Feature Article: Joe Biden Wants to Remove Carbon From US Electricity Generation by 2035. Is That Realistic?

In 2019, the United States consumed 4,118,000 gigawatt hours of electricity (4,118 billion kilowatt hours, or 4,118 terawatt hours).

By generation fuel, 63% was produced using fossil fuels (24% coal, 38% natural gas, 1% other). Nuclear accounted for 20%, and Renewables the remaining 17% (7% wind, 7% hydro, 2% solar, 1% other).

Realizing the Biden plan’s goals will require some combination of four changes:

1) A large increase in nuclear generation (very unlikely);

2) A large increase in the use of carbon capture and storage technologies (which, given current technologies, would produce a substantial increase in electricity prices);

3) A very large increase in generation from wind and solar; and

4) A very substantial increase in the use of utility scale battery storage technologies (both to reduce the use of gas fired plants that run to meet demand peaks, and to support the expansion of intermittent generation technologies like wind and solar).

The expansion of utility scale battery storage is by far the most uncertain of these options, yet seems to be the critical assumption that underlies the Biden plan. Given its disruptive potential, it is important to understand the underlying issues.

Let’s start by clarifying the confusing terminology used battery storage discussions. Megawatts are a measure of the power rating of a battery – how fast it can discharge stored energy. The more power that can be released over a given period of time, the more powerful the battery.

In contrast, Megawatt Hours are a measure of how much energy is storied in the battery – how much energy it can discharge.

Mathematically, a battery’s Megawatt Hours equal its power times the length of time (in hours) it can discharge its full power.

These highlight a fundamental battery tradeoff: You can get full power for short period of time, or a lower amount of power for a longer period of time. For example, a 60MW battery that stores 240 MWh of energy can discharge 60MW for four hours, or 30MW for eight hours.

Here’s an example. There are basically two types of power generation units. “Baseload” units have a high rate of capacity utilization, while “Peaking” units tend to be older and less efficient, and only generate power when daily electricity demand is at its highest point (note that these peak periods are of varying lengths, depending on the season and location of demand – e.g., areas with large cities tend to have longer peaks).

In the United States in 2018, almost 90% of electricity generated from natural gas was produced by highly efficient combined cycle baseload plants, while only 10% was produced by less efficient (and often ore polluting) combustion turbine peaker plants.

According to a recent study, New York’s gas fired peaker plants have a capacity of 4,500 Megawatts (4.5 gigawatts). The study’s authors concluded that, if cost efficient batteries that could discharge for 6 hours were installed, they could replace 275 Megawatts of this gas fired peaker capacity. Batteries that could discharge for 8 hours could replace 500 MW of gas fired peaker capacity. And if the 6-hour batteries were paired with new solar generation facilities, 1,804 MW of gas peaker capacity could be displaced.

That sounds great, doesn’t it? That’s certainly the kind of story you see a lot these days, extolling the brave new (greener) world of renewables (especially solar) paired with batteries that is just around the corner.

But how real is this vision?

Let’s look at six issues that you don’t read about as much: (1) Grid control; (2) Profitability of the battery storage revenue model; (3) Batteries’ cost; (4) Batteries’ long term safety and reliability; (5) Batteries’ use of rare earth minerals; and (6) How climate change will affect future solar radiation and wind.

Grid Control

Buried in the New York study was this critical sentence: “We did not consider [battery] charging constraints if multiple gas fired peakers were removed” and replaced with solar generation. The underlying assumption of pairing batteries with solar generation is that there will be times during the day when all the generating facilities on a grid will be producing more electricity than is required to satisfy total demand. During those periods, it is assumed that solar facilities will be allowed by the system operator to recharge their paired battery.

But by regulation, system operators are generally required to “dispatch” (i.e., take power from) producers in order of their cost, with the lowest cost facilities (e.g., nuclear plants) being dispatched first, and the most expensive ones (e.g., gas peakers) dispatched last. What happens if solar is so cheap that is frequently dispatched and has insufficient time to recharge the battery that is supposed to meet peak power demands (e.g., late on a winter afternoon when the sun has gone down and people come home and turn on a lot of electrical appliances)?

