Energy Density and Land Use: It Matters

How much land would we need to use to power all of the US for a year by utilizing just one energy generation source?

This question is much, much more complicated than it sounds, but it’s a curiosity I’ve had for a long time. As a seasoned industry professional, I feel like I have an intrinsic knowledge of how much space each generation source requires. That said, I haven’t sat down and actually attempted to do the research and math – until now.

Energy production and consumption are linked but not directly one to one. The EIA has a fantastic article demonstrating the different source sectors, electric power sector, and end-use sectors. They do a wonderful job summarizing it to a level most people can understand, and this graphic captures the general idea:

Energy generation and energy consumption comes from a myriad of sources and have a slew of end users. As such, much of this analysis will be a hypothetical thought experiment to demonstrate land use required for each electricity generation source.

The first pass at this analysis used a theoretical engineering approach rather than real world data. I was reminded that resources exist (i.e. the EIA and Google Earth) to look at real world examples of how much land we’re using, and “over-engineering” problems often leads to wrong answers with terrible conclusions. Assumptions are listed, and all links to resources are below. For simplicity, I only look at the differences between solar, wind, oil, gas, and nuclear.

I used the EIA’s map to identify and look up power capacities across the country. It’s a great resource if you ever want to understand where power plants and power infrastructure exist in the US: https://www.eia.gov/state/maps.php

I also included a spreadsheet at the end of this post with all of my calculations and data. Feel free to download, review, copy, and critique my work. If you happen to catch an error, comment or email me! You won’t hurt my feelings! 😉

Total Demand

Based on the EIA’s estimate, we need 113 Quadrillion BTUs (aka a shit-ton) of energy to power the US annually. I prefer to think about energy in metric, so I converted all values to kWh. Total US consumption is approximately 3.3117*10^13 kWh, or 33,117,000,000,000 kWh. Assuming load is flat (horrible assumption, but we’ll use it anyway), then that means we need:

33 * 10^12 kWh/year * (1 year / 365 days) * (1 day / 24 hrs) = 3.78 * 10^9 kW.

Summary

Because most people won’t bother to step through the entire analysis, here’s a summary of my estimates.

And here’s a map showing the relative sizes of each of those land areas relative to the US. Rectangles are colored by solar (yellow), wind (orange), oil (green), gas (red), and nuclear (purple…but like, where is it??).

Solar 🌞

I’m aware of a solar power plant in southeast New Mexico. Using EIA’s map, you can see its nameplate capacity is 10 MW:

I’ve driven past this one on my way to our oil and gas assets in New Mexico. Mapping it on Google Earth shows it’s about 0.18 square miles:

Simple calculation: 10 MW nameplate capacity per 0.18 square miles gives it an energy density of ~55.6 MW/mile^2. Assuming these solar plants were generating power at max capacity all the time (which is – of course – asinine because #nighttime), then we’d require ~68,000 square miles to power the US. Put on a map, that area covers most of New Mexico

‍

Wind 💨

There’s a certain kind of ethereal feeling driving through a wind power plant. If you’ve ever flown over Texas or driven through any of the major farms there, then it’s easy to be mezmerized by the scale and tenacity of the farms. The word hypnotic comes to mind…

For the wind land area, I searched for one of the densest power generation areas I could find. Enter Nolan, TX:

I narrowed in on these seven power plants with a cumulative nameplate capacity of 1961.2 MW:

I did my best to trace the total area in Google Earth and came out to roughly 133 square miles:

That means wind power has approximately a power density of 15 MW per square mile. Scaling that up, it would take a substantial portion of Nebraska, Kansas, and Oklahoma to provide all of the energy the US needs for a year. Keep in mind, that’s IF the wind is blowing ALL the time…

Oil 🛢

I selected a six mile by six mile area in the heart of the Permian Basin in New Mexico. I’m biased to this area because it’s very close to where my existing assets are. It also happens to be one of the most prolific basins across the country, so these averages will likely be high.

In 2020, this area produced 25,577,504 barrels (bbl) of oil. Converted to power density, this area has a power density of approximately 138 MW/square mile. Scale that up to what we’d need to power the US for a year, and we’d need about 27,400 square miles. This is shown as the green square in the image below. Solar (yellow) and part of wind (orange) areas are shown for scale:

Gas 🔥

I chose a prolific area of Pennsylvania with a high concentration of natural gas wells for the gas example. These wells have likely been used over the past decade to keep lights, heat, and air conditioning on all across the east coast. In total, this area covers about 374 square miles.

