« Back to Energy Insider
Do Northern California and Southern Australia Herald the End of Baseload Generation?
by Bob Shively, Enerdynamics President and Lead Instructor
Since the mid-20th century, it's been widely accepted wisdom: For electric utilities to economically serve customers they must build generating systems rooted in large baseload plants. But with the recent growth of renewables in various global markets, the need for large baseload plants is in question. Developments in markets such as Southern Australia and Northern California suggest that the concept of baseload generation may soon be considered a thing of the past.
How Load Historically Has Been Served
The following depiction of a dispatch stack is representative of most utilities around the world:
Baseload: Baseload units serve the minimum level of customer loads. These large units have low operating costs but typically have a high initial capital cost. The units operate most of the hours out the year at or near rated capacity allowing the large capital costs to be spread across large amounts of megawatt hours (MWh). Capacity factors (which represents the amount of output relative to rated capacity) typically range from 70 to 90+%. Since the units are typically turned on and left on, fast start times and the capability to ramp over a wide range of output are not important design factors. Typical units used for baseload historically include run-of-the-river hydro, coal and gas steam turbines, and nuclear steam turbines.
Intermediate load: The next tranche of load occurs for the approximate 16 hours in a day when the bulk of the customers are up and actively using power. Units that serve this type of load must be flexible enough to ramp up and down in response to the changing load curve, and ideally they can be turned on in the morning and shut down in the late evening. Since the units run fewer hours, variable costs higher than baseload units are generally considered acceptable, and lower capital costs are desired since the costs must be spread across fewer operating hours. Typical capacity factors range from 30 to 60%. Historically, intermediate loads have been served by coal or gas steam turbines that can run at partial loading during the off-peak hours and then ramp up to serve these loads. In recent years, this role has been filled by gas combined-cycle turbines.
Peak load: The last tranche of load occurs for only a few peak hours, typically late in the day in the summer and in the morning and evening in winter. Peak load units must have quick start times and be able to ramp up and down rapidly. These units are often run only a few hours during the day, and thus it is acceptable to have high variable operating costs. But capital costs are ideally kept low since the cost of units is spread over so few hours. Typical capacity factors may be 5 to 10% or even lower for some resources. This role is frequently performed by hydro and pumped hydro resources that can be controlled, gas combustion turbines, and, in recent years, controllable demand response loads.
Serving Loads with Growing Penetrations of Renewables
Adding significant amounts of wind and solar to the grid results in a paradigm shift. Now a significant amount of electric supply is not controllable and is variable in nature. The new resources share some interesting characteristics with typical baseload units and in other ways are the exact opposite.
From a cost standpoint, renewables share the baseload characteristic of low operating costs but higher upfront capital costs. This means that utilities prefer to take the output of renewable units whenever it is available. And given the operating restrictions on baseload units, system operators do not like to ramp the units much meaning that controllability of baseload units is low. This is similar to renewables (at least from the standpoint of the ability to ramp up, renewables can usually be ramped down by simply curtailing their power although this is not a popular alternative). But the key difference is variability. Traditional baseload units are expected to have similar output every hour (except during planned or forced maintenance outages), whereas renewables are a variable resource whose output depends on weather.
As renewables grow, system operators shift the focus on operating traditional units to serve all loads to operating them to serve the “net load’ – that portion of load not served by renewable resources. The so-called Duck Curve from California shows how this evolves as more renewables are added:
Source: California Energy Commission Staff 
Forecasts for some scenarios suggest that eventually solar power may push net loads in some regions to below zero. To maintain system balance, operators in this case would either need to export power to another grid, have available storage options, have flexible loads willing to use more power, or curtail solar output.
Will Renewables Be the Death of Baseload?
In 2015, Australian professor Mark Diesendorf raised eyebrows in the utility industry with his paper titled “Do We Need Base-load Power Stations?” The paper suggested that “base-load power stations are unnecessary to meet standard reliability criteria for the whole supply-demand system, such as loss-of-load probability or annual energy shortfall.” Diesendorf suggested that future dispatch curves may look more like this than the traditional curve shown above:
Source: Energyscience.org 
In the last few months, it has become clear that Diesendorf’s thoughts are not just idle speculation. In May, the state of South Australia closed its last baseload unit, the 520 MW Northern coal unit. Now, the dispatch curve in the state shows large amounts of renewables supplemented with flexible gas generation:
Source: RenewEconomy 
Later in June the major utility in Northern California, Pacific Gas and Electric (PG&E), announced a settlement that will result in closing their 2,240 MW Diablo Canyon nuclear unit by 2025. The settlement outlined PG&E’s plan to replace Diablo’s power with a mix of reduced loads through energy efficiency, flexible loads, storage, and renewable power. Once this occurs, one of the largest utilities in the U.S. will be operating without any traditional baseload units.
In most cases, major infrastructure such as electric grids take many decades to transition once new technologies are introduced. Many regions will continue to use baseload units for years to come. But markets such as Southern Australia and Northern California show us that, in some cases, transitions happen more quickly than we expect. And yes, it is possible to run an electric system without baseload units.
 For more explanation of the Duck Curve see Energy Currents article Renewables Require System Operators and Designers to Rapidly Respond to Changing Load Curves at https://blog.enerdynamics.com/2016/06/20/renewables-require-system-operators-and-designers-to-rapidly-respond-to-changing-load-curves/
 Available at: http://www.energyscience.org.au/BP16%20BaseLoad.pdf
Giles Parkinson, Wind and solar power become the new “base load” power for South Australia, May 16, 2016, available at http://reneweconomy.com.au/2016/wind-and-solar-become-new-base-load-power-for-south-australia-99364
 See Joint Proposal for the Orderly Replacement of Diablo Canyon Power Plant with Energy Efficiency and Renewables, available at http://www.pge.com/includes/docs/pdfs/safety/dcpp/MJBA_Report.pdf
 See for instance this blog by Vaclav Smil: http://blogs.scientificamerican.com/the-curious-wavefunction/vaclav-smil-e2809cthe-great-hope-for-a-quick-and-sweeping-transition-to-renewable-energy-is-wishful-thinkinge2809d/