By Craig S. Brooker, P.E.
Across the US power markets, there are needs for new peaking capacity. Some may value peaking differently; however all markets value the main characteristics of a peaking facility: quick start/ramp rate, emissions compliance, flexible operation, and output. Historically, aeroderivative gas turbines have been the benefactor of the peaking market given their characteristics that match the demands. With the advancements in frame machine technology and the projections of low gas prices, the default answer to fulfill peaking market needs may not be so clear cut anymore.
Traditionally, frame turbines have had poor ramp rates, start times and a negative perception of maintenance due to starts charges. In addition to those deficiencies, historically an SCR on frame engines in simple cycle was more complex due to higher exhaust gas temperatures. The aggregate market impact has resulted in a peaker market dominated by aeroderivative turbines. However, given the advancement of frame gas turbines and SCR technology over the past decade, many of these shortfalls have been addressed. This includes faster start up times, improved ramp rates and overcoming the technical challenges of combining a hot SCR with a simple cycle frame engine. Thus the frame machine has advanced such that it is technically feasible for the peaking market. More importantly, now that it can compete technically, the market may realize the significant capital cost advantage of a simple cycle frame turbine. This evolution challenges the conventional thought that a peaking plant equates to an aeroderivative engine.
This article aims to provide a snapshot comparison of frame versus aero gas turbine technology and how these engines have technically evolved over time.
|PJM Demand Curve|
I. Peaking Market Fundamentals
Regardless of what electricity market you are in, peaking plants are valued for their ability to be responsive and flexible to market demands. They meet those demands by using technologies that have low turn-down, fast start times and quick ramp rates. There are various reasons why these performance attributes are valued more so for peaking operations than baseload operations, ranging from specific market structures to a region’s generation mix. Peaking plants, which are traditionally gas turbine simple cycle plants, typically only operate during peak load times, which are seasonal and/or cyclical and therefore limit their operating hours.
Four years of data (2012-2015) of simple cycle generating units across the United States shows that 90 percent of plants had an operating profile consisting of a capacity factor of 15 percent or less and 150 starts or less based on data obtained via Velocity Suite. The average capacity factor for a peaking plant results in low operating hours as shown in Figure 1. With low operating hours, it is in the generator’s best interest to respond quickly and generate as many megawatt-hours as possible within that plants specific infrastructure constraints (i.e. gas supply, transmission, etc.) when dispatched in order to capture the peak energy prices and maximize operating revenue from the energy market.
The ancillary services market (up/down regulation and spinning/non-spinning reserves) exemplifies the flexible and responsive characteristics of a peaking plant. Regulation helps to maintain grid stability by tightly controlling the system frequency to around 60 Hz, which means having to respond to rapid load changes that happen every few seconds. In the most stringent regions reserves are required to have at maximum a 10 minute response time to increase its’ current generation level from either online and synchronized status (spinning) or offline and non-synchronized status (non-spinning). Figure 2 summarizes various ancillary services and their response times.
|Ancillary Service Response Times|
A region’s generation mix also plays a role when discussing peaking operation. For example, a region that has a high amount of renewables will require units that have a quick response to not only the shoulder hours of generation of solar assets, but also the volatility of wind generation. One can observe the impacts of these differing operating characteristics in Figures 3& 4.
Another market that peaking units participate in is the capacity market. Although this market does not value responsiveness and flexibility as much as the energy and ancillary services markets, it does provide a source of revenue which is directly tied to the capital cost of the pant. Plant capital costs will be discussed later in section IV.
II. GT Technology Evolution
One of the key factors in power plant performance is the technology around which that plant is designed and constructed with the key technology being the gas turbine. This section looks at general gas turbine attributes, both historic and recent.
Both simple cycle and combined cycle plants can showcase responsiveness and flexibility through inherent gas turbine technology and plant design. The primary focus of this article is on simple cycle plants and gas turbine technology. Historically, frame gas turbines have had attributes not conducive to peaking market applications. They have had long start times, poor turn-down, slow ramp rates and start penalties, an aspect not previously discussed. When compared with the aeroderivative gas turbines, there is a stark difference, as shown below. That difference has led to the automatic correlation between simple cycle peaking plants and aero engines.
|CAISO Duck Curve|
In addition to performance gaps between the technologies, the frame machines’ equivalent hours factor for starts was a large detriment to the engine for project evaluation.
Aeroderivatives had an advantage over frame engines when they were first introduced. Both technologies have realized improvements over the years, however, frames have realized greater improvements which has eroded the long held advantage of aero’s.
Within the last 5-10 years, frame gas turbines have made significant technological advancements which have been mainly driven by the larger frame models and adopted into the smaller sized products. Just a few examples of improvements are blade tip clearances, thermal barrier coatings, combustors, blade design and manufacturing processes. These advancements have led to both performance improvements applicable to the peaking market and also removal of the equivalent hours starts penalty.
Another aspect to consider for a simple cycle gas turbine plant is emissions control. In today’s market it is highly likely that emission control technology will be required, mainly selective catalytic reduction (SCR) technology. The next section will provide a high level summary of historical applications of SCR technology.
|CAISO Single Day Generation Profile|
III. SCR Implementation Snapshot
Frame engines have been challenging for SCR’s due to their high exhaust gas temperature and the potential impact on catalyst materials and life. That said, when looking at the projects that have given a bad reputation to frame engines and SCR’s, the majority of the problems have arisen from improper installation and/or engineering design. Table 3 shows highlights of frame engine SCR applications.
Early on complications were present due to catalyst manufacturing and plant engineering and construction issues. However, as catalyst technology has matured and design and installation practices have improved through experience, there has been demonstrated success for frame engine SCR’s.
Given the technology advancements of frame engines and the demonstration of successful SCR applications, the frame engines advantages can now be considered. The next section will show a levelized cost of electricity comparison between frame and aero engines.
IV. COE Comparison
Frame engines have always had a lower all in capex than aero engines, even after including a more expensive SCR system as shown in Tables 4 and 5. When operating a low number of hours as a peaking plant does, capital costs become a majority factor in determining a plant’s cost of electricity. As depicted in Table 4, the capital cost between the two technologies is considerable. In addition to the per unit basis advantage, there is also an absolute magnitude of the cost advantage, as shown in Table 5.
Although aero engines have an advantage when it comes to efficiency and Long Term Service Agreement (LTSA) costs, these factors become less of a contributor to the COE than the capital cost when considering low capacity factors and low fuel prices. It is widely expected that natural gas prices will stay low and this will continue to counteract the efficiency advantage of the aero. In addition with low capacity factors, the LTSA has a small impact at best on the overall COE (Figure 5).
Aero engines have long been the de facto technology solution for simple cycle peaking plants given their advantageous performance attributes for a peaking application. However, with modern frame gas turbines and the advancements they have realized over the years, the synonymy of aero engine and peaking application should be challenged. As this paper has outlined, frame engines have realized improvements for start times, ramp rates, turn down, emission levels and the LTSA basis. In addition to the gas turbine itself, the SCR technology and application have matured as demonstrated by the success of new builds and retrofits.
Each project has its unique challenges and characteristics that contribute to the eventual selection of a technology. Given the capital cost advantage and the ability of frame engines to now meet the needs of peaking applications across mid to large capacity ranges, these products should be seriously considered when determining technology selection.
Craig S. Brooker is a market research analyst at Mitsubishi Hitachi Power Systems Americas.