By Brian Reinhart, Kandi Howell and Kyle Kropf, Black & Veatch
Operators are under increasing pressure to find ways to improve the dispatch capability of their combined cycle facilities. Increased penetration of renewable generation has resulted, and will continue to result, in greater differences between peak and off-peak net load demand. Like it or not, combined cycle plants are under increasing pressure to chase variable wind and solar resources.
Several options can be employed to improve the reliability and flexibility of combined cycle plants. Each option must be considered within the context of the overall facility and the future outlook of power market pricing. This article highlights recent combined cycle dispatch trends and outlines plant dispatch improvements in the following areas:
- Part Load
- Peak Load
Select case study examples are presented to illustrate real-world opportunities.
EVALUATING the OPTIONS
Low gas prices, widely accepted to be stable for the foreseeable future, coupled with an unprecedented number of coal plant retirements, have placed gas-fired combined cycle plants in the slot of “baseload” generation. The reality, however, is that the combined cycle fleet (along with the remaining coal fleet) is largely responsible for filling the gap between inflexible nuclear generation and variable wind and solar. This need to fill the gap, or match net load demand, will play an increasing role in the economics of combined cycle plants.
Weekly reports issued by the Electric Reliability Council of Texas (ERCOT) and made available to the public provide a good illustration of how ERCOT load and wind output do not necessarily correlate well, leaving dispatchable generators responsible for filling a highly volatile gap, as presented in Figure 1. The example presented in Figure 1 shows a randomly selected week in spring where net load demand ranges from about 12.5 gigawatts (GW) to 41 GW within the span of a few days.
Many markets are experiencing extremely low off-peak prices and fewer high dollar on-peak prices; therefore, increased reliability and flexibility will be instrumental in keeping combined cycle plants economical. Improvements discussed herein can help enhance a combined cycle’s ability to make money for a utility through increased revenue and decreased cost.
Options should be evaluated on a cost/benefit or net present value basis, where project costs and incremental operational costs are weighed against avoided costs and increased revenues.
Improvement options and project examples are presented in the following sections.
IMPROVEMENT OPTIONS AND EXAMPLES
Startup and ramping characteristics have historically been a secondary consideration for combined cycle plant design. Long start times and slow ramping capabilities are largely a result of a lack of perceived economic value, coupled with less apparent design ramifications. Additionally, operating and maintenance practices may not have been established with frequent cycling events in mind. The following are some common areas of potential improvement.
“This need to fill the gap, or match net load demand, will play an increasing role in the economics of combined cycle plants.”
Review Maintenance Practice
Often, changes in operating modes can lead to maintenance plans and strategies that differ from those used in the original design or commissioning. Figure 2 shows the operational impact caused by operating mode changes. Preventive maintenance plans were historically developed with time-based strategies, for example, monthly, quarterly or annual tasks or inspections. These time-based strategies used assumptions of expected operating profiles, starts/stops and ramp rates. Moving to more operating and condition-based maintenance tasks can improve equipment reliability. Condition-based maintenance and online condition monitoring continue to play an important role in managing overall equipment reliability and high impact equipment failure reduction.
Review and Audit of Startup Sequence
The start sequence should have been established on the basis of major equipment original equipment manufacturer (OEM) recommendations during plant commissioning. Plant commissioning engineers sometimes fail to follow OEM recommendations by either taking a more aggressive approach or, possibly, a more conservative approach according to their personal experiences with other, seemingly similar combined cycle plants. Over the years, previous plant management and operations might have taken similar stances and decided to modify start logic. Often, the first step in evaluating potential improvements to a plant’s startup capabilities is to perform an audit for which engineers develop an independent conceptual start sequence using equipment OEM recommendations. The conceptual start sequence can then be compared against current practices to determine what might, and what will not, work. The result may be a modified start sequence that can serve as a baseline for all further start improvement considerations. Implementation may be as simple as a revised operating procedure and distributed control system (DCS) logic.
