Over a long period of time, if you are “in” the stock market on the best days, and avoid the stock market on the worst days, your overall returns are astronomically higher than if you just put you money in and go away for a few decades.
This truism of playing the market has many variations but here’s one: If you missed the best 40 performing days between 1993 and 2013, a $10,000 investment in the S&P 500 shrunk to $8149; if you avoided the worst-performing days during that period, your $10,000 investment would have been almost 900% higher. Twenty years ago, this example was used in an industry presentation to illustrate for an audience of control-system engineers where electricity deregulation was headed. Today, this reality has come to pass, but with a few unexpected twists.
Imagine you are responsible for a fleet of simple-cycle gas turbines and combined cycles. The wind energy penetration in your balancing area is rising past 40%. Coal units are being retired because of uneconomic environmental investments. Your GT assets are operating in a casino-like environment, and wind enjoys the house odds. What’s more, your ability to make money when you are “in the market” (dispatched and online) is less than the chance of losing your shirt if you say you can be online and you fail.
In other words, it’s about reliability. You make money filling in around wind producers. Value isn’t about producing energy or getting paid for capacity. It’s about being able to start when you say you’re going to start, putting online whatever load the balancing authority demands, and remaining online until told to shut down. There is no wholesale market; it’s all day-ahead and real-time. Real time means pushing the start button within 30 seconds of receiving the call from the system operator.
It’s about extreme flexibility. In this world, it might be that a simple-cycle Frame 7 starts and stops up to seven times a day; three starts is a regular “day at the office.” Such GTs must be capable of 10-min start times, because that’s the threshold set by the system operator.
This need for flexibility ripples through the supply chain. The gas transmission folks are managing your starts and stops too. Your fuel contracts probably aren’t optimized for 20 separate one-hour fuel deliveries over the course of a week. Your large coal-unit brethren in the fleet are cycling from minimum to maximum load within an hour, and the smaller coal units start and stop daily. Every fossil unit in the fleet, in fact, operates with wind energy as its master.
You don’t really know from day to day or hour to hour if you are going to make any real money. What everyone does know is that the wind guys and gals get $38/MWh, according to one Midwestern power producer, because of the federal and state subsidies. Their risk has to do with the weather, not whether they’re going to get paid.
What the GT fleet manager also knows is that the penalties for not “doing what you say you’re going to do” are severe. You can lose an unbelievable amount of money in a few hours. In one balancing area, whole groups pulled their GT fleets out of the market because of losses incurred for “overpromising and under-delivering.” In other words, they took their money and left the casino licking their wounds.
Good luck to you if your combined cycle was intended for base-load operation. Example: A combined cycle that came online at the end of 2004 was designed for 40 to 50 starts per year and had an LTSA with the GT vendor to cover the expected high number of equivalent operating hours and high capacity factor. Plants like this often have a separate startup boiler and a relatively long startup/warmup period. They were not designed for cycling.
Under the new market/balancing area regime with significant wind penetration, the value of the example combined cycle’s two gas turbines changed dramatically. Both units had a total of 20 starts one year, 34 starts in 2014, and about that number the following year. One reason is that starting costs are high, since these units were never designed to start frequently; they were designed to be efficient at their maximum load point (like all baseload units).
In the brave new world of casino electricity, the GT units in the aggregate have been dispatched for energy only during one unseasonably hot summer over the last three years. F-class Frame 7s, capable of delivering 150 MW, might start up just to deliver 25 MW within 10 minutes, run for an hour, and shut down. Thus, the term “peaking machine” could become a misnomer. These are ancillary-services machines, despite historically low natural-gas prices.
What price O&M? Maintenance philosophies have to adapt when your assets essentially pick up the chips falling from the table. Plus, GT OEMs care less and less about supporting and servicing old peaking machines. Management is likely to allocate funds commensurate with your run time—that is, not much.
You need to find and hire the kind of people who can solve problems and fix equipment, comb through historical outage reports, and develop a long-term maintenance plan and strategy. Controls are a good place to start. Replacing old fuel valves with more flexible ones helps reduce start times. Electronic valve control assists to accelerate ramps on synchronization. Watch for dead spots in the hydraulics. They can plug up when a unit sits for months on end without operating. It may not be efficient to maintain a higher-than-normal lube-oil temperature when units are offline but it helps keep the machine ready when called upon.
Consider conducting borescope inspections semi-annually. Increase the frequency of electrical testing, especially for critical circuits, beyond the maintenance guidelines or even beyond what the failure history might suggest. The same holds true for combustion hot-gas-path inspections, generator testing, and rewinds. Be proactive about starting-package issues. In some cases, new control systems or changes to the existing control logic and permissives may be required to ensure that a GT can participate in the 10-min-start market.
Also critical is to adapt your maintenance strategies based on the larger fleet for that particular engine model, not just the operating history of your unit, or the like machines within your ownership. Failures and equipment maladies experienced by other owner/operators should be factored into operating, inspection, and maintenance routines. The value of participating in user groups cannot be under-estimated here.
Been here before. As a historical note, the casino electricity market may be relatively new to the electricity industry but the need for new powerplants to adapt to changing realities on the ground is not. In the 1970s, a fleet of large supercritical coal-fired units, also designed for base-load operation, were added to the grid. Many of these units were almost immediately relegated to cycling because of the nuclear units coming online (which were not allowed to cycle) and a reduction in electricity demand growth caused by recessions, energy crises, and the general economic malaise of that decade.
More recently, dozens of combined cycles installed in the 1997-2005 period were intended for base-load operation, or at least reasonably high capacity factors. They also were built when the forward price curve for natural gas was “reasonable.” The economy tanked in 2001, 9/11 occurred, and by 2007, gas prices were as high as $14/million Btu. Many of these brand new facilities were soon operating at capacity factors around 10-25%, rather than the expected 60-80%.
However, today’s situation is unique because of (1) higher renewable-energy penetrations in many parts of the county and (2) greater market mechanisms being applied in the balancing areas. The reality is that intermittent, less-predictable renewable energy assets are favored and are given highest priority for remaining online. They are replacing highly predictable, dependable base-load nuclear, coal, and gas-fired powerplants at the top of the queue.
Until further notice, renewable assets enjoy house odds. GRiD