|Xcel Energy’s coal-fired Cameo Station in southern Colorado ran a pilot program from July 2010 to Dec. 2010 that integrated concentrating solar generation until the Cameo plant was retired at the end of 2010. A parabolic trough solar field provided thermal energy to produce supplemental steam for power generation in order to decrease overall coal consumption, reduce emissions, improve plant efficiency, and test the commercial viability of concentrating solar integration. Courtesy: Xcel Energy|
It may seem counterintuitive, but the strategic coupling of simple- and combined- cycle technologies with renewable generation could establish the conditions necessary for adding more renewable megawatts to transmission grids around the world.
Although many nations and most U.S. states have goals for increasing the percentage of electricity generated by renewable energy, meeting those goals means facing several challenges. In particular, the variable nature of wind and solar power creates difficulties for grid stability, especially as the percentage of variable power increases. The large penetration of wind power in some parts of the U.S., for example, affects the existing grid in terms of system capacity, harmonics, safety, and protection.
Fortunately, two separate but related technologies are already being developed and deployed to help resolve these problems: a “smart grid” and fast-start gas-fired generation. The use of these technologies could facilitate the deployment of more renewable generation and help states and countries meet their renewable portfolio goals.
However, questions remain. How will new generation assets influence the back-up power demand and its daily profile? How well will the design of the new conventional plants currently at the planning stage in the U.S. (mostly combined-cycle plants) cope with rapid changes in demand due to the ever-increasing penetration of renewable power? [Editor’s note: For more on this issu, see the top story in this issue’s Global Monitor.] Will anticipated CO2 capture and sequestration legislation require these facilities to develop higher efficiency and reduce their carbon footprint? Will smart grid initiatives change the way gas turbine loading is done? This article attempts to answer these questions and offer solutions for integrating renewable sources with conventional fossil-fueled plants.
Currently, all of the major equipment suppliers are offering solutions for these problems based on their specific gas turbine technologies and capabilities. However, contradictory requirements for maintaining high efficiency and emissions at part load, uncertainty about CO2 capture legislation, and difficulties in financing projects that employ innovative solutions make the process complex and highly challenging—not only for original equipment manufacturers (OEMs), but also for project developers and engineering, procurement, and construction contractors.
What the Grid Wants—and What It Gets
Electric utilities and grid operators would like renewable energy sources to behave as conventional, dispatchable power plants. For example, they would like to see a constant level of voltage from wind and solar plants. But wind farms and solar power plants require reactive compensation. In fact, rigorous reactive compensation standards are likely to become a reality for renewable energy power plants in North America, Europe, and Asia. An example is the recent “Interconnection Standards Initiative Draft Straw Proposal” set forth by the California Independent System Operator in the spring of 2010.
Meanwhile, solar power plants are being asked to meet power factor constraints, provide voltage control, and follow low- and high-voltage ride-through requirements. Renewable energy power plants, particularly solar plants, also must be able to provide day-to-day voltage support to maintain smooth and stable system voltages, even if the plant’s power output varies due to clouds, insufficient wind speed, or other factors during the course of the day.
It is imperative that renewable sources remain connected to the grid when they are most needed, particularly during power system disturbances, to help the grid recover.
Wind Power Variability. Because wind resources shift suddenly and dramatically, wind farms do not operate all the time; therefore, additional capacity is needed when they are not producing power, and differences between forecast and actual production have to be balanced.
Balancing and backup come at a cost, as does building new transmission infrastructure. These facts apply to wind energy just as they apply to other power-producing technologies that are integrated into electricity grids.
And although a wind turbine’s output is more variable and less predictable than that of conventional generation technologies, from a system operations perspective, the output of a single wind farm is just as irrelevant as the demand of a single consumer. The real challenge is matching the simultaneous collective demand of all customers with the entire available production from all sources. That has been the guiding principle of grid operation since its inception and will remain so regardless of which technologies are used.
That said, wind power is different from other power technologies, and integrating large amounts of it into the existing power system is a challenge. Here are some of the reasons:
- According to the Electric Reliability Council of Texas, less than 10% of total wind capacity is counted as being “available” during peak summer days.