In 2020, all of these gas wells produced 1,425,071,050 thousand cubic feet of gas. Converted to MWh, that’s approximately 428 million MWh (dayyyuuumm) over 376 square miles. Scaling this up, we’d need approximately an equivalent amount of land mass as oil to power the US for a year (~29,000 square miles aka damn near the size of Pennsylvania):

Keep in mind, the surface use for wind turbines, oil wells, and gas wells is relatively small, but we have to space them apart in order to capture all of the energy. While a windmill might only take up a fraction of an acre, we have to space them hundreds of feet apart in order to effectively capture the laminar flow of the wind. Similarly, a shale well pad might only occupy 400′ by 400′, but we have to place them about a quarter to a half mile apart in order to effectively drill the horizontal wells and capture all of the resource in the subsurface. The estimated required area then is smaller and we could place additional assets in between them – but it’s really not worth it when we look at the land use required for nuclear.

Nuclear ☢️

Now we get to my favorite: energy via fission! Just for fun, here’s a map of the current nuclear power plants in the US:

We’ve selected relatively new power plants for all of the other analysis, so let’s look at one of the newest in the US: the Watts Bar Nuclear Power Plant. Unit 1 has a nameplate capacity of 1,167 MW, and Unit 2 has a nameplate capacity of 1,165 MW giving them a combined total of 2,332 MW (EIA has them listed as 2,539.8 MW nameplate, so let’s use that):

Zooming in on Google Earth, we can see the vast amount of land the Watts Bar Nuclear Power Plant occupies… wait for it!!! ….

…1.1 square miles.

Scaled up to power the entire US, and we’d need just about 1,600 square miles

‍

Comments/Conclusions ✅

Keep the following facts in mind:

• Oil and natural gas wells will deplete after we produce all of the resource. It’s not a viable long term plan. The above land uses are only for one year. What happens when the wells go dry in 20 to 40 years?
• Windmills and solar panels don’t have an infinite life. As they’re manufactured currently, they will need to be replaced. Furthermore, the land areas shown above only account for the NAMEPLATE capacity aka they don’t take into consideration the additional storage or charging capacity we would need from renewables if we want to heat our homes at night or charge our electric vehicles when the wind isn’t blowing.
• Nuclear power plants are permitted in the US to last 40 years, but many of them are renewing their permits for the 2nd time to extend their life up to 80 years. There’s no reason to think we won’t extend their permit up to 100 years 20 years from now.
• This post isn’t intended to look at the typical rebuttals against nuclear aka, “But what are we going to do with the nuclear waste?! But what happens if there’s a catastrophic accident?! GAHAAAAA!!!” In short, these are problems we can solve, and I intend to address them in future posts.

If you truly care about the environment, slowing/stopping/reversing climate change, and preserving land for future generations, then stop questioning which energy technology to support. There’s a short answer to the big problem.

Caveats, Complications, & Conundrums

Answering questions inevitably leads to more questions. Here are follow up points, assumptions, and/or questions that came to mind while exploring this idea.

• Capacity factor of renewables wasn’t consider for the total land use i.e. just the nameplate capacity was used. If we actually wanted to keep the lights on all the time, we’d likely need to triple the land mass used.
• There are a tremendous number of efficiencies lost in converting raw materials (i.e. oil and natural gas) into electricity. These weren’t accounted for.
• There’s also a ton of land used for pipelines, refineries, manufacturing facilities, filling stations, etc. that wasn’t accounted for with oil and gas.
• Land use for mining uranium wasn’t considered for nuclear, but it’s assumed to be trivial. Prove me wrong.
• How much land would we need to use if we included:
  • Battery storage?
  • Energy storage using dams?
  • Hydro power?
  • Geothermal power?
  • Concentrated Solar Power (Solar was done for commercial scale PV)?
  • Nuclear waste for 1000 years?
• How does the addition of waste impact the land use?
  • How much space does nuclear waste consume?
  • How much space will PV panels and windmills consume?
  • How much space does oil and gas development leave polluted or contaminated?

References 📝

• EIA Energy Usage: https://www.eia.gov/energyexplained/us-energy-facts/
• https://www.eia.gov/state/maps.php
• Enverus Prism Tool

‍