permitted for zero liquid discharge using the following parameters: 30-year life, 7 percent discount rate (cost of money), 85 percent capacity factor, $1.25/MMBtu fuel costs, $100/acre-ft/yr water costs, $0.20/1,000 gallons water treatment costs, $7/kW-month replacement capacity costs, and $51/MWh replacement energy costs. For a wet cooling system, the busbar cost came to $36.40/MWh. For the parallel condensing system, the busbar cost came to $37.20/MWh, a 2.2 percent increase. For the dry cooling system, the busbar cost came to $38.80/MWh, a 6.6 percent increase. The bulk of the higher cost for the dry system could be attributed to the cost of replacement power.
SPX Technologies is commercializing a new technology called the Air2Air Water Conservation Unit that offers another route to reduced water consumption. “The A2A unit is an additional heat exchanger mounted in the upper portion of a standard cooling tower that recondenses and returns up to 30 percent of the evaporated water to the cooling circuit,” said Ken Mortensen, manager, environmental programs, with SPX Cooling Technologies. “Compared to a standard tower, we have to install a larger structure and a series of patented fill packs. There is also slightly higher fan power usage to draw in the additional air to recondense the water.” Based on field studies conducted by SPX, the average power usage may be in the range of 113-300 MWh per MCF (million cubic feet) of recovered water.
When incorporated in new tower designs, the addition of the A2A unit may add 10 to 50 percent to the installation cost of the tower. “With the added ability to collect the recovered, pure water in a side stream, the A2A unit costing stacks up incredibly well compared to other pure water production units like reverse osmosis,” said Mortensen. “It is also an environmentally responsible solution in that it allows plant operators to utilize waste heat from the plant to generate pure water.” SPX has completed validation testing of the unit at its Kansas City research and development center and is planning the first scale field installation, on a heavily instrumented tower, at an operating power plant later this year.
Siemens Westinghouse is also developing a water recovery technology, called WETEX (Water Extraction from Turbine Exhaust). As reported in the November 2003 issue of Power Engineering, this technology aims to use a liquid desiccant to remove water vapor from combustion flue gases. “We have just completed an 18-month research program that included a pilot scale test of a WETEX system that extracted water from representative flue gases for both combined-cycle plants and coal plants,” said Phil Deen, thermal cycle engineering manager with Siemens Westinghouse. “Based on the better than expected results, we are confident that WETEX, when used in conjunction with dry cooling/heat sink technologies, will be fully capable of producing enough water to cover all of a plant’s cycle make-up needs.” The estimated cost to produce water of reverse osmosis quality is $10 to $15 per thousand gallons, which includes amortized capital costs, plant performance impacts and O&M costs.
Siemens Westinghouse and the Energy and Enviornmental Research Center at the University of North Dakota have partnered in this development and are currently preparing for the second R&D phase, which will include refined parametric testing on a pilot scale level. Once the initial parametric testing is conducted and the results evaluated, the pilot scale system will be modified for efficiency and operational improvements and subsequently will be used in a long-term operation study. The long-term operational study will provide the technical foundation for a proposed third phase of testing in a scaled commercial application. Siemens Westinghouse is currently qualifying potential industrial partners for participation in this commercial demonstration phase.
Niagara Blower also offers an alternative to air-cooled technology. Its Wet Surface Air Cooler (WSAC) has been used successfully at simple-cycle and combined-cycle plants, coal gasification facilities and bio-fuel plants for various applications: water cooling, wastewater cooling, aux loop cooling, vacuum steam condensing, boiler blowdown cooling, gas turbine inlet air cooling and vapor condensing.
In a WSAC, the fluid/vapor to be cooled or condensed flows through tube bundles as part of a closed-loop system, and heat is rejected by means of latent (evaporative) heat transfer. A significant advantage is that makeup water can come from alternate sources such as open tower blowdown, reverse osmosis discharge, condensate, produced water, pond water, gray water or sewage effluent, according to Peter Demakos, president of Niagara Blower. Additionally, because the water is only used to wet the exterior of the tube surface, higher cycles of concentration can be achieved.
In a water conservation scenario implemented at a 525 MW combined-cycle plant in the Southeast, the WSAC is being used to cool the aux loop while using existing tower blowdown as makeup. If the plant operates 5,000 hours over the course of a given year, this configuration will save 35 to 40 million gallons/yr compared to a conventional cooling tower configuration.
For water-limited applications, when not enough water is available to use evaporative cooling for the entire load, a hybrid unit incorporating a dry (finned section) and a wet section can be used. The wet-surface air cooler can be designed to operate either wet or dry, further reducing the need for makeup water. Compared to a dry air-cooled condenser system, the WSAC offers lower installed and maintenance costs, smaller footprint (about 70 percent less ground space), lower parasitic energy use (about 60 percent less), and lower fluid outlet temperatures (as low as 95 F at 110 F ambient).