We must dramatically step up our efforts to achieve the US energy transition to 100% carbon free and 100% electric. It will be totally worth it!
Part 1 estimated that by 2050 we need to expand renewable energy around 27 times the amount the US produced in 2019. This implies an annual increase equal to as much as 7 times the increase in 2019.
The foremost technical obstacle to such an energy system is seasonal imbalances between renewable supply and electricity demand, which can be solved through massive overbuilding of wind and solar resources.
The energy system of 2050 will require that we pick up the pace of our effort through some combination of (1) Greening the grid, (2) Electrifying everything, (3) Investing in energy efficiency, (4) Aggressively pricing carbon, and (5) Making public sector investments in research and other areas in which carbon pricing is insufficient to achieve the result we need.
The electric system of 2050 will be around 2.5 times the size of the electric system today due to electrification. But electricity is so much more efficient at producing the final “energy services” (heat, motive power, light, etc.) that total energy use will be far lower than it is today.
Unfortunately, electricity is harder to store for long periods than fossil fuels. Fossil fuels are nothing but stored solar energy, so they lend themselves naturally to the idea of “being saved till you need it.” Advances in battery technology have virtually solved the problem of storing electricity for short duration at reasonable cost, but have not solved the problem of “seasonal storage” where energy must be stored for much longer periods of time and the storage might be needed as little as once a year or even less. This is likely to be a winter season problem in the US.
Part 2 will consider in more depth (1) The nature of overbuilding to meet seasonal imbalances, (2) Alternatives to overbuilding to meet all seasonal imbalances and (3) The relative cost of this “dramatic step up.”
Overbuilding Or A Combination Of Options?
The seasonal problem can be solved through overbuilding renewable energy resources. Overbuilding will likely carry a price tag far lower than a hypothetical battery resource even with much lower battery costs versus today. Still, the overbuilding required leads to an oversupply of wind and solar equal to 25% of total annual production, making it an expensive part of the solution. Might there be some less expensive alternatives?
I looked at different levels of overbuilding using the spreadsheet model developed for Part 1. The more we overbuild, the more the benefits of additional overbuilding decline.
Consider an additional 3 scenarios beyond the 90% and 100% carbon free scenarios for 2050: overbuilding to 92.5%, 95%, and 97.5% carbon free. The remainder of electricity demand would be satisfied by the lowest cost option to get there, which I conclude is less likely to be more overbuilding above around 95%.
Each bar in the graph below represents the incremental curtailments of wind needed to avoid the same incremental quantity of remaining fossil fuel energy between 90% and 100% carbon free (each increment represents 2.5% of total energy).
The fossil fuel generation for each segment is 248 terawatt-hours (TWh), and for the 90–92.5% segment we get 180 TWh of curtailment to save the same 248 TWh. For the 97.5% to 100% segment we need 1,308 TWh of curtailment, or about 8 times as much. This is what economists call an upward sloping incremental cost curve. As we get closer to the magical 100% it becomes more and more difficult.
We can also examine the cost of replacing each increment of remaining fossil fuel development including the cost of curtailed energy. To develop cost estimates, I assumed that the incremental generation came from wind energy, which is better suited to meet winter demand than solar, at a cost of $30 per megawatt-hour (MWh), which is a 25% reduction from Lazard’s most recent cost estimate, and thus builds in sustained improvements in lowering wind cost at a steady annual rate.
The costs, including curtailment cost, of this final step to get to 100% clean energy range from $51.70 per MWH for the 90–92.5% increment to $187.50 per MWH for 97.5–100%. Again, this shows an increasing marginal cost curve.
That $187.50 for the final 2.5% of total energy use could seem like a lot when you are comparing it to assumed $30.00 for wind energy in 2050. If you think so, you should look at Lazard’s estimated cost of energy report (2020 version just released) or at my article in CleanTechnica comparing 2019 levelized cost numbers. In 2020, the levelized cost of peaking power from combustion turbines is estimated by Lazard with a midpoint of $174 with no carbon pricing. This is only slightly below the final increment cost. We can just rely on overbuilding for that final 2.5% and pay barely more for it than we would pay today for peaking power!
