The Cost Of Wind And Solar Intermittency

By Earl J. Ritchie, Lecturer, Department of Construction Management

Until relatively recently, generation of electricity with wind and solar has not been cost competitive. Growth has largely been due to subsidies and renewable energy mandates. Due to decreasing cost, wind and solar are now cost competitive with fossil fuels in favorable locations.

The continuing decrease in wind and solar costs is a very positive development. However, this trend may reverse as the percentage of variable renewable energy (VRE) energy that isn’t available on-demand but only at specific times, such as when the wind is blowingreaches high levels. Countries such as Germany that have integrated significant amounts of wind and solar have already seen price increases.

The levelized cost of electricity

Comparisons of electrical generation cost are usually based on the so-called levelized cost of energy (LCOE), an estimate of the total cost of generation expressed in dollars per megawatt hour ($/WMh). The calculation includes capital costs, operating and maintenance costs and fuel cost. It is affected by assumed utilization rate and interest rates.

The most widely cited levelized cost estimates are those of the U.S. Energy Information Agency (EIA) and the investment firm Lazard. Although these estimates are useful for comparison, they exclude such costs as network upgrades, integration and transmission, which can become significant as renewables penetration increases. As the International Energy Agency (IEA) put it in the context of integrating variable renewable energy, “comparison based on LCOE is no longer sufficient and can be misleading.”

Levelized cost estimates are based on a large number of assumptions, not least of which is the future cost of fossil fuels. There are some differences in these estimates, with Lazard showing unsubsidized utility scale solar and onshore wind as competitive with natural gas and the EIA not.

The table shows national averages. For wind and solar, location is very important; they are in places locally cheaper than natural gas combined cycle. For the purposes of this discussion, these differences are not significant. The more important point is the added cost of factors not included in the levelized cost.

The sources of integration costs

As described by Mark Delucchi and Mark Jacobson, “any electricity system must be able to respond to changes in demand over seconds, minutes, hours, seasons and years, and must be able to accommodate unanticipated changes in the availability of generation.” Traditionally, this is handled by base load and peak load plants, which handle the minimum load and increases above that level, respectively. This is an oversimplification, since supply is managed by the minute using a variety of sources with different response times.

Wind and solar are non-dispatchable, meaning that they are not under the control of the operator. They only generate electricity when the wind blows or the sun shines. This adds integration costs, shown conceptually below.

Source: Ueckerdt, 2015

 When variable sources are a small fraction of electricity supply, the cost of integration is low. The current level of deployment is below thresholds where the cost of dealing with intermittency becomes significant.

There are numerous possible solutions to intermittency. These include diversification, redundancy, storage and demand shifting. That redundancy and storage add cost is obvious. Diversification also adds cost in control equipment and transmission capability between geographically separated sources.

Demand shifting can theoretically lower cost by reducing the peak capacity needed. It is often discussed jointly with efficiency improvement under the term demand-side management.

One issue in demand management is illustrated in this graph of daily load for a location in Australia. Solar is only available when the sun shines and peaks around midday. As solar generation increases, the average load on the remainder of the system decreases, but the peak is barely affected. Dispatchable sources must make up the difference between the midday low and the evening and morning peaks. This relationship is called the “duck curve.”

Source: Ledwich 2015

Measures to shift usage from peak periods include education, jawboning, differential pricing and control of end use by the utility through the smart grid. Education, jawboning and even differential pricing have had limited success to date. Time of day pricing and end-use control require a smart grid, with attendant cost.

Wind power typically will generate throughout the day, but it has its own limitations. It is less predictable, more variable over short periods than solar, may be seasonal and may need to be shut down when the wind is too strong.

The graph below shows generation for one day on the island of Crete. Renewables penetration reaches a peak of 60%, accommodated by curtailment of diesel and gas generation. Even so, average annual renewable share is only 20%, and some difficulties were encountered during peak renewables generation periods.

The Crete example is typical of existing systems in that balancing is done with fossil fuels. Balancing may also be done by dispatchable renewable energy, primarily hydroelectric and biomass, and with storage.

What’s the best generation mix?

Due to the wide variety of generating sources and unique local circumstances, there is considerable flexibility in the design of generating systems. The trade-offs in cost and environmental benefit are complex.

Hundreds of studies which address increasing the share of renewables have been published. These vary greatly in scope and sophistication. Some do not include cost analysis or ignore integration costs. Adequate analysis of high levels of variable generation requires that balancing demand within short time frames be included.

The sample of published scenarios below illustrates the wide range of possible combinations. Wind and solar range from less than 20% to over 80%. The mix is influenced by availability of other sources, and by ideology.

Source: Modified from Cochran 2014

Big differences result from design choices, such as whether expansion or retention of some fossil fuels are included. Accepting periods of inadequate capacity is also a factor.

Most scenarios with high percentages of renewables rely on substantial reduction in growth of electricity demand. It’s questionable how realistic this is, particularly if strong growth in electric automobiles is anticipated.

What is the integration threshold?

There is no threshold, per se. The cost of managing intermittency is nonlinear and depends upon the mix and location of dispatchable and non-dispatchable sources, the match of local demand patterns with variable source pattern, and various other factors.

Based on model studies of Germany and Indiana, Falko Ueckerdt found integration costs began to become significant at 20%. As of 2015, only four countries have variable renewable energy over 20%.

Hawaii Electric recently approached 50% renewables; however, the share of wind and solar was only about 15%. Even so, they have requested a 6.9% rate increase based partly on the cost of renewables integration, and estimate the cost of grid upgrades necessary to reach 100% renewables as $8 billion.

Champions of wind and solar have characterized integration cost estimates as ploys to discourage renewable energy, but integration costs are real.

Isn’t it being done already?

The poster child for variable renewable energy is Denmark, reported to be over 50% in 2015. Denmark’s success is often used to illustrate that high levels are readily achievable. This is misleading in that Denmark is a small country tied into the European grid. Variable wind power is balanced with hydroelectric and other sources in adjacent countries. De facto share for the system is lower. Denmark’s installed wind capacity ranks ninth among EU countries and represents less than 4% of EU.

Source: EIA 2016

Germany’s combined wind and solar has the largest capacity in Europe and is second highest per capita. Despite Germany’s progress, the share of variable renewable energy for electrical generation is less than 25% and has been achieved at significant cost. The renewable energy surcharge is 22% of household electricity price.

Even at relatively low levels of renewables share, there is a clear correlation between the share of variable renewable energy and the retail price of electricity. This is largely due to feed-in tariffs and net metering, which transfer renewable subsidies costs to the retail customer.

 The range of published integration cost estimates at higher shares of wind and solar is very broad and dependent upon both parameter assumptions and model structure. I will discuss these in a later post.


Earl J. Ritchie is a retired energy executive and teaches a course on the oil and gas industry at the University of Houston. He has 35 years’ experience in the industry. He started as a geophysicist with Mobil Oil and subsequently worked in a variety of management and technical positions with several independent exploration and production companies. Ritchie retired as Vice President and General Manager of the offshore division of EOG Resources in 2007. Prior to his experience in the oil industry, he served at the US Air Force Special Weapons Center, providing geologic and geophysical support to nuclear research activities.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: