Climate Act Renewable Intermittency Challenge

One of the difficulties associated with describing the challenge of the Climate Leadership & Community Protection Act (Climate Act) is that many of the concepts are difficult to describe to the general public.  Tyler Duren writing at Zero Hedge published an article, The Renewable Intermittency Challenge, that did a good job introducing the challenges associated with intermittency.  This post expands upon his article because I think he underestimates the difficulty of a solution.

I have been following the Climate Act since it was first proposed. I submitted comments on the Climate Act implementation plan and have written over 300 articles about New York’s net-zero transition because I believe the ambitions for a zero-emissions economy embodied in the Climate Act outstrip available renewable technology such that the net-zero transition will do more harm than good.  The opinions expressed in this post do not reflect the position of any of my previous employers or any other company I have been associated with, these comments are mine alone.

Climate Act Background

The Climate Leadership & Community Protection Act (Climate Act) established a New York “Net Zero” target (85% reduction and 15% offset of emissions) by 2050 and an interim 2030 target of a 40% reduction by 2030. The Climate Action Council is responsible for preparing the Scoping Plan that outlines how to “achieve the State’s bold clean energy and climate agenda.”  In brief, that plan is to electrify everything possible and power the electric gride with zero-emissions generating resources by 2040.  The Integration Analysis prepared by the New York State Energy Research and Development Authority (NYSERDA) and its consultants quantifies the impact of the electrification strategies.  That material was used to write a Draft Scoping Plan.  After a year-long review the Scoping Plan recommendations were finalized at the end of 2022.  In 2023 the Scoping Plan recommendations are supposed to be implemented through regulation and legislation.

Over many years New York electric planners have developed modeling procedures that project the resource adequacy necessary to maintain current reliability standards that keep the lights on when needed the most. The current reliability procedures were developed for generation resources that can be turned on, ramped up, ramped down, and turned off as needed, have well understood forced outage rates, and do not necessarily stop operating all at the same time.   Wind and solar resources do not have any of those characteristics which makes future reliability planning much more difficult.  This post looks at the problem of intermittency.

Intermittency Challenge

All quotations below are from the article The Renewable Intermittency Challenge.  To be fair my criticisms of this work are based on the presumption that intermittency is a significant problem when the energy system is intended to get to a net-zero target.  Author Tyler Duren introduces the challenge by giving an overview of the electric system:

The U.S. has a dynamic electricity mix, with a range of energy sources generating electricity at different times of the day.

At all times, the amount of electricity generated must match demand in order to keep the power grid in balance, which leads to cyclical patterns in daily and weekly electricity generation.

The graphic below, via Visual Capitalist’s Govind Bhutada and Sabrina Lam, tracks hourly changes in U.S. electricity generation over one week, based on data from the U.S. Energy Information Administration (EIA).

It may difficult to read the summary of the renewable intermittency challenge in the previous figure.  It says “Unlike conventional sources of electricity, wind and solar are variable and location-specific.  This is a concern for grid operators, especially as more renewable capacity is deployed”.  I agree with this description.

Duren goes on to explain that the electric load is met with three types of power plants.  He describes daily load cycles and the use of these power plants.  In my opinion, some peaking power plants are not normally used for daily load variations.  Some of these units only are operated at times of high loads like an extended period of hot weather and high loads.

The Three Types of Power Plants

Before diving in, it’s important to distinguish between the three main types of power plants in the U.S. electricity mix:

• Base load plants generally run at full or near-full capacity and are used to meet the base load or the minimum amount of electricity demanded at all times. These are typically coal-fired or nuclear power plants. If regionally available, geothermal and hydropower plants can also be used as baseload sources.
• Peak load or peaking power plants are typically dispatchable and can be ramped up quickly during periods of high demand. These plants usually operate at maximum capacity only for a few hours a day and include gas-fired and pumped-storage hydropower plants.
• Intermediate load plants are used during the transitory hours between base load and peak load demand. Intermittent renewable sources like wind and solar (without battery storage) are suitable for intermediate use, along with other sources.

This simplistic overview did not explain the difficulties facing a system that relies on intermittent wind and solar.  In an article co-authored with Russel Schussler we explained some of the issues with peaking power when most of the energy is supplied by wind and solar.