It is interesting to consider the carbon emissions avoided for each increment of overbuilding, shown below:
Overbuilding for the first increment would only cost $52 per ton to avoid additional carbon emissions, while the final increment would cost $378 per ton. Compare the above to the CO2 fee projection from the Energy Innovation Act (my preferred carbon pricing solution). The fee by 2050 for this act would range from $300 a ton to $450 a ton, so the 95–97.5% rate is well below the minimum and the 97.5–100% rate is midway between the minimum and maximum.
But just because a carbon fee is set at a certain level does NOT mean that all opportunities to reduce carbon below that level are accepted. The carbon fee creates an incentive, but rational utilities would not choose a solution that reduced carbon at a cost equal to the fee (say $136 or $378 a ton) if there were a solution that could reduce the same carbon at a lower cost. The carbon fee becomes the UPPER BOUND of prices that utilities would accept.
My intuition tells me that above 95% (or above $83 a ton or above a cost of $87 per MWh) utilities would want to look for other options that could satisfy energy demands and reduce carbon emissions more economically. Thus, there would be a blend of overbuilding with other options. What can we say about those other options?
Options For The 95% To 100% Solution:
Here are some possibilities that come to mind. Please suggest some more in the comments!
Take advantage of a warming climate: How much warmer will it be by 2050? It’s possible that electric demand will dramatically shift with warmer winters and hotter summers (counterbalanced by climate migration north). Maybe forecasted winter peak months will disappear at some point. How much of that will happen by 2050?
Maximize hydro use for seasonal storage: For 20 years, our family owned a home on a TVA power lake in North Georgia. Most days, the Lake Chatuge hydro plant was operated like a battery, with the lake level rising a little at night and falling a bit during the day, to generate power when demand was highest. But a major purpose of the lake was to provide recreation and help tourism, so daily use never drew down the lake much. After staying at a high level all summer and fall, the lake was drawn down in late fall to the typical winter pool level. This 10-foot drop in lake level represented a huge amount of power available one time per year. This seasonal storage was inexpensive since the dam was already built. Every hydro resource could be evaluated to see if such opportunities are available.
More home and business weatherization: Tightening building envelopes, adding insulation, and increasing HVAC efficiency focuses energy savings on the winter months. Carbon pricing will provide greater incentives which could be combined with advertising, financing and subsidies, especially for low income, to transform our buildings so they consume much less energy in winter. Some of this is already factored into my forecast for energy demand. To go beyond the baseline we will need a massive program of national and community efforts that pays off in many ways.
Green hydrogen: Renewable energy can produce renewable hydrogen, which can in turn be stored to produce electricity in winter. Hydrogen is used today in the chemical industry and it could be used as a substitute for the fossil fuel component of our scenarios. Producing “green hydrogen” could utilize the otherwise wasted renewable energy and it will find a role in aviation and in some industrial processes. However, renewable energy with batteries is likely to be far less costly for most applications. Using renewable electricity to produce green hydrogen to in turn produce renewable electricity at a different time requires capital investment and results in efficiency lost. I think green hydrogen will play a role in solving the seasonal storage problem along with its other uses, although the scale of this contribution remains to be seen. Michael Liebreich wrote two excellent articles in Bloomberg New Energy Finance on this subject for those interested in a deep dive, here and here.
Biofuels: Biofuels may play a role in various sectors, including electricity. Methane from trash or waste, ethanol from corn, or plain old firewood can be used to produce electricity or to directly produce heat. Just like hydrogen, biofuels can be stored inexpensively for long periods of time and then help solve the seasonal storage problem. We already produce a lot of ethanol to blend with gasoline, and we will not need it for that purpose once ground transportation is electric. Some of this ethanol could be used for heating, air transport, or for generating electricity.
Demand response: Utilities can pay large users to shut down during winter months. They can also design seasonal rate structures which reward using less electricity in high demand / low wind and solar months. If electric rates are redesigned to be highest in the winter months, then every electric customer can profit from being part of the solution. Once we have a surplus of curtailed renewable electricity, there will also be an incentive for businesses to orient their operations around directly tapping into the low cost of this resource.
Carbon capture: One option is to continue to burn fossil fuels and directly capture the carbon OR to just capture carbon from the air to offset fossil fuel burning elsewhere. A net negative form of energy production would be to burn biofuels and capture the carbon. These options are uneconomic today, but they could become so with high carbon prices and technology improvements.
