Where the Energy Goes
Farming has become much more energy efficient in recent decades, as improved machinery and production practices resulted in less fuel consumed for each unit of output. Last year, the industry consumed about 1,800 trillion British thermal units (Btu), representing 1.8% of total US energy demand. That is significantly less than the peak of 3.1% of total US energy that the farm sector consumed in 1979. (The Btu total combines EIA figures for direct energy costs and Gro calculations of energy consumed for indirect costs, especially fertilizers.)
The use of fuel, largely for planting, harvesting, and transporting crops and livestock, is the main source of farmers’ direct energy consumption. In 2018, farmers spent $12.4 billion on fuel, down from $16.4 billion in 2013, largely because of lower oil prices. Crop farms last year spent $7.8 billion on fuels, and livestock farms spent $4.5 billion. Diesel’s share of total fuel consumption has been growing over time, as farmers have shifted from gasoline engines to machinery with more efficient diesel engines. In 2018, on-farm diesel expenditures were $8.1 billion, 65% of total fuel expenses. Direct agricultural fuel consumption can be calculated by dividing annual fuel expenditure amounts by the price of each fuel product.
This chart shows agriculture’s annual direct and indirect energy use, measured in trillion British thermal units, or Btu. Energy consumed on the farm, such as diesel fuel for planting or electricity for irrigation, is a direct expense. Indirect energy is used off-farm for producing inputs consumed on the farm, such as fertilizers and pesticides. More efficient machinery and production practices have reduced total fuel consumption since a peak in the 1970s.
Irrigation is a big factor in driving up energy consumption for some farmers. Pump engines use electricity, diesel, or propane to extract water from an underground aquifer. According to the USDA’s 2013 Farm and Ranch Irrigation Survey, farmers spent $2.7 billion on water pumping expenses in 2012, $1.8 billion of which was attributed to electricity.
Indirect energy consumption includes the use of fuel and feedstock in the manufacturing of agricultural chemicals such as fertilizers and pesticides. These can be calculated for fertilizers, for example, based on application rates, which are reported by USDA NASS, combined with public information on the energy intensity needed to produce each of the chemicals.
Fertilizer application rates per acre vary by state and crop. The chart on the left shows nitrogen fertilizer application rates for corn planted in various states in the Corn Belt. In 2018, Ohio farmers were the most intense users of nitrogen at 175 pounds per acre. The chart on the right shows average phosphate fertilizer application rates for soybean plantings. Illinois soybean farmers were the top users in 2018 at 75 pounds per acre.
In the US, about half of all chemical nutrient applications are in the form of nitrogen, which is vital to a plant’s ability to develop proteins and enzymes. From 1960 to 2018, annual use of nitrogen fertilizers rose more than fivefold to an estimated 13.2 million tons, underlining its central role in increasing farm productivity.
Ammonia is the main source of nitrogen in fertilizers used in crop production. Ammonia can be directly applied to soil as fertilizer or used as a raw material to produce nitrogen fertilizers, such as urea, ammonia nitrate, and nitrogen solutions. The process involves extracting hydrogen from natural gas (or coal or oil) and nitrogen from the air. The two react with the help of catalysts, high temperature, and high pressure. To produce one ton of ammonia, it takes 33 million Btu of natural gas, roughly one-third the amount to heat an average US home for a year. Still, producing ammonia requires about half as much energy today as it did in the 1960s.
As with direct energy costs, expenditures on indirect energy inputs vary greatly by region and crop. Corn fertilizer expenses averaged $110/acre in 2018, compared to just $24/acre for soybeans. Chemical expenses ranged from $95/acre for rice to $7/acre for oats.
Expenses per acre vary widely by crop. The chart on the left shows fuel expenses, including electricity, are much higher for rice on a per acre basis than for other crops. On the right are per-acre fertilizer expenses, most commonly including nitrogen, phosphorus, and potassium compounds. Fertilizer expenses are much higher for corn and rice than for other crops.
Although pesticides require energy to facilitate chemical reactions, and for storage, packaging, and transport, these chemicals have a smaller energy footprint than do fertilizers.
Gro calculations find that indirect energy use on farms amounted to 580 trillion Btu in 2018, in line with our calculated 10-year average of 578 trillion Btu. Seventy-five percent of indirect energy use is related to nitrogen fertilizer production.
Other indirect energy uses in agriculture are for potash, which requires around 2 million Btu per ton of production, and phosphates, which use 70,000 Btu per ton.
How Gro’s Data Can Measure Energy Use
The Energy Information Administration (EIA) once a year publishes a national number for farm energy consumption for three sectors: crops, livestock, and forestry. Using the EIA’s framework, Gro data can be used to sharpen those energy demand estimates to the state or county level, and with weekly or even daily frequency.
In addition, the USDA’s NASS calculates total farm expenditures by categories, including energy, for each state. And the USDA’s ERS reports farm expenditure data broken down by crop and per acre planted.
Gro’s area planted and yield models can be used to predict crop production by county, which should be directly related to local electricity and diesel demand at harvest. Daily harvest pace can be calculated by combining weather forecast data with historical crop progress reports.
Estimating energy demand for planting can be done in a similar fashion. With historical percentage planted data for each state and weather forecast data, users can estimate daily planting progress and predict when diesel demand will surge.
This map shows annual farm expenditures on petroleum-based fuels, by state. In 2018, farmers in California, Texas, and Iowa, three of the largest producers of agricultural products, incurred the most expenses on fuel, of between $800 million and $1 billion per state.
Building a Propane Demand Model
Crop drying is a fuel-intensive farm activity, and the amount of fuel used varies by the type of crop and its moisture content. High-temperature dryers are usually powered by propane.
Propane demand in the US Midwest exhibits heavy seasonality, mostly centered around the grain harvest season. The surge in propane demand can create local bottlenecks and supply shortages, depending on demand from other uses, such as home and building heating. Using data in Gro, including harvest progress, moisture content at harvest, and production forecasts, we construct a daily propane demand model for the corn drying around this year’s harvest period.
First, we use Gro’s yield and area planted models to calculate a production estimate for each county in the Corn Belt. Next, using NASS harvest progress data, we estimate a daily harvest amount. Crop drying tends to occur within 24 hours after harvest, so harvest volume estimates are lagged by a day to create a daily crop drying estimate. Using NASS data on crop moisture at harvest, combined with propane demand intensity per percentage point of moisture removed, we compute the total amount of propane needed to dry the corn crop at harvest. In the chart below, Iowa’s daily propane demand estimates are provided for the 2019 corn harvest period.
Propane demand rose strongly from Oct. 7, when 15% of the corn crop was harvested, to Nov. 10, when the crop was 66% harvested. Crop moisture content is also higher than normal this year, at 21% compared to a five-year average of 17%. Higher moisture levels require more propane for crop drying. At its peak, propane consumption topped 200,000 gallons per day in the state.
This chart shows the estimated daily propane demand needed to dry Iowa’s corn crop from this year’s harvest. We modeled propane demand for each Iowa county, then summed those amounts to arrive at a statewide total.
Agricultural energy demand varies widely from state to state and by crop or livestock. Modeling energy demand for specific regions can be helpful for a wide range of industry participants to predict local energy prices or forecast crop growing expenses. Gro’s database includes key factors in agricultural energy use, including diesel fuel expenditures, fertilizer application rates, and crop planting and harvesting pace. In addition to the Iowa propane demand model presented above, Gro’s data can be used to expand energy modeling to other regions and energy inputs.