Duren goes on to show how electricity generation meets load on an hourly basis.

Zooming In: The U.S. Hourly Electricity Mix

With that context, the table below provides an overview of average hourly electricity generation by source for the week of March 7–March 14, 2023, in the Eastern Time Zone.

It’s worth noting that while this is representative of a typical week of electricity generation, these patterns can change with seasons. For example, in the month of June, electricity demand usually peaks around 5 PM, when solar generation is still high, unlike in March.

Natural gas is the country’s largest source of electricity, with gas-fired plants generating an average of 176,000 MWh of electricity per hour throughout the week outlined above. The dispatchable nature of natural gas is evident in the chart, with gas-fired generation falling in the wee hours and rising during business hours.

Meanwhile, nuclear electricity generation remains steady throughout the given days and week, ranging between 80,000–85,000 MWh per hour. Nuclear plants are designed to operate for long durations (1.5 to 2 years) before refueling and require less maintenance, allowing them to provide reliable baseload energy.

On the other hand, wind and solar generation tend to see large fluctuations throughout the week. For example, during the week of March 07–14, wind generation ranged between 26,875 MWh and 77,185 MWh per hour, based on wind speeds. Solar generation had stronger extremes, often reaching zero or net-negative at night and rising to over 40,000 MWh in the afternoon.

Because wind and solar are often variable and location-specific, integrating them into the grid can pose challenges for grid operators, who rely on forecasts to keep electricity supply and demand in balance. So, what are some ways to solve these problems?

Duren suggests that these challenges can be solved.  His suggestions will be the focus of the remainder of this post.  Rafe Champion describes the issue of wind droughts that undermines the ability of these solutions to work in a system that relies heavily on wind and solar.  If the renewables are only intended to augment the existing system much of the following discussion is appropriate.  However, the only reason to install wind and solar is to mitigate climate change which requires a net-zero solution so I do not believe there is any reason to consider limited penetration of renewables.

Solving the Renewable Intermittency Challenge

As more renewable capacity is deployed, here are three ways to make the transition smoother.

• Energy storage systems can be combined with renewables to mitigate variability. Batteries can store electricity during times of high generation (for example, in the afternoon for solar), and supply it during periods of peak demand.
• Demand-side management can be used to shift flexible demand to times of high renewable generation. For instance, utilities can collaborate with their industrial customers to ensure that certain factory lines only run in the afternoon, when solar generation peaks.
• Expanding transmission lines can help connect high-quality solar and wind resources in remote regions to centers of demand. In fact, as of the end of 2021, over 900 gigawatts of solar and wind capacity (notably more than the country’s current renewable capacity) were queued for grid interconnection.

Energy Storage

Duren makes a common mistake when describing the electric system. He only talks about average conditions.  His description of energy storage systems only addresses daily variations: “Batteries can store electricity during times of high generation (for example, in the afternoon for solar), and supply it during periods of peak demand.”  The bigger problem is that wind and, to a lesser extent, solar can also be subject to longer periods of reduced output that complicates energy storage requirements.  In a net-zero electric system I believe that wind droughts are a fatal flaw because existing energy storage technology is too limited and too expensive.

Francis Menton writing at the Manhattan Contrarian described work by Bill Ponton that addressed energy storage requirements for the Climate Act transition plan over longer time periods than a day.  They evaluated “Initial Report on the New York Power Grid Study”  which includes the following table of how much wind, solar and storage capacity needed to meet the net-zero transition targets of the Climate Act .

Menton describes the energy storage provisions:

But far more absurd is the provision in this Report for prospective energy storage. Note the numbers in the table above — 3 GW in 2030 and 15.5 GW in 2040. As usual they leave out the duration of the batteries. But Ponton wrote to the lead author of the Report from the Brattle Group, a guy named Hannes Pfeifenberger, to get the information. Result:

I asked one of the principal authors of the NY Power Grid study report, Hannes Pfeifenberger, how did he intend to balance fluctuations in wind power and he stated that the biggest factor was 17 GW of battery storage with a maximum duration of 6-hr, totaling 102 GWh. His response is surprising. I calculated that with wind power capacity of 84 GW,  there was 59,851 GWH of wind energy curtailed and 48,071 GWH of gas turbine energy used. In theory, the curtailed wind energy could be stored and then subsequently discharged to substitute for the energy provided by the gas turbines, but would require energy storage of 12,000 GWH. 