Nuclear or geothermal: Carbon free “baseload” technologies like nuclear or geothermal might achieve a cost breakthrough that reduces the overbuilding needed for wind and provides a component of power without seasonal variations. Again, the economics currently do not favor these approaches versus solar and wind with batteries. This is not a targeted solution to the need for seasonal storage, but it would lessen the need as a “side benefit.”
Summary: There are many excellent options and the default of relying solely on wind and solar overbuilding is not all that bad anyway. And we don’t need to decide now how to handle seasonal storage. We just need to focus on producing gobs of wind and solar, improving energy efficiency, and beginning the long process of electrifying everything. Solving the seasonal storage problem does not need to be worked out until after a lot of that is done — say 2035–2050. We can continue to research these options and roll out pilot programs for the next 15 years. By then, carbon pricing can be high enough to guide the market to choosing the best mix of options. Carbon pricing will ensure that the climate benefits of each option are factored into the cost all along the supply chains of all the industries involved.
Will It Be Too Expensive?
We are talking about a big increase in wind and solar: Assuming a capacity factor of 22% for solar and 40% for wind, the 2050 energy system would need capacity of about 2.2 TW of solar and 2.1 TW of wind with the 100% case. If we go with the 95% overbuilding solution, the energy system would need 1.5 TW of wind, for a reduction of .6 TW (600 GW).
At the end of 2019, the US had about 55 gigawatts (GW) of solar and 104 GW of wind, according to data from EIA, for about 159 GW total. We need to expand that total 27 times by 2050 for the 100% case and 23 times for the 95% case. We need to add about 140 GW and 114 GW per year respectively, 7 or 5.75 times as much as we added in 2019.
In my judgement, such an order of magnitude increase will not happen on its own, especially with expiring subsidies for wind and solar. It’s likely to happen eventually on its own, but not at the speed needed for our 2050 net zero goal.
We need government policy to support the rapid advance required!
Fortunately, greatly expanding wind and solar won’t be an economic burden! Assuming $1000 per kilowatt cost of solar and $1,300 for wind (near current estimates from Lazard) that comes out to $4.5 trillion for the 100% case ($148 billion per year) and $3.7 trillion ($120 billion per year) for the 95% overbuilding case.
These costs could be lowered by finding a use for the curtailed renewable energy. The 100% overbuilding case has no fossil fuel generation and 2,465 TWh of wind and solar generation that isn’t needed but that we have to pay for. In the 95% case, 497 TWh of fossil fuel generation remains and must be eliminated by one of the other options and we have 686 TWh of unused renewable energy available to help. Hydrogen electrolysis might work for that. Carbon capture might work for that.
Is $120 or $148 billion all that much to spend per year? Step back and compare this cost to what US consumers spend on other things. Here are some comparisons for 2019 for what US consumers spent on other things: tobacco — $100B, restaurant meals — $680B, alcohol — $250B, or health care — $3,700B. And don’t forget defense spending — $720B in the 2020 budget. It doesn’t seem such a large amount compared to our current consumer spending.
And the investments are roughly in line with what we are already investing in energy. According to Bureau of Economic Analysis data, in 2019 the US economy invested $106B in the power industry, and $116B in oil and gas exploration. Wow. If we could just move the oil and gas budget to the power sector we would almost be there. We could continue to extract oil and gas from existing wells but stop building new pipelines and new exploration.
There’s room in our collective budgets for the transition. We can move to a whole new source of energy and stop killing ourselves and possibly pay less for the whole thing! And once we are transitioned to 100%, the nation’s fleet of solar and wind farms will require much smaller investments going forward than the fossil fuel industry would have required. A livable world will truly be a better future!
So yes, it’s a lot of money, but it’s not a big lift compared to other spending, and it’s peanuts compared to the climate damages of not doing it!
We need to increase our investments in renewable energy sixfold. We need find a way to step up our game and make the investments we need. The results will be profound. The air will be incredibly clean. Illness and death due to fossil fuel air pollution will cease. The US will no longer depend on foreign countries to meet our energy needs. The US will be a leader in combating climate change. We will no longer need to scar the earth for fossil fuel drilling and mining. Electrification will provide a better experience for the many ways we use energy, and once our initial transition investments are made, we will have an era of cheap energy.
By including carbon pricing as the centerpiece of our policy, the investments we make will mostly be paid by the private sector. No need to tap the federal budget to a great extent. And if we adopt the Energy Innovation Act, US residents will be getting a nice monthly dividend to add to all of the great salaries people will be making in the new jobs we create.