- The PJM Interconnection regional transmission organization credits wind with about 13% capacity factor during peak periods. PJM coordinates the movement of wholesale electricity in all or parts of 13 states and the District of Columbia.
- Midwest ISO operators curtail thousands of megawatts of wind daily; 1,800-MW swings over hourly periods are common.
Without energy storage or fossil-fuel backup, integrating large quantities of wind power is difficult. Figure 1 provides one example of the discrepancies between wind power generation and demand.
|1. Mismatch. In this example from grid operator PJM, during peak load demand, wind production is close to its minimum. The economic impact is evident in the graph on the right, which shows that the premium price for power occurs when wind production is lower. Source: PJM|
Solar Power Variability. Figure 2 presents a dramatic illustration of the intermittent nature of solar photovoltaic (PV) power. It is important to evaluate not only the rate of change in generation, but also its magnitude. In seconds, the system can go from full output to 20% output and back again. At higher levels of PV penetration, such variability will significantly affect grid operation and power factor.
|2. Sun and shade. This chart shows a five-day period in the summer for several locations across Texas, a span that includes clear periods and periods with intermitten sunshine. Night hours have been omitted. Hourly data is from Aug. 10 through Aug. 15, 2005 and was derived from the 1991–2005 National Solar Radiation Data Base. El Paso data has been adjusted from its local Mountain time zone to coincide temporarily with the Central time zone. Source: Texas State Energy Conservation Office|
For direct electrical power generation renewable technologies such as wind and solar PV, there is no way at present to match grid demand. Though some forms of energy storage exist for solar thermal (molten salt is one example), only conventional fossil-fueled generation systems can cover the gap for other solar and wind generation.
The remainder of this article looks at technologies for bridging that gap and the design and operation considerations they raise.
The Smart Grid Part of the Solution
A smart grid differs from a conventional grid in that it is able to apply digital control to electricity supply and demand. A smart grid uses the analysis of vast amounts of data plus two-way digital communication to optimize the delivery of electricity from suppliers to consumers and, in some cases, to control demand at consumers’ homes or businesses. The ability to remotely fine-tune power generation and delivery can save energy, reduce costs, and increase reliability and transparency.
A smart grid is made possible by sensing, measurement, and control devices equipped with two-way communications capability that are applied to electricity production, transmission, distribution, and consumption parts of the power grid. Information about grid conditions is communicated to system users, operators, and automated devices, making it possible to dynamically respond to changes in grid condition. For example, automated demand-side management programs can allow system operators to reduce electricity demand during periods of low renewable plant output that would otherwise result in an overall shortage of power on the grid. (To learn more about the smart grid, use the Smart Grid tab at the top of the POWER home page at https://www.powermag.com to view previous articles on the subject.)
A fully developed smart grid includes an intelligent monitoring system that keeps track of all electricity flowing in the system. It is also capable of better integrating conventional and renewable sources of power generation, such as solar and wind. When combined with emerging technologies that are improving the ability of renewable power generators and system operators to predict more accurately, in smaller time increments, the output of wind and solar plants, a smart grid can maximize the benefit of these resources.
As regional and national grids incorporate smart technologies, the grid’s new capabilities will affect the behavior of simple-cycle and combined-cycle plants. In combination with larger quantities of renewable generation, the requirements for gas turbines will certainly change:
- The tendency will be to move toward smaller and dispersed plants.
- The type and frequency of cycling operation will be significantly different than is seen today.
- More-frequent cycling will affect component life and plant efficiency and will result in much more stringent environmental emissions at part-load operation.
- A significant reduction in peak load should occur due to better grid management and intermittent renewables, while some baseload resources will see an increase in the percentage of their utilization share.
