Over the Past Few Years, High-pressure Fog Cooling Units for Power Augmentation have been successfully installed in hundreds of turbines throughout North America and around the world. Not surprisingly, major OEMs are now offering fog systems in their latest models. “Due to increased customer demand for fogging, we are including fog units in certain turbines,” said Ernie Smith, Siemens-Westinghouse’s Auxiliary Standards Group manager. “Within a few years, I expect to see around 25 percent of new turbines being shipped with a fog system.”
Photo 1. Fog system skid. Photos courtesy of Mee Industries.
With after-market retrofits in high demand, many plants will soon be ordering fog systems for the first time. Based on interviews with dozens of facilities, therefore, this article explains how to specify and install a high-pressure fog system.
Don’t Reinvent the Wheel
High-pressure fog has achieved maturity in the inlet air cooling field. As such, there is no need to reinvent the wheel for each new installation, yet some facilities have wasted many months attempting to do just that.
At one East Coast facility, for instance, a fog system project begun in early 1998 still isn’t finalized. In addition to the manufacturer, two groups of consulting engineers were asked for input. “The supplier lacked experience and too many design changes were made,” said the utility’s specifying engineer. “Having so many different opinions definitely slowed the whole process down.”
Instead of installing proven equipment, each step of the process was subjected to extensive R&D, with every part of the design questioned. The facility itself has failed to achieve any real benefit from the system, while at another of its sites, a more experienced company came in earlier this year and installed fog in 11 turbines within a few weeks, resulting in an immediate power boost and a rapid payback.
Focus on Results
Utilities are in the business of power production, not fog system design. It is unwise, therefore, to attempt to control all the peculiarities of fog system design. It’s more beneficial to clearly define the performance criteria to meet, and ask the fog companies to supply specifications for a system that can achieve these criteria. One Southeastern utility fell into the details trap, spending many hours specifying every single aspect of the system, testing each part and trying out all possible arrangements. By neglecting to define performance requirements in terms of MWs, however, the utility achieved poorer-than-expected returns and the project was turned over to a more reliable vendor.
Turnkey vs. Piecemeal
Figure 1. Fog system schematic. Courtesy of Mee Industries.
Modern fog systems come in attractive, pre-engineered, skid-mounted packages (Figure 1 and Photo 1). The alternative is to cobble together the various components, some from the fog system vendor, some from other vendors. In the majority of cases, the turnkey approach is recommended.
Photo 2. PLC system controller. Photo courtesy of Mee Industries.
The programmable logic controller (PLC), for example, is occasionally the subject of debate regarding the turnkey approach. The PLC is a computerized controller that translates weather data into fog system operating instructions (Photo 2). As air temperature and humidity are constantly shifting, the amount of water needed to saturate the air varies. Only by using a fairly complicated control algorithm is it possible to accurately monitor the optimum amount of fog needed to obtain maximum evaporative cooling without overshooting (usually termed overspraying or intercooling) or undershooting. Many turbine operators, however, purposely inject more water than will evaporate in the inlet air stream to increase the amount of power boost. The excess water is carried by the air stream into the compressor where it evaporates when the air is heated by the compression process. But a good control algorithm is still required, since the amount of overspray will vary (for a given water injection rate) depending on the ambient temperature and humidity. Even in sites where intercooling is specified, therefore, it may not be a good idea to omit the PLC.
Florida Power & Light opted for a turnkey fog unit with a stand-alone control system. Fog unit control could have been integrated into the existing turbine control system, but the utility opted for stand-alone control. “The supplier takes full responsibility for designing and building the complete system, thereby eliminating complex interfacing problems and the potential for finger-pointing,” says senior contracts agent Paul Seiler. Additionally, turnkey units save time during the installation phase.
Admittedly, most power producers possess the know-how required to incorporate fog system control into the existing plant control system. At one of Indeck Idaho Management’s plants in Rupert, Idaho, for instance, the fog system was hooked up to the facility’s existing Solar Allen Bradley 540 control system. As this unit already controlled the entire plant, Indeck decided to reconfigure the vendor’s software to fit the existing PLC rather than use the vendor’s PLC. “This involved considerable man hours and is contingent on the skill level of the people involved,” noted maintenance manager Chris Harriman. The plant installed 50 feet of conduit, pulled six pairs of conduit from the control room to the skid, added a transformer and spent about 30 man hours on reconfiguring software code.
With regard to cost, omitting the PLC seems to make little sense in a large or mid-range fog system, as it involves less than 5 percent of the total bill-an amount that can be recouped over the short term in reduced labor costs, water and energy usage, and in increased power output. Perhaps in smaller plants, however, where the PLC may account for a much larger percentage of the total fog system cost, it might be worth considering omitting this component. But take into account the potential financial repercussions, as well as exposure to human error in running the system manually.
In order to quote low, some manufacturers are forced to economize on such elements as paint coatings, metal grade or pumping system quality. This can result in a poor quality system that requires significant maintenance time.
Pump speed, for instance, should be carefully checked. To lower costs, some vendors specify fewer or smaller pumps, then operate them at higher speeds. Positive displacement, reciprocating pumps, however, are best run at low speeds, to reduce pump wear and vibration and increase the life of the pumps. By specifying larger pumps that are run at lower speeds, the system life is greatly increased.
One East Coast utility plant is a case in point. This facility had to replace several pumps shortly after installation due to malfunction. Florida Power Corp., on the other hand, installed a turnkey fog system with no compromise on quality and the system performed well without the need for more than routine maintenance.
Nozzle Manifold Position
Photo 3. Nozzle manifold in gas turbine inlet structure. Photo courtesy of Mee Industries.