While it is not yet clear how regulators will make this tradeoff, it seems very likely that it will end up being litigated.

A closely related issue is the minimum number of hours that regulators and system operators will require battery storage operators to be able to provide. For example, a four hour minimum is likely to result in many more battery storage facilities than an eight hour requirement (which is why the PJM system operator has recently set an 11 hour requirement for storage operators to have grid access – a decision which is also getting challenged: see, “Capacity Value of Energy Storage in PJM” by Astrape Consulting for the US Energy Storage Association and the Natural Resources Defense Council).

More broadly, as you add more intermittent (or variable renewable) generating sources like wind and power to an electricity grid, along with fast responding batteries in multiple locations, maintaining control of that grid becomes much more challenging.

In essence, grid control involves the simultaneous solution of four problems:

1) Power generation and transmission capacity must be sufficient to meet peak demand for electricity

2) Power systems must have adequate flexibility to address variability and uncertainty in demand (load) and generation resources

3) Power systems must be able to maintain steady frequency

4) Power systems must be able to maintain voltage within an acceptable range.

Multiple “smart grid” projects seek to meet these challenges; however there are still obstacles they must overcome to achieve their goals (e.g., “The Role of Smart Grids in Integrating Renewable Energy” by the US National Renewable Energy Laboratory, and “Smart Grid: Status and Outlook” by the Congressional Research Service).

Profitability of the Battery Storage Revenue Model

Perhaps the most famous battery storage facility in the world is The Hornsdale Power Reserve in South Australia, more commonly known as the Big Tesla Battery.

Built in 2017, it can (before a planned expansion) deliver 129 Megawatt Hours of energy for about one hour.

70MW of its 100MW power is used for maintenance of frequency and voltage control (i.e., “grid stability”, or “Frequency Control and Ancilliary Services”, or FCAS).

30MW is used for “load shifting” – storing excess electricity generated by the adjacent wind farm, and sold for short periods into the market for a higher price during peak demand periods, in competition with gas peakers and other battery storage facilities.

Initially, provision of FCAS services was very profitable; however, as more battery storage facilities have opened competition has increased.

Moreover, so called “park spreads” from storing and reselling energy have been thin, due to both “round trip” efficiency losses (only about 80% of the originally generated energy can be resold to meet peak demand), and margin pressure due to competition from other peak suppliers (e.g., other battery storage operators and gas generators with excess capacity they can ramp up).

On the bright side, a recent review by the US Energy Information Administration (“Battery Storage in the United States: An Update on Market Trends”) identified a larger number of potential sources of revenue for battery storage operators:

• Frequency regulation helps balance momentary differences between electricity demand and supply within the transmission grid, often in order to help maintain interconnection frequencies close to 60 Hertz.

• Spinning reserve is the unused dispatchable generating capacity of online assets that provides grid frequency management, which may be available to use during a significant frequency disturbance, such as during an unexpected loss of generation capacity. This reserve ensures system operation and availability. Dispatchable generators are those that can be turned on or off in order to meet immediate needs of the system.

• Voltage or reactive power support ensures the quality of power delivered by maintaining the local voltage within specified limits by serving as a source or sink of reactive power (the portion of electricity that establishes and sustains the electric and magnetic fields of alternating‐current equipment).

• Load following supplies (discharges) or absorbs (charges) power to compensate for load variations—this application is a power balancing application, also known as a form of ramp rate control.

• System peak shaving reduces or defers the need to build new central generation capacity or purchase capacity in the wholesale electricity market, often during times of peak demand.

• Arbitrage occurs when batteries charge during periods when electrical energy is less expensive and discharge when prices for electricity are high, also referred to as electrical energy time‐shift.