Elimination of Purge on Startup
Purging the gas turbine and backend equipment during startup can be a lengthy process that lasts anywhere from five to 20 minutes. Eliminating purge on startup is possible by modifying the fuel and ammonia supplies. National Fire Protection Association (NFPA) 85 allows for elimination of purging at startup as long as the operator implements proper valving arrangements (i.e., triple block valves, double venting, valve on/off position switches and pressure switches or transmitters) and performs certain actions (valve proving at shutdown and continuous monitoring of valve positions and pressures). Purge credits can be maintained for 8 days or indefinitely if the pressurized pipe method is performed using air or inert gases as a sealing medium.
This relatively inexpensive plant modification allows for much faster gas turbine power production. This approach does require careful consideration for how to maintain continuous monitoring during shutdown, which can be considered onerous when accounting for issues such as position feedback noise.
When a change from baseload to cycling operation is considered, the plant water treatment system should be reviewed to ensure that it is sized appropriately. Because of startup blowdown, each startup requires additional makeup water, which sometimes requires modifications to increase the capacity of the existing water treatment system. This often results in adding capacity to the raw water inlet screen and filters and expansion of the demineralized water treatment system. Adding demineralized water storage tanks can reduce the size of the water treatment system depending on the cycling frequency of the unit. Alternatively, one solution that could drastically reduce startup water usage is the addition of a condensate polisher. Condensate polishers purify the condensate of various impurities, such as iron from the feedwater, which can concentrate in the heat recovery steam generator (HRSG) and impact performance and component life. In addition, condensate polishers can reduce the unit startup times resulting from water chemistry-related delays, minimize condenser leaks and reduce boiler cleaning frequency.
Heat Retention and Preheating
HRSG and steam turbine ramp rates are dictated by starting temperature. Two methods of improving ramping rates during startup are retaining the heat absorbed from the last time of operation and preheating equipment and piping in advance of a start sequence. A stack damper, coupled with insulation up to the point of the stack damper, is utilized to isolate the HRSG from the stack during shutdown to minimize natural draft cooling of the HRSG. An auxiliary steam supply, often from an auxiliary boiler, can be utilized to reduce startup times and the impact of cycling on the unit. Auxiliary steam is supplied to the HRSG drums, condenser sparging, steam line warming and steam turbine seals to warm systems preceding a startup. While it may be a long and difficult process to add an auxiliary boiler to an existing site, it is possible. Ideally, the site already has an auxiliary boiler with sufficient capacity, whether the boiler was used for other purposes and not properly tied into the cycle or whether tie-ins are already present but not being used to their full extent. In addition, auxiliary boilers can be equipped (or retrofitted) with heating coils in the auxiliary boiler mud drums to further reduce their startup time.
Thermal Decouplin of Steam Turbine
During startup, steam turbine admission temperature cannot be controlled using interstage attemperation. The reason is that interstage attemperators have only limited control of the steam temperature leaving the second to last superheater and reheater sections, and spray flow is restricted to maintain superheat before the final superheater and reheater sections. Much of the heat available in the gas turbine exhaust is subsequently absorbed by the final sections, and the resultant steam conditions to the steam turbine are largely unaffected by the interstage conditioning. Because of this, a traditional approach to starting a combined cycle entails holding the gas turbine at a very low load, with reduced exhaust temperature, for a period of time sufficient to warm the steam turbine and ancillary equipment. This works well from an equipment preservation perspective but results in a very slow startup. Adding final point attemperation to the main steam and hot reheat piping between the HRSG and steam turbine allows better control of steam turbine admission temperatures, freeing the gas turbine to ramp at a rate allowed by the HRSG. By adhering to the HRSG OEM pressure section ramp limitations, the gas turbine ramp time to minimum emissions compliance load (MECL) can be reduced, improving combined cycle responsiveness and minimizing startup emissions.
Typically, plant startup procedures call for visual checks and manual operation of valves (i.e., steam line and HRSG vents and drains). The time frame required for this process depends on whether the start was anticipated, plant staff availability to perform and other conditions. Installing proper instrumentation and block valve actuation maximizes control room operator capabilities and reduces start times. Typically, a start sequence audit will reveal such opportunities for automation.
