Farming, in industrial countries, hasgenerally become an industry itself. As such, it has national energydemands comparable with other industries and thus competes for energysupplies, particularly petroleum. Annual global oil production hasbeen rising for many years, but an important economic transitionperiod for all industries will arrive when annual production peaks,then declines as discovery drops off and the remaining oil becomesharder to extract. That time is not far off, even by optimisticestimates. For example, a recent assessment by the US GeologicalSurvey increased its previous global estimate of remaining oil, bothdiscovered and undiscovered, by an unprecedented 37percent.1 During the century ending in 1990, the worldconsumed an amount of oil equal to this increase, 625 billion barrelsof oil. In stark contrast to the 100 years, using current annualproduction as a guide, 27 billion barrels, the increase will lastlittle more than 25 years during the early part of the twenty-firstcentury. Since oil production will be approximately symmetric on bothsides of the peak, the increase will delay the peak of annual globaloil production by half this period, or 12 years.2 Addedonto current estimates of the peak year, the 12-year delay wouldplace the peak only one human generation from now, or about threedecades. The highly touted Caspian Sea oilfields in the former SovietUnion would postpone the peak only a year.
The importance of national foodsecurity dictates that we should reduce farming's dependence onfossil fuels. As part of The Land Institute's mission to use natureas measure for developing sustainable agriculture and culture, theSunshine Farm Project has been exploring the possibilities of farmingwithout fossil fuels, fertilizers or pesticides. We have beenconducting energy accounting on the Sunshine Farm to assess theextent to which modern farms can run, essentially, on sunlight. Aslate as the 1950s, many farms ran mostly on sunlight, relying ondraft horses and using crop rotations for soil fertility instead ofcommercial fertilizers. Nowadays however, agriculture uses far moreenergy nowadays in the form of farm inputs, while at the same time,there are current renewable energy technologies that were notavailable in traditional farming. Energy accounting reveals patternsof energy consumption that can be examined to determine which changesin farming practices would save energy. For example, the use ofanimal manure as a source of nitrogen requires less energy than thatrequired to manufacture commercial fertilizer.
The accounting on the Sunshine Farmis done by determining the weight of everything going into and out ofthe farm and by entering this information into an accountingframework that we established for our computer database. The weightsof inputs are converted into values of energy by computercalculations based on embodied energy factors (the energy used inindustry to mine, process and fabricate raw materials and ores intofarm inputs) from the professional literature. Likewise, theliterature is used to provide caloric energy factors for convertingour marketed farm outputs into calories, like in our dailydiet.
Our energy accounting shows that theSunshine Farm could supply about 40 percent of its embodied energyneeds through animal feed, electricity, and biodiesel fuel for thetractor and vehicles. That is, a 4.5-kilowatt photovoltaic array onthe Sunshine Farm has been converting sunlight into electricityrequired for the workshop tools, electric fencing, water pumping, andchick brooding. About one-fourth of the cropland on the Sunshine Farmhas been devoted to soybeans and sunflowers for biodiesel fuel thatcould be commercially produced to meet all our field operations andoff-farm transportation. Both the array and biodiesel are renewableenergy; they produce more energy than they consume.3Almost three-fourths of the animal feed consumed on the Sunshine Farmhas been supplied by the oats, grain sorghum and alfalfa produced onthe farm for the draft horses, beef cattle and poultry. The remaining60 percent of the farm's embodied energy needs is imported onto thefarm through purchased inputs including amortized capital, such assupplies, commercial feed and seed, buildings, fences, water lines,tractors, vehicles, equipment, and the photovoltaic array.
These results do not include thelifestyle energy support for a farm family, a large unknown rangingfrom the austere Amish to typically affluent Americans, although wedo keep track of human labor on the Sunshine Farm. Nor do theyinclude some prorated part of the food processing, marketing,distribution and preparation sectors that nationally consume moreenergy than farming.4 Our focus is to determine how muchenergy farms can supply for their inputs. Family support andcommercial sectors are not unique to agriculture but are common toall industrial activity; hence, they are social considerations beyondthe scope of our project.
