Dispatchable generation refers to sources of electricity that can be started or brought on-line at the request of power grid operators, according to demand on the grid. Some dispatchable clean energy sources are: hydroelectric, geothermal, nuclear, ocean thermal. Examples of non-dispatchable clean energy sources are wind, solar, and ocean waves. All forms of energy storage are designed to dispatch power on command. Examples include lithium batteries, flow batteries, pumped hydro, compressed air, spinning masses, capacitor banks, hydrogen, to name a few.
The predominant, legacy dispatchable energy source is the peaker plant (gas turbine). As recently as 2015 there were hundreds of these plants sprinkled around California. The majority of dispatch power in California continues to be gas-fired.
Why is it important?
The primary purpose of dispatchable generation is to provide load-matching and peak-matching capabilities on the grid. Electricity on the grid does not exist statically, it’s dynamic and always in motion. Load-matching is the ability to balance electrons on the wires between multiple sources and multiple recipients: generation must always equal consumption. If the load isn’t balanced at a particular point, a lightning-bolt-like event will occur as electrons discharge to the ground (possibly correlated with a power blackout somewhere else).
Peak-matching is a form of load-matching, but occurs over much shorter intervals of time and so is given its own name. One example is the daily 3-hour ramp during the late afternoon that we hear about so much from utilities. California’s balancing authority (called the Independent System Operator) forecasts daily electric demand and issues system dispatch orders, so as to maintain equilibrium between generation and demand on the grid. The process is iterative as the day wears on; the magnitude of dispatch orders shrink or expand to match the observed load.
How does it work?
Dispatch plants are useful only insofar as they can respond quickly to changing grid conditions. Each kind of plant varies in start-up time. The fastest plants to dispatch are grid battery systems (both lithium & flow) which can dispatch in milliseconds. Hydroelectric systems respond in about 1 minute. Peaker plants’ response time (following business-as-usual) is 5-10 minutes. But business-as-usual requires that these plants be kept running all day long in standby mode (i.e. combusting gas, but disconnected from the grid) until they are called for. At that time the turbine then spins-up to operational speed and connects to the grid, about a 5-10 minute procedure. Cold-starts require at least 30-minutes before dispatching power.
California’s power load and dispatch response (using peaker plants) over a 9-year period is shown in the “duck curve” figure below. As more solar enters the system, less fossil fuel combustion is needed in the daytime, thereby increasing the size of the duck’s belly. More non-solar power is needed in the afternoon as the Sun sets, yielding a rapid ramp-up of dispatch power.
How will it work with clean energy?
The next figure is a simulation of the US grid in 2050, assuming 100% zero-carbon energy. Baseload power is represented as relatively flat lines at the bottom of the graph. Baseload power is the background generation that occurs reliably around the clock, day after day. The other colors represent real-time dispatch of the system in order to satisfy the minute-by-minute load. This model is illustrative for suggesting the way California’s grid will behave in 2045, after achieving 100% zero-carbon energy. Large power sources will be rare and so baseload power generation won’t dominate like it does now (or did in the recent past). Power dispatch will dominate because of 1) variability of solar and wind generation, and 2) extremely large number of modest-sized power sources (i.e. cf’d today). Graph from Vibrant Clean Energy Co.