When the gas turbine is started or ramped at high rates of change, rapid changes occur in steam production and nitrogen oxide (NOx) production. Modifying the steam attemperator controls and possibly the drum level controls to include a feed-forward loop informed by gas turbine ramp rate settings improves responsiveness in maintaining adequate boiler feed pump head and improves the consistency in admission steam temperature to mitigate side effects of fast ramping, such as temperature excursions, load holds, load runbacks and trips. Ramping the gas turbine quickly can result in rapid changes in NOx production. Since NOx is typically measured only at the stack, fast ramping can lead to over injecting or under injecting ammonia. Over injecting ammonia can result in higher ammonia slip, potentially exceeding permit limits, and greater conversion of ammonia salts, increasing the potential of deposition on the low-pressure (LP) evaporator and economizer sections. Under injecting can result in stack NOx emissions exceeding permit limits. To mitigate these issues, the addition of feed-forward controls should be considered.
– Elimination of Purge on Startup
The example presented is a 3-on-1 combined cycle (three gas turbine/HRSG trains supplying steam to a single condensing steam turbine generator) based on Mitsubishi G-Series gas turbines. Figure 3 is an example startup curve showing the combined cycle block load at startup before and after applying NFPA 85 purge credit retrofits.
Assuming a dispatch profit margin of $6 per megawatt-hour (MWh) and 150 starts per year, the estimated benefit of incorporating natural gas and ammonia supply purge credit designs would result in a profit of approximately $1,400 per start, or $210,000 per year. These savings are likely understated because the assumed purge time used was five minutes. Purge times on the order of 10 to 15 minutes are more typical for combined cycle plants.
– Thermal Decoupling of Steam Turbine
A feasibility study was performed for reducing startup times on a 1-on-1 combined cycle based on a Siemens-Westinghouse 501FD3 gas turbine, Nooter/Eriksen HRSG and Fuji Electric Type KN steam turbine. One of the options considered included the addition of final point attemperators. Adhering to the HRSG OEM pressure section ramp limitations, the gas turbine ramp time to MECL could be reduced by about 140 minutes, as shown in Figure 4, during a cold start event.
Part load performance and turndown capabilities are becoming increasingly important as the difference between net load demand peak hours and off-peak hours grows. The benefits of keeping a unit online rather than cycling the unit off and on can be huge. A unit online is much more reliable and responsive than a unit offline. In addition, avoiding start/stop cycles can help preserve the life of the asset. The following are some common areas of potential improvement.
Boiler Feed Pumps
Operating a boiler feed pump at low flow increases pump stresses and strains on both the pump and feedwater control valves, causing premature wear or, in severe cases, failure. Solutions to mitigate this include upgrading the pump to a design more suited to frequent starts/stops and implementing a variable frequency drive (VFD) or fluid coupling drive. When modifying an existing feedwater pump(s) is considered, a VFD is often evaluated as the lowest installed cost option because of the minimal physical changes required to the pump and motor; whereas, a fluid coupling requires shifting the motor to accommodate for the space of the coupling. Additional benefits of VFDs include the ability to use a single VFD for multiple redundant pumps in addition to improving the plant heat rate at lower loads. However, fluid couplings are typically the more cost-effective approach for a new-build combined cycle because they have a lower capital cost and can be designed into the pump layout.
High energy piping, from small bore attemperator feed piping to the large bore steam lines, is also impacted by the increased thermal cycles, often leading to pipe support failure and resulting in inadequately supported piping. The best solution to mitigate this impact is to take a proactive approach to performing high energy pipe support audits every couple of years.
If unsure of where to start, the best practice is to perform a balance-of-plant equipment audit to ensure that equipment is operating within its design capacity, determine any potential impacts associated with part load operation and identify potential opportunities to improve part load turndown and/or efficiency.
– Boiler Feed Pump Upgrades
Black & Veatch recently supported a client with a 2-on-1 General Electric (GE) 7FA combined cycle plant that was experiencing boiler feed pump reliability concerns. The plant was designed for baseload operation, and to save capital cost, the original feedwater pumps were provided with only a balancing drum and slinger type oil lubrication. Because the feedwater pumps had no active control of the shaft axial position, the balancing drum was damaged as the pump accelerated to speed on each startup. Black & Veatch was engaged to study the feedwater pump failure information, perform a system assessment and make recommendations for pump improvements. Discussions with the client and the pump OEM led to the purchase and installation of a new pump with active thrust bearings and forced oil feed lubrication, which is better suited for cycling operation. In addition, the new pump was able to reuse the existing soleplate, motor and piping connections, thus minimizing the outage time for installation.
Black & Veatch’s design also focused on improving pump operation minimum flow control, including improved piping layout, tight shut-off valve and elimination of concerns with flow accelerated corrosion of the high velocity pump minimum flow piping. The piping system hydraulics and piping stress were analyzed to ensure successful long-term operation. The improvements eliminated reliability and maintenance issues with the original minimum flow control valve and provided the flow capacity required for cyclic operation of the existing feedwater pumps. Feedwater pump suction piping was rerouted to eliminate an intermediate high point in the suction piping that was an operating concern with potentially catastrophic effects on pump operation. Finally, an in-depth study of the pump protection instrumentation and logic, comparing the existing systems with industry best practices, was performed; control system changes for pump protection were recommended from this list.
Increasing the peak output capacity of a combined cycle must be carefully evaluated as balance-of-plant design limitations, often in the areas of electrical export, water/steam production, steam turbine flow, heat rejection, and air permit limits can increase project cost or even render certain upgrades infeasible. In general, a combined cycle already equipped with supplemental HRSG duct firing and an oversized steam turbine is often more amenable to such upgrades. The following are some common areas of potential improvement.
There are generally two forms of evaporative cooling: wetted media and fogging. Wetted media evaporative cooling requires a large surface area typically made of a cellulosic or glass fiber material that is wetted using a filtered water stream. As the air passes through the wetted material, water is vaporized, and gas turbine compressor inlet temperature approaches wet-bulb temperature. The moisture-laden air is cooled and at a higher density, thus increasing mass flow and output of the gas turbine, often with the additional benefit of improved gas turbine efficiency. As a general rule, wetted media evaporative cooling is capable of cooling the gas turbine inlet air by about 85 percent of the difference between the dry-bulb and wet-bulb temperature. Typically, no modifications are required for the bottoming cycle or electrical systems, provided that some margin currently remains. Fogging technology works in a manner similar to wetted media except that the water is introduced into the inlet airstream in an active manner. Atomized high purity water is injected into the inlet airstream through an array of nozzles. Water injection is controlled using a weather station, which is often mounted on the fogging pump skid. Since gas turbine filters and inlet filter houses are not designed to pass saturated air, the compressor inlet air temperature is typically maintained slightly above wet-bulb, by approximately 1 to 2° F (0.56 to 1.11° C). A schematic of a fogging system is presented on Figure 5.
Chilling reduces the gas turbine compressor inlet temperature, thus increasing gas turbine output and often reducing gas turbine heat rate by removing heat from the inlet airstream through indirect cooling. Chilling works by sending a chilled coolant (often water or a water/glycol mixture) from a condenser to a finned tube heat exchanger, or chiller coil, situated in the gas turbine inlet airstream. The chiller coil then removes heat from the inlet airstream and returns the warmed coolant to the chiller unit, which is typically an electrically powered mechanical vapor-compression refrigeration cycle. Unlike evaporative cooling, chilling works through indirect cooling, and no water is imparted into the gas turbine inlet airstream. In addition, chilling is capable of reducing the compressor inlet air temperature below the ambient wet-bulb temperature, potentially as low as 45° F (7° C).
Gas Turbine Upgrades
Gas turbine upgrade options vary depending on the OEM and model. Upgrades can be as common as extended life hot gas path parts or minor software upgrades, or more extreme, such as an entirely new compressor or hot gas path design. Upgrades can be performed to improve reliability, maintainability, performance or a combination of these. Unlike evaporative cooling or chilling, many gas turbine upgrade options improve performance throughout the year. In addition, much more significant improvements are often possible. Drawbacks might be high initial cost, higher long-term maintenance fees and possibly having to permit the upgrade as a major modification.
Steam Turbine Upgrades
Steam turbine upgrades are something to consider for a plant’s next major overhaul. Advances in steam turbine technology (improved airfoil profiles, sealing techniques, materials and cooling techniques), along with the potential to “open up” the steam turbine flow path for increased output, can have large returns, especially for a combined cycle already equipped with supplemental HRSG duct firing and an oversized steam turbine.
– Evaporative Cooling
A client was interested in increasing output during summer conditions for a 3-on-1 combined cycle based on GE LM6000 gas turbines, two pressure non-reheat steam turbines, and an air-cooled condenser. Two evaporative cooling technologies were considered: wetted media and fogging. Wetted media evaporative cooling is considered to be a proven passively controlled technology requiring minimal electricity. However, its limited ability to improve performance, coupled with the need to retrofit the existing inlet air filter housings for the added space and weight of the wetted media and water trough, made that option unappealing.
Fogging is capable of depressing the gas turbine inlet air temperature better than wetted media. In the above application, because of the inlet air duct arrangement, the retrofit installation would have to be mounted just inside the inlet air louvers, upstream of the filters. Since the filters are not designed to pass saturated air, the inlet air temperature would need to be maintained above saturation; a 1° F (0.56° C) margin was selected as the target control. The fogging nozzle array could be placed farther downstream, just upstream of the compressor inlet; however, water agglomeration on birdscreen and structural members just upstream of the bellmouth has the potential to impinge on the compressor blades. Fogging relies on electricity to drive high-pressure (HP) positive displacement pumps and a source of high purity demineralized water. In addition, there is some risk associated with fogging because it is an actively controlled technology; a potential exists for issues caused by overspray. As with wetted media evaporative cooling, no modifications were identified as being required for the bottoming cycle or electrical systems. Combined cycle hot day operation net plant output was estimated to increase by about 5 to 8 percent, and net plant heat rate was estimated to decrease by about 1 to 1.5 percent for both evaporative cooling options considered, depending on hot day relative humidity.
– Steam Turbine Upgrades
A client wanted to increase the summer capacity of a 3-on-1 combined cycle based on GE 7F.04 gas turbines, Foster Wheeler HRSG with supplemental duct firing capabilities and an oversized Toshiba two casing steam turbine with combined HP and intermediate-pressure (IP) casings and two flow LP casings. A proposal to upgrade the steam turbine was reviewed, and a thermodynamic model of the plant was created to assess feasibility of the upgrades and balance-of-plant limitations. Proposed upgrades included new fully bladed HP-IP rotor, HP-IP nozzle diaphragms, HP inner casing, new LP blades for the first two stages, diaphragm packing seals, gland packing seals and other associated steam turbine hardware along with necessary generator stator, rotor, hydrogen cooler, excitation and protection upgrades.
The preliminary assessment revealed that the level of duct firing would be limited by the HP turbine exhaust pressure limit because of restrictions in the reheat piping and LP turbine flow passing capabilities.
The upgrade proposal was revised accordingly to reflect the discovered limitation, resulting in significantly diminished guaranteed fully fired steam turbine output.
The isolated phase bus duct and generator circuit breaker were found to be undersized, which would require further attention and potential replacement should the project proceed. The estimated combined cycle net maximum increase in power was about 4.5 percent.
Combined cycle plants are under increasing pressure to chase variable wind and solar resources. Combined cycle flexibility is expected to play an increasingly important role in keeping the plants economically viable through improvements in reliability and flexibility.
The options described and examples presented in this paper demonstrate that there are many avenues to consider. Whether the objective is to minimize downtime, improve start responsiveness or maximize output and/or heat rate, there are solutions that could offer high payoffs. Understanding and evaluating the options is paramount to the success of a combined cycle facility.