The purpose of the renewable energytechnologies in our project is to reduce our dependence on fossilfuels but not our dependence on local energy systems. Virtually allfarms are part of the local community in many ways, energy being noexception. For example, from our exploration of biodiesel production,we have learned that biodiesel fuel with quality satisfactory toengine manufacturers can be produced by farmers' co-operatives butnot by individual farms producing various, unregulated home-brews ofbiodiesel fuel. Also, although our photovoltaic array has a bank ofbatteries and could stand alone, it is connected to the electric gridof the local power company to sell excess electricity.
Just as important as the monetaryincome from the sale is the fact that the excess electricity is nowon the grid for use by the local community. In other words, in anall-solar future, given certain limits of energy production by solartechnologies compared to current conventional technologies (seebelow), it will likely be considered uncivic to own a personalelectrical technology located near a grid but not connected to it.This would prevent the use of the technology's potential excesselectricity by other people, and in effect, would be squandering someof the hard-won, solar-based energy embodied in the manufacture ofthat technology. Since any solar technology exposed to the weatherwill slowly deteriorate whether or not it is used, there would belittle to be gained from the selfish idea of operating a personalsolar technology only when the owner needs its energy. The obligationto sell excess electricity would be quite contrary to the currentpopular notion of achieving energy self-sufficiency in order todisconnect oneself from the grid. This notion is made possible by thepresent industrial economy with its abundance of fossil fuels andmineral resources.
Some important consequences forfarming result from some differences between fuel crops and solartechnologies such as heat collectors, photovoltaics, wind turbinesand hydroelectric. The annual energy efficiency of crop production byphotosynthesis is less than 1 percent, far below the range of 10-25percent for the conversion of sunlight into useful energy by solartechnologies. Hence, to produce a given amount of useful energy, fuelcrops require ten to 100 times more land area than solartechnologies.5 Solar's smaller land requirement is why theSunshine Farm uses a photovoltaic array as its primary source ofelectricity instead of a generator operated on a renewable fuel. Fuelcrops' larger land requirement explains why the national productionof ethanol from corn grain for use in gasohol raised a furiousethical debate over diverting substantial cropland from food to fuelproduction. Moreover, solar's smaller land requirement would make itmore desirable than fuel crops for powering tractors on farms ofabout 100 acres or less in an all-solar future, but the technologywill require further development.6
In the energy accounting of a farm,solar technologies are treated in the same manner as conventionalenergy technologies. As an imaginary example, if a farmer built acoal-fired electric plant on his/her farm, it would be ludicrous tocount its marketed electricity as an output of farming. Instead, wenote the amount of electricity consumed annually by the farm (absentthe plant) and calculate the embodied energy of the coal required toprovide that electricity through the farmer's plant. The coal'sembodied energy is the farm input representing the electricityconsumed by the farm. This is the same conventional procedure usedfor electrical consumption by any farm; i.e., the coal-fired plant isusually owned by a utility company and is located far away from thefarm. Analogous to the imaginary example, we do not regard the outputof a farmer-owned solar technology as a farm output. Instead, we notewhat percentage of the technology's output is consumed by the farm.We then take the energy embodied in the mining and manufacture ofthat technology and adjust it by the output percentage to obtain areduced value that becomes the farm input representing the electricalconsumption of the farm.
Agriculture's potential to provideenergy as well as food for society can be ascertained by the energybalance of various farms and national agricultural systems. Wedetermine the energy balance of a farm by comparing the caloricenergy of its marketed outputs with the embodied energy of itspurchased inputs. Not counting the photovoltaic array, the SunshineFarm marketed 1.7 calories for every calorie purchased. The outputsand inputs are qualitatively different types of energy, but sincetheir units are the same, this is an energy ratio of 1.7 to 1 foroutputs to inputs, or simply 1.7. Inclusion of the photovoltaic arrayby the above accounting procedure had little effect on the SunshineFarm's energy ratio, but this was because the energy ratios for thearray and the farm were about the same to beginwith.3
The Sunshine Farm's energy ratio iscomparable to values at the upper end of the range for mixed crop andlivestock farms (Table1).7 Aconventional Pennsylvania dairy farm had a ratio of 1.8. Amish farmsfrom six communities had ratios of 0.7-1.6, greater than the ratiosof 0.3-0.6 for nearby conventional farms, except for a group of mixedfarms in the Illinois Corn-Belt region that averaged 2.0. This strongcontrast between Amish and conventional farms is all the moreremarkable considering that the former also energetically providetheir own field traction while the latter do not. That is, Amishfarmers raise draft horses fed by their crops, while conventionalfarms do not make their tractors and fuel, but purchasethem.