The Gas-Plus-Renewables Option
As wind and PV capacity increases, these renewable sources are used to meet intermediate and peaking loads. However, during periods when PV or wind generation output is low, additional backup power is needed to ensure that the grid demand is met. Figure 3 illustrates the need for spinning reserves or storage. In the absence of large electrical, thermal, or pumped storage options, providing backup power and maintaining spinning reserve will be a major role of fast-starting and rapid-loading gas turbines.
|3. Solar backup. This scenario from a DOE study shows that as solar photovoltaic generation accounts for an increasing percentage of total generation, the need for spinning reserve or storage also increases. Source: DOE|
Gas turbines are particularly well-suited to operate in conjunction with wind and PV sources (see sidebar). Their well-known fast-start and fast-ramping capabilities are better able to meet rapid changes in grid requirements than coal, steam, or nuclear plants. Consequently, until alternative solutions are widely available, there is a real need for manufacturers to adapt gas turbines specifically to compensate for renewable power variability. Grid codes and customers are continuously demanding more operational flexibility, faster starts, and accelerated loading response times.
As a rule of thumb, for each installed 400 MW of wind power, 100 MW of gas-fired backup power is required. Hence, the requirement that gas-fired generation support renewable generation is driving modern power plant design to strongly focus on operational flexibility.
When compared with a continuous baseload regime, gas turbine operation over wide power ranges not only increases fuel consumption but also impacts NOx and CO emissions. The inability to achieve premix combustion operation at low power levels makes this particular requirement difficult to meet.
The good news, according to a Brattle Group study, is that a wind-plus–gas turbine plant could achieve at least 75% reduction of the maximum possible CO2 emissions. It is obvious that emissions-free power from wind generation is compensating for some of the conventional fossil-generated power.
The bad news is that the economics of gas turbine operation under these conditions are different than for a conventional standalone gas plant. All parties involved in determining the price of electricity must account for increased costs incurred by the gas turbine power generators. Forcing these facilities to operate at less than full capacity reduces their revenue stream. Cycling operation also affects maintenance schedules and gas turbine availability.
In response to these diverse requirements, OEMs have developed both heavy-duty and aero-derivative gas turbines with greater capability to support a wide range of operational flexibility enhancements, enabling customers to effectively use equipment for peak and cycling applications. The relative ease and speed of installing gas-fired generation also gives it an advantage when it comes to meeting emergent and urgent power demand.
Some features of OEMs’ solutions to the new demands on gas technologies are examined below.
All major manufacturers have realized the importance of fast start-up and rapid loading for gas turbines in simple-cycle operation, particularly for installations designed for cycling operation. Current gas turbines can ramp at the rate of 3% per minute (though their efficient operating range is narrow), which is a much higher rate than that of pulverized coal plants. The coal-fired steam cycle exhibits substantial power losses each time a steam turbine is shut down or restarted. Major ramp-downs in under 15 minutes may require wasting (venting) the steam, and major ramp-ups in less than 15 minutes may be impossible.
Here is a short list of published features of gas turbines in simple-cycle operation that demonstrate this technology’s attractiveness for supporting variable renewable generation:
- GE has a 100-MW gas turbine (LMS 100) capable of a 10-minute start for lower-megawatt applications and a recently announced GE-7FA.05 gas turbine, also with 10-minute start-up capabilities. Similar start-up and ramping features are offered by Siemens and Mitsubishi Heavy Industries (MHI).
- Due to a unique sequential combustion design, Alstom turbines’ efficiency at part load is higher than others. The same capability of shutting off one combustor at part-load operation offers an additional advantage for operating in this manner: continuous operation at close to 30% of baseload while remaining in compliance with baseload emissions levels.
- An alternative to frame gas turbines are the aero-derivatives. One of the owners of large wind farms (Westar Energy) uses GE’s LM6000 gas turbines to meet demand for peak load and to cover for shortfalls in wind farm generation. The LM6000 can reach its full load output from cold start in less than 10 minutes and operate for 1 hour or less. When a number of LM6000 gas turbines are on standby, they can be dispatched immediately and are able to respond to situations when high-speed and wide-ranging wind fronts are cutting wind turbines’ production by hundreds of MW.
The real challenge for the industry is to develop capabilities for fast start-up and rapid loading of equipment without affecting its availability and reliability. Cycling operation must be performed without increasing the number of equivalent operating hours or accelerating the maintenance schedule. To that end, here are some of the actions initiated by OEMs:
- Use high-starting-reliability systems for the gas turbine and balance of plant.