• Load management provides a demand side customer‐related service, such as power quality, power reliability (grid‐connected or microgrid operation), retail electrical energy time‐shift, demand charge management, or renewable power consumption maximization (charging the battery storage system during periods when renewable energy is greatest so as to consume the maximum renewable energy from the battery system, i.e. charging with solar during the day or charging with wind during high wind periods).

• Storing excess wind and solar generation reduces the rate of change of the power output from a non‐dispatchable generator in order to comply with local grid requirements related to grid stability or prevent over production or over‐production penalties. Non-dispatchable generators cannot be turned on or off in order to meet immediate needs and are often intermittent resources (generators with output controlled by the natural variability of the energy source, for example wind and solar).

• Backup power, following a catastrophic failure of a grid, provides an active reserve of power and energy that can be used to energize transmission and distribution lines, provides start‐up power for generators, or provides a reference frequency.

• Transmission and distribution deferral keeps the loading of the transmission or distribution system equipment below a specified maximum. This application allows for delays in transmission upgrades, avoids the need to upgrade a transmission system completely, or avoids congestion‐related costs and charges.

• Co‐located generator firming provides constant output power over a certain period of time of a combined generator and energy storage system. Often the generator in this case is a nondispatchable renewable generator (for example, wind or solar).

Batteries’ Cost

The cost of battery storage has declined in recent years. Most recently, Lazard and Company estimated that the levelized cost of storage (LCOS, which includes associated costs beyond the battery itself) for a 100 MW battery that could delivery this power for four hours (i.e., that could deliver 400 Megawatt Hours of energy) is now between $189 and $325 per Megawatt Hour, without subsidies.

In comparison, a new combined cycle gas generating plant costs between $44 and $68 per Megawatt Hour.

In “Cost Projections for Utility-Scale Battery Storage”, NREL most optimistic case estimates that by 2030 the cost for a four-hour system could fall to $124 per Megawatt Hour ($207 middle case), and $76 per Megawatt Hour ($156 middle case) by 2050.

To put this in perspective, in March this year, Florian Mayr published an analysis of a recent solar plus storage deal in Arizona (“Battery storage at US$20/MWh? Breaking down low-cost solar-plus-storage PPAs in the USA”). After a masterful piece of financial analysis, he concluded that behind the deal’s complicated terms, the estimated cost of the storage was about $310, which he noted was “within the rage of aggressive, but realistic quotes we observe in the industry today.”

But that is in a state that has optimal conditions for solar plus storage installations. It is clear that the levelized cost of storage is going to have to come down much more to support more widespread deployment of this model.

Another aspect of the cost issue was highlighted in the aforementioned EIA report: “There are two major challenges in determining the profitability and cost of battery storage systems.

“First, quantifying the competitiveness of a battery storage technology with other technologies operating on the grid must consider the individual markets that the storage technology is planning to be used in and what revenue opportunities exist for the technology.

“The second challenge involves the degradation of the system over time, which is the lasting and continuous decrease in either a battery’s power or energy performance or both and is linked to use or age of a battery component or system.

“The performance can be characterized by the full cycle power input and output at an agreed‐upon charge/discharge rate. There are two general options that can be employed to ensure reliable performance during a storage system’s lifetime:

(1) Overbuilding: adding more storage or discharge capacity behind the inverter than is needed, so that as the system ages it will maintain a capacity at or above the contracted capacity required of the system.

(2) Continual Upgrades: replacing some portion of the storage system to maintain the agreed‐upon performance during its lifetime.

“The two approaches to meeting performance requirements affect the installed capital costs of the system. Overbuilding storage capacity leads to a higher initial installed capital cost, while continual upgrades lead to higher operation and maintenance costs throughout the lifetime of the storage facility. Therefore, comparing only the normalized capital cost of various battery systems, [as we have one in this analysis] does not capture the variation in the lifetime costs.”

Batteries’ Long Term Safety and Reliability

In “Why Is the Utility Industry Less Bullish on Grid-Scale Storage?”, Kavya Balaraman observed that, “In Utility Dive's 2020 State of the Electric Utility survey, 27% of participants said they expect their organization will significantly increase grid-scale battery storage in the next 10 years — a significant reduction from 37% in 2018, and 34% in 2019.”