The Pennsylvania dairy farm, IllinoisCorn-Belt farms, and Sunshine Farm had the greatest farm energyratios, partly because most of their marketed output (69-95 percentof caloric energy) was crops (Table1). In contrast, crops on theother farms constituted only 14-29 percent of marketed output, andanimal products accounted for the remaining share. On the otherfarms, feeding much of the crops to animals incurred large metabolicenergy losses and thus lower farm energy ratios. Another reason forthe large farm energy ratios of the Pennsylvania dairy farm, SunshineFarm, and some communities of Amish farms was the low amount ofpurchased inputs per acre of cropland (Table1). In other words, lessenergy-intensive production helps yield a higher farm energy ratio.The average per-acre marketed crop output of the Illinois Corn-Beltfarms was so great that they had the largest farm energy ratiodespite a large amount of purchased inputs per acre (Table1). These results arecorroborated by 15 hypothetical farm energy budgets computed byGerald Leach in his 1976 book, Energy and Food Production: thegreater farm energy ratios were clearly associated with fewerpurchased inputs and with a larger proportion of outputs devoted tocrops.
National energy ratios for farming invarious countries display a range of values similar to mixed farms(Table2). Industrial countries,such as the US and those in Europe, have energy ratios near 1.0 orless, like conventional mixed farms. This is not surprising becausethe agricultural structure of industrial countries is generally a mixof energy-intensive crop and livestock production. For example, aboutthree-fourths of the crops in the US are fed tolivestock.8 While much of it is fed in large feedlots andoperations, this value is a common proportion on individual mixedfarms, thus resulting in similar energy ratios at the national andindividual levels of farming. As an example of the effect of adoptingcapitalism, China's agriculture has rapidly increased its use ofcommercial fertilizers and other inputs, resulting in a nationalenergy ratio of 1.2 in 1978. This value is surely lower now andreflects an industrial economy instead of one that relied on peasantfarming methods not that long ago, as described in Farmers ofForty Centuries.9
Although the energy ratio of 1.0 forUS agriculture is based on data a quarter century old(Table2), this data is nearly thelatest available from which an energy ratio can be readilycalculated. It was computed from summary data reported by Stanhillthat, in turn, was based on the extensive 1974 USDA national surveyof agricultural energy inputs.10,11 The USDA surveyrepeated in 1978 also gives the same energy ratio.12 Sincethen, less detailed estimates of some agricultural energy inputs havebeen made, but with no computation of the embodied energy in theinputs.13,14 Recent econometric studies have shown thatenergy productivity in US agriculture has increased since 1980 due toimprovement in technology and farm management stimulated by theenergy price shocks of the 1970s, the withdrawal of energy-requiringmarginal cropland through the federal Conservation Reserve Program,and the economy of scale accrued from continuing increases in averagefarm size (although the latter is not desirable for ruralcommunities).15,16 Hence the current energy ratio for USagriculture must be greater than 1.0, but it cannot be directlycalculated from these econometric studies because the measures ofenergy productivity were based in part on financial data orindices.
National energy ratios for farmingare higher in less-industrialized countries. For example, agriculturein Egypt, Pakistan and Australia have respective ratios of 1.8, 2.9,and 3.1, again because of fewer purchased inputs and fewer crops fedto animals (Table2). Australia relies mostlyon low-input crops (e.g., wheat), free-range animal rearing, andextensive use of leys (i.e., grazed legume cover crops) instead ofcommercial fertilizer for cropland nitrogen needs. Likewise inAmerica, before the tremendous increase in commercial fertilizers,pesticides and irrigation after World War II, American farming in1940 had a national energy ratio of 2.3 (Table2).