102 GWh versus 12,000 GWh. So, as usual with the studies we can find for places like New York and California, they’re off on the storage requirement by a factor of more than 100.

I have tried to make my own estimates of energy storage requirements and they are the same order of magnitude as described here.  The main point is that Duren does not address this problem at all.

Demand-Side Management

Simplistic evaluations of net-zero transition programs also suggest that reducing load through programs like demand-side management can be a viable solution.   There are two issues.  The first is that net-zero programs refer to the entire economy and emission reductions from transportation and residential heating, hot water, and cooking, rely on electrification which necessarily increases load.  This is a particular problem when loads are highest.  Duren writes that “Demand-side management can be used to shift flexible demand to times of high renewable generation. For instance, utilities can collaborate with their industrial customers to ensure that certain factory lines only run in the afternoon, when solar generation peaks.”  There is a limited amount of load shifting possible when temperatures are coldest, electric vehicle charging and battery capabilities are lowered, and everyone needs electricity to keep warm.

Expanding Transmission Capabilities

Another favorite solution of naïve energy analysts is predicated on the concept that the wind is always blowing somewhere.  Duren writes: “Expanding transmission lines can help connect high-quality solar and wind resources in remote regions to centers of demand. In fact, as of the end of 2021, over 900 gigawatts of solar and wind capacity (notably more than the country’s current renewable capacity) were queued for grid interconnection.”  In general that is true and that might work for a net-zero electric system on average.  Unfortunately, the coldest and hottest weather, and thus the highest load, is strongly correlated with high pressure systems that also have the lowest wind resources. 

In a presentation describing my skeptical concerns about the Climate Act I addressed this problem in detail.  In brief, consider the following weather map of February 17, 2021 during the period of the Great Texas Freeze. This event was associated with an intensely cold polar vortex huge high-pressure system.  Remember that winds are higher when the isobars are close together.  On this day there are light winds from New York to the southeast, west, and north including the proposed New York offshore wind development area.  There are packed isobars in northeastern New England, in the western Great Plains, and central Gulf Coast where wind resources would be plentiful.  For New York to guarantee wind energy availability from those locations, wind turbines and the transmission lines between New York and those locations would have to be dedicated for our use.  Otherwise, jurisdictions in between would claim those resources for their own use during these high energy demand days.  It is unreasonable to expect that this could possibly be an economic solution.

Conclusion

I liked the introduction of Duren’s article.  He gives a good overview of the balancing act necessary to constantly match generating resources with the load.  The graphics illustrate his points well too.  He does not explicitly address net zero targets and that avoids addressing the problem that what works for today’s systems will not work when the system relies on intermittent renewables.  In the context of a net-zero electric system that eliminates fossil-fired generation, I do not believe his conclusion that there are readily available solutions that address the intermittency challenge is correct.

If the future electric system has to rely on weather-dependent resources, then the question of worst-case availability has to be addressed.  Analyses done to date in New York State have all shown that there are wind lulls in the winter that are difficult to address using existing energy storage technology.  The Integration Analysis and the New York Independent System Operator Resource Outlook analyses of future resource requirements have both concluded that the solutions recommended by Duren are inadequate and a new dispatchable emissions-free resource (DEFR) must be developed to address the wind lull intermittency problem.  The Resource Outlook notes “The lead time necessary for research, development, permitting, and construction of DEFR supply will require action well in advance of 2040 if state policy mandates under the CLCPA are to be achieved”. 

In addition to the challenge of developing an entirely new resource, in order to determine the resource capabilities necessary for the worst-case, a comprehensive analysis of wind and solar resource availability is needed.  If that analysis does not get the right resource availability, then there will not be enough energy available to supply everyone who needs it during the worst conditions.  If either of these challenges is not met adequately, people will freeze to death in the dark.