Balaraman noted that, “Matthew Raiford, manager of the Consortium for Battery Innovation, a research organization focused on lead batteries, said that ‘there’s also the looming issue of safety — the last couple of years have witnessed some high-profile safety-related issues with battery storage, including an explosion at an Arizona Public Service facility last April and multiple storage-related fires in South Korea…I would venture to guess that over the next few years, there will be a refocusing in the market, looking at things like safety, reliability and the kind of technical economics of utilizing these systems’”.

Raiford’s quote is familiar to anyone who has tried to sell a new technology to a utility company. They are, quite rightly, very conservative customers. They know all-too-well that regulators and customers are risk averse, and that they face significant downside costs for self-inflicted errors. Despite the enthusiasm of battery storage supporters, don’t expect utilities and their regulators to jump n the bandwagon any time soon.

Batteries’ Use Of Rare Earth Minerals

Lithium-ion batteries are the most widely used technology in grid-scale storage today. However, their manufacture relies on continuing supplies of reasonably priced lithium, cobalt, and a number of rare-earth minerals.

In light of growing US-China tensions, the continuing supply of the rare-earths cannot be guaranteed. Moreover, supplies of both lithium and cobalt are highly concentrated in a small number of countries (e.g., the Democratic Republic of the Congo for cobalt).

A recent report by UNCTAD highlighted the risks this poses to the world battery industry (“Commodities at a Glance: Special Issue on Strategic Battery Raw Materials”).

How Climate Change Will Affect Future Solar Radiation And Wind

Another critical assumption that underlies Joe Biden’s green power vision is that climate change will not have a negative effect on current levels of solar radiation and wind.

There is evidence that this assumption is more uncertain that many people would acknowledge. For example, in the Northeast United States, since the 1950s available solar energy in the summer has decreased, as the warming Great Lakes have produced more clouds (see, “Examining the Climatology of Shortwave Radiation in the Northeast United States”, by Hanrahan et al).

And in “Southward Shift Of The Global Wind Energy Resource Under High Carbon Dioxide Emissions”, Karnauskas et al find that, “the use of wind energy resource is an integral part of many nations’ strategies towards realizing the carbon emissions reduction targets set forth in the Paris Agreement, and global installed wind power cumulative capacity has grown on average by 22% per year since 2006.

“However, assessments of wind energy resource are usually based on today’s climate, rather than taking into account that anthropogenic greenhouse gas emissions continue to modify the global atmospheric circulation.” The authors “apply an industry wind turbine power curve to simulations of high and low future emissions scenarios in an ensemble of ten fully coupled global climate models to investigate large-scale changes in wind power across the globe.”

Their “calculations reveal decreases in wind power across the Northern Hemisphere mid-latitudes and increases across the tropics and Southern Hemisphere, with substantial regional variations. The changes across the northern mid-latitudes are robust responses over time in both emissions scenarios, whereas the Southern Hemisphere changes appear critically sensitive to each individual emissions scenario…Established features of climate change can explain these patterns: polar amplification is implicated in the northern mid-latitude decrease in wind power, and enhanced land–sea thermal gradients account for the tropical and southern subtropical increases.”

Conclusion

Substantial greenhouse gas emissions reductions via widespread deployment of wind and solar power, battery storage, and the Smart Grid is a seductive vision that may very well win Joe Biden votes in this November’s election.

As a practical matter, however, it is based on six critical assumptions that are far more uncertain than its supporters acknowledge.

Even if the probability for each of them having their most optimistic outcome is 90% (a very optimistic estimate), their joint probability is only 53% (90% to the 6th power). If the individual probabilities fall to 80%, the joint probability declines to just 26%.

To answer the question we started with, no, Joe Biden’s goal of removing carbon from US electricity generation by 2035 is not realistic.