Over the years, fog nozzles have been installed in several different positions in the turbine inlet air system (Photo 3). In a few cases, the nozzle manifolds have been placed upstream of the air filters, requiring the addition of a fog droplet filter to remove fog from the air stream. If properly engineered, this type of installation can work well. Special attention should be paid to the design of the droplet filters as a few unsuccessful installations caused excessive wetting of the filters.
Generally speaking, for an evaporative cooling type fog system, the best position for the nozzle manifolds is just downstream of the air filters. “This is the most common location for fog installations that are primarily to be used as evaporative cooling systems. It allows the most evaporation time for the fog and requires only one or two days to install,” reports Thomas Mee III, CEO Mee Industries. However, if the system is to be used for both evaporative cooling and fog intercooling (overspray), then it may be best to avoid spraying too much water upstream of the silencers. Some operators have reported that spraying excessive amounts of fog before the silencers results in frequent compressor fouling, due to water collecting on the silencers and washing dirt from the silencer panels into the compressor. When the nozzle manifolds are moved downstream of the silencers, the compressor fouling problem goes away. In some cases, it may even make sense to locate evaporative cooling manifolds upstream of the silencers and fog intercooling nozzle manifolds downstream of the silencers.
At TransAlta Energy Cogeneration in Mississauga, Ontario, for example, the differences between nozzle location can clearly be seen. Due to scheduled maintenance considerations and the plant’s desire to be ready for the summer season, fog was installed upstream of the air filters in one turbine and downstream in another. Both are GE LM6000 aeroderivative, base-loaded turbines. “The performance gain from the fog system placed upstream of the air filters was 1 MW less than the other unit with the nozzles in the downstream position,” said Lorne Reddy, assistant plant manager. “We are planning to consolidate both systems downstream of the air filters to achieve maximum cooling potential.” In this case, the fog nozzles installed upstream of the air filters didn’t perform as well because much of the fog was removed from the air stream by the droplet eliminator before it evaporated.
Impaction-pin Fog Nozzles
In recent years, great strides have been made in fog nozzle technology. Most experts agree that droplet diameter for inlet-air fogging should not exceed 30 microns. At smaller droplet sizes, cooling potential increases and the potential for compressor blade erosion decreases. With inlet air cooling fog systems, therefore, it is critical to specify high-grade, stainless steel impaction-pin nozzles, which can maintain droplet size below 10 microns with low energy consumption and maintenance requirements. Due to the creation of such fine droplets, most turbine operators elect to exclude mist eliminators downstream of the fog system, allowing fog to enter the compressor where the fog intercooling effect gives an additional power boost.
To reduce costs, one Southern utility used swirl-jet type nozzles rather than impaction-pin nozzles. Unfortunately, these nozzles are more suited to agricultural use and don’t produce a small enough droplet to function well in inlet air cooling. As a result, the plant experienced disappointing levels of power augmentation and reported some compressor blade wear due to the bombardment of large droplets.
The industrial application of fog has been around since the late 1960s, but fog systems have been applied to gas turbines only for the past 10 years or so and, to date, there are fewer than 200 fog installations on gas turbines. Some of these installations have been less than perfect and have resulted in damage to turbine components or high maintenance requirements. It’s best, then, to get a list of all the installations done by each vendor and to contact the users to find out what their experience has been.
The bids are in and the vendor with the best value has been given the contract. Now, however, is too late to discover that the chosen supplier is unable to cope with your delivery needs. One utility in the northern Great Lakes area, for instance, discovered that the fog system it ordered couldn’t be delivered before August, defeating the purpose of expediting the project for the summer season. Another plant in Wisconsin experienced delays due to the uncertainty of environmental permitting. “When everything was finally approved, the supplier was slow in reacting to our needs,” said the facility’s projects engineer, adding a further two months to the process.
An example of a rapidly executed installation is Florida Power and Light. “Fog systems were installed in 11 GE 7001B turbines (both peaking and base-loaded units) within a five-week period, producing an extra 2 to 3 MW per turbine,” said senior engineer Jay Blum. “The nozzle manifolds for our base-loaded turbines were installed in the morning and in use in the afternoon.”
Specifying liquidated damages for late delivery is a good way to get vendors to “put their money where their mouth is” and guarantee on-time delivery.
The time required for testing and commissioning in the fog business varies greatly, as one Indianapolis utility discovered. “Several pumps had to be replaced shortly after installation due to leaks and other failures,” said the plant’s specifying engineer. “Proper testing would have isolated these problems prior to shipping and onsite commissioning by a representative of the manufacturer would have made the whole process a lot smoother.”
Dime Wise, Dollar Foolis
Regardless of the cutthroat nature of the bidding process, keep in mind that there is often a major and speedy return on investment from installing fog. “It doesn’t take a rocket scientist to work out the math,” says Drew Wozniak of Louisville, Ky.-based Caldwell Energy & Environmental. “Fog has all but supplanted media-type evaporative cooling.” At City Utilities of Springfield, Mo., for instance, the plant manager calculated that with the high electricity prices per MWh last summer, the fog system paid for itself in two days by adding 5 MW onto the 40 MW being produced by each of its FT4-TP turbines. Because such gains can easily be lost due to a lack of testing, shoddy equipment or late delivery, it behooves the user to select carefully. As well as considering the amount of the bid, steer purchasing decisions toward vendors who can demonstrate high quality, on-time delivery and excellent after-sale support.
Drew Robb is a Los Angeles-based writer specializing in engineering and technology. In 1978, he graduated from Strathclyde University in Scotland, majoring in geology. He has written almost 200 articles for trade publications and has regular columns in several engineering publications. He can be reached at [email protected]