- Implement complex control systems capable of providing adequate ramp rates for each specific state of the hardware.
- Employ a high degree of start-up automation for both gas and steam turbine.
- Implement measures aimed at heat retention during shutdowns, such as stack dampers and the use of auxiliary steam.
- Provide sophisticated monitoring systems for major equipment conditions, allowing operators to evaluate the impact of accelerated start-up or cycling operation on component life.
- Allow the gas turbine to rapidly ramp without the constraints of the heat-recovery steam generator (HRSG) and steam turbine.
Heat-Recovery Steam Generator. It should be remembered that for CCs, the element requiring the most time to reach baseload is not necessarily the gas turbine. For HRSGs, the most appropriate solution to accelerate the start-up process is the use of the Benson-type high-pressure (HP) circuit. Following are some of the well-known mechanisms affecting the performance and integrity of the HRSG components in cycling operation:
- Low cycle fatigue
- Thermal shock
- Oxidation and exfoliation
- Differential expansion
- Corrosion fatigue
- Corrosion in tubes
- Flow-accelerated corrosion (FAC)
- Corrosion product migration
All components in an HRSG are subject to the operating-life-affecting mechanisms listed above. However, some components may be more vulnerable because of their location, construction, or exposure. Critical components in an HRSG generally include these:
- Superheater and reheater outlets
- Tube-to-header joints in hot sections
- Drum to downcomer nozzle in HP drum
- Bent portion of the heat transfer tubes
- Bypass valves
These need to be designed and monitored more closely for any kind of life-affecting conditions. Solutions offered by OEMs include:
- Designing hot section outlets to minimize side-to-side variation.
- Using full-penetration welds, generating a joint with longer fatigue life.
- Limiting the use of dissimilar materials.
- Designing an adequate draining system aimed at reducing quenching effect.
- Employing various methods to keep the drums warm during shutdowns.
- Equipping the stack with a stack damper.
- Using special alloys to mitigate the exposure of critical components to FAC.
- Including special features such cascading bypass to minimize thermal shock during start-up.
Steam Turbine. Design and operability improvements in steam turbines in CC operation have allowed overall start-up and turbine rolling to baseload in record times. As mentioned above, new and complex control features and stress measurements for steam turbines permit CCs to respond much faster than before, particularly during hot starts and load-following mode. Without appropriate flexibility for the start-up times of steam turbines, the viability of CC as a power-controlling element for renewables will be significantly reduced.
It should be noted that the most logical solution for this type of application is a 1 x 1 (one gas turbine, one HRSG, one steam turbine generator) arrangement. Other configurations (2 x 1 and 3 x 1) will require larger turbines with a longer start-up time. A modern G or H class gas turbine in a 3 x 1 CC configuration might require a 400-MW to 500-MW steam turbine and therefore be more suitable for baseload operation.
Considering these constraints, the preferred configuration of CCs for renewable back-up power must be selected as a result of detailed feasibility studies combining the start-up curves of all CC components and their control systems.
Integrating Solar and Fossil Generation
An additional role that a CC can play in the deployment of renewable power, specifically solar thermal, is to accept the steam produced by a solar thermal source into its steam cycle. This arrangement is called integrated solar combined cycle (ISCC). By including an additional source of heat, such as solar energy, the efficiency of the system is dramatically increased. Annual electricity production is increased because the steam turbine is already in operation, avoiding lost time for start-up. During solar operation, the steam produced by the solar heat source offsets the loss of power typical for a CC when the ambient temperature is higher.
Hybrids involving conventional coal-fired plants are also possible in regions with reasonably good solar conditions, as was the case for Xcel Energy’s Cameo Generating Station in southern Colorado, pictured at the top of this article. For these plants, where the steam pressures and temperatures are higher than for ISCC, the type of solar conversion technology used (linear Fresnel, parabolic trough, or tower) will dictate how solar is integrated into the plant. Table 1 summarizes the types of technology and their thermal output.
|Table 1. Summary of concentrated solar technologies. Source: Bechtel Power|
When planning to integrate steam generated by solar energy into a CC, several questions must be answered. What solar technology should be used? How much solar energy should be integrated into the CC? Where is the best place in the steam cycle to inject the solar-generated steam? Unfortunately, there are no simple answers to these questions. A detailed economic analysis must be performed to determine the levelized cost of electricity for the specific site under consideration. This analysis must look at different MW thermal solar inputs to the CC and different solar technologies that generate differing steam conditions.