To understand the challenge ofobtaining energy from agriculture, the energy ratios for mixed cropand livestock agriculture can be compared with energy ratios thatindustrial society has obtained from nonrenewable and renewableenergy sources. The highest energy returns have come fromnonrenewable fossil fuels. Drilling for petroleum and natural gascurrently yields an energy ratio of 10, well below the ratio of 100or more in the 1940s when oil was literally gushing out of the wells(Table 3). The energy ratio for coalalso declined during that time with a current value of 30 at the minemouth or stripped coalfield. Including combustion of coal in electricpower plants drops this value to a national average of 9. The ratiocan go as low as 2.5 for western, strip-mined coal that istransported more than 1,000 miles to Midwestern power plantsoutfitted with pollution scrubbers. Nearly the same energy ratioapplies to the current extraction and use of natural gas in electricpower plants.
The mining and processing of uraniumore and its use in nuclear light-water reactors to produceelectricity result in an energy ratio of only 4, which includesstorage of spent fuel (Table3 ). This value certainlywill not justify the future exposure of society to the inherent risksof nuclear reactors and their fuel cycle from mining to burial. Therisk is simply too great for large-scale civilian or militaryaccidents, sabotage, or mismanagement with huge economic,environmental, and social costs. None of the energy ratios inTable 3include decommissioning of atechnology (i.e., moth-balling, dismantling, or disposal) at the endof its useful life, and this energy input, relative to the otherinputs, will be far larger for nuclear reactors than for the othertechnologies.
The energy ratios for renewableliquid fuels are much lower than for extraction of petroleum.Conventional agricultural production of grain, starch, or sugar cropsand some chemical processing yield ethanol fuel with small energyratios in the neighborhood of 1-2, with methanol from tree productionslightly higher (Table3 ). The energy ratios forbiodiesel fuel, which can be chemically processed from variousvegetable oils, would not be much higher than ethanol and methanol.The ratios for ethanol and vegetable oil also include an energycredit for the by-product spent mash and meal cake,respectively.3
The energy ratios for renewable solidfuels, namely 6-13 for biomass production by intensive farming ofcrops and trees, are lower than for mining of coal, the analogousfossil fuel (Table 3). The energy ratios forbiomass production are higher than renewable liquid fuels because thelatter require chemical processing that introduces either substantialenergy inputs or chemical losses for the liquid fuels and thus lowerstheir energy ratios. Moreover, the output of fuels derived from seed,like ethanol and biodiesel, is lower than biomass because seed isonly a small portion of the above-ground plant that constitutesbiomass.
Subsequent use of biomass furtherreduces the energy ratios resulting from biomass production.Gasification of biomass crops to produce a gaseous fuel results invalues of 2-5, lower than the current ratio for extraction of naturalgas, the analogous fossil fuel (Table3 ). These values are alsonot much better than the above ratios for renewable liquid fuels.Direct combustion of biomass crops for heat might give slightlyhigher ratios, including an energy input for the embodied energy inthe furnace or boiler. Heat can also be obtained by flat-plate solarcollectors, including storage, with energy ratios of 2-5, the lowervalues including fuel-operated heating systems as back-up. Directcombustion of crop residues or biomass for electricity results inenergy ratios between 3 and 4, although advanced cogeneration ofelectricity and heat, not yet commercialized for biomass, may yieldvalues twice as high.
Another renewable fuel is biogas,mostly methane, produced from anaerobic digestion of agriculturalmaterials such as manure, crop residues, or wastes. The range ofenergy ratios in a table of results from eight European countries was1.7-5.6, the higher values in southern countries with warmer climatesthat reduce the energy input required to keep the digester warmenough for sufficient microbial digestion.17 However, ifwe use some other studies to add estimates for collection ofagricultural materials and for amortized embodied energy indigesters, then the range becomes 1.5-3.1, less than the ratio forextraction of natural gas, but comparable to renewable liquid fuels(Table 3).18,19 The lowervalues, associated with northern countries, do not decrease muchbecause the two additional inputs are small compared to the heatingrequirement for the digesters. Finally, if an agricultural materialis going to be regularly dedicated to biogas production, or someother renewable fuel, then it should be regarded not as a secondaryby-product with just collection costs but as a primary product withits prorated share of the production inputs for its source. Usingprorated production inputs for crop residues as representative ofagricultural materials in general, the range of energy ratios for theEuropean countries would then become 1.4-2.0.
Analogous to the results forrenewable fuels, most solar-related technologies for the productionof electricity, including reliable back-ups, generally have energyratios less than the national average for coal-fired electricityincluding coal mining. One exception is hydroelectric systems withdammed water as inherent energy storage, which commonly attain anenergy ratio of about 10. Upper values of 8-10 have been reported inenergy ratios for photovoltaic arrays that electrochemically convertsunlight directly into electricity and for parabolic-thermalreflectors that collect heat for driving steam or gas turbines togenerate electricity (Table3 ). In comparison,wind-electric turbines have achieved upper values twice as high.However, it remains to be seen if these upper values can be typicallyachieved so as to become averages. Unlike current centralized powerplants, these smaller technologies will be distributed in locationand thus will have less embodied energy in transmission lines.However, despite being connected to local electric grids, sometechnologies may require energy storage and/or back-upsystems.
Energy conservation and efficiencyreduce the consumption of energy with the result that energy savingseffectively constitute an output available for other uses. The energyratios for conservation are comparable to nonrenewable energy sourcesand are often an order of magnitude greater than renewable fuels andsolar technologies. For example, double-pane windows and ceilinginsulation prevent losses of heat in winter and coolness in summerthat are equal to 136 and 61 times, respectively, the energy expendedin producing and installing the windows and insulation(Table 3). Passive solar designreduces heating and cooling in new houses by an amount equal to 10-25times the energy spent in manufacturing the passive components fromraw materials and in constructing new houses with them. In otherwords, much more energy would generally be made available from agiven amount of energy inputs invested in conservation and efficiencythan in renewable energy sources. As documented by the US Departmentof Energy, during 1979-1986 the US obtained 7 times as much newenergy from savings through conservation and efficiency than from allnet increases in domestic energy supplies based on fossil fuels,nuclear power, and renewable sources.20,21 Henceconservation and efficiency should be fully developed, as well asrenewable energy sources.
This brief review shows thatindustrial society has been powered by nonrenewable energy sourceswith energy ratios much greater than the values of 2.0 or less formixed crop and livestock farms. With energy ratios this low,agriculture will not be a net source of renewable fuels orelectricity. Simply, if outputs from mixed farms were converted intofuels or electricity, typically half of the energy in the outputswould be lost during the conversion processes.22 Inconjunction with the energy ratios of 2.0 or less, this implies thatfor any mixed farm the potential output of marketed fuel orelectricity would be less than the embodied energy required in themanufacture of the farm's purchased inputs. Hence, beyond the foodand feed already marketed by mixed farms, there would be no netoutput available as energy for society. The same conclusion alsoapplies to US agriculture for which the energy ratio is greater than1 but probably less than 2, as elaborated above. The nation exportsone-fourth of its grain production, and one might think that it couldbe converted into a lot of energy for society, but the resultingconverted energy would provide less than one-half of the embodiedenergy in the farm inputs used by USagriculture.4,8
For mixed farms or US agriculture tohave some net output available as energy for society, the energyratios must be raised by reducing purchased inputs and increasingmarketed outputs. Many farmers have been using less purchasedfertilizers and pesticides, mainly to cut expenses. Farms couldsomeday, like the Sunshine Farm, energetically supply their own fuelsand electricity instead of purchasing them. Inputs can also bereduced by utilizing biological efficiencies in crops and animals,such as letting animals obtain their own feed through grazing andforaging which involve no embodied energy, in contrast to feedingthem machine-harvested grain and hay.
As far as increasing the amount ofmarketed outputs from mixed farms, increased crop yields will not bean option under a regime of fewer commercial inputs in farming. It isenvisioned that yields will be maintained not quite as high