How to integrate steam generated by a solar technology obviously depends in large part on the steam conditions that can be generated by that technology. One must remember that all power generated in the steam cycle of a CC is “free” from a fuel perspective. In other words, steam cycle power is generated without burning any additional fuel (all steam cycle power is generated based on the energy provided in the gas turbine exhaust gases). Thus, one must be careful to not just substitute the free energy from solar power for the free energy in the gas turbine exhaust gases. When integrating solar into the steam cycle of a CC, it is important to try to maximize the use of both sources of free energy.
Next we look at how various solar technologies can be integrated into CC power plants. Because technologies are evolving and improving, each technology has been categorized based on fluid temperature capability: High temperature is >500C/>932F); medium temperature is 400C/752F; low temperature is 250C–300C/482F–572F. Medium-temperature technology is presented first, as it is the most proven technology.
Medium-Temperature Solar Technology. The most common medium-temperature solar technology is the parabolic trough. Studies have indicated that, for parabolic trough systems that can generate steam up to ~380C, it is best to generate saturated high-pressure (HP) steam to mix with the saturated steam generated in the HRSG HP drum. Integrating HP saturated steam into an HRSG is very common in integrated gasification combined-cycle plants.
Solar thermal input to an ISCC can also be used to reduce the plant’s fuel consumption. Reducing gas turbine fuel consumption also reduces gas turbine power and exhaust energy. For the same plant net output with 100 MWth of solar energy input, plant fuel consumption is reduced by ~8%.
High-Temperature Solar Technology. Solar tower systems can generate superheated steam at high pressure and up to 545C. These conditions allow admission of solar-generated superheated steam directly into the HP steam line to the steam turbine. In addition, steam can be reheated in the power tower much as it is in the HRSG. This minimizes impact on the HRSG, because superheating and reheating of solar steam take place in the solar boiler.
Low-Temperature Solar Technology. Most linear Fresnel systems fall into this category. These systems generate saturated steam at up to 270C/55 bar (518F/800 psia). (More recently, the technology has been enhanced to reach higher temperatures.) This pressure is too low to allow integration of the steam cycle into the HP system. Basically, two options exist:
- Generate saturated steam at ~30 bar (435 psia) and admit to the cold reheat line.
- Generate steam at ~5 bar (73 psia) and admit to the low-pressure steam admission line.
As with the other solar systems, taking the feedwater supply from the optimum location in the steam cycle is of great importance to maximizing system efficiency. At the same time, low-temperature systems allow less flexibility in selecting the feedwater takeoff point, because the takeoff temperature must be below the saturation temperature of the steam being generated.
An increasing number of states and countries mandate that a portion of new generation must be renewable. In the absence of adequate storage solutions, the energy generated by wind or solar typically has to be absorbed into the grid regardless of load demand.
Predicting the output of renewable generation in time to allow grid operators to adjust to sudden losses of many megawatts could be a daunting task—at least until a fully functional smart grid is in place. Another way to compensate for shortfalls in power and maintain grid stability is to use gas turbines in simple- or combined-cycle operation.
OEMs have introduced many creative solutions to allow fast start-up and operation at part load without affecting equipment availability and reliability. Large frame industrial turbines and, in particular, aero-derivatives—with their rapid output rate increase—are suitable options for maintaining grid stability and meeting customer demands for power. ISCC also offers a solution in that the steam output of solar thermal plants could be combined with that of conventional gas-fired units, typically resulting in lower fuel consumption and lower capital costs than a standalone solar plant.
As the renewable power generation portfolio continues to grow, so too will the role of the gas turbine industry.
— Dr. Justin Zachary ([email protected]) is a POWER contributing editor, technology manager for Bechtel Power Corp., and a Bechtel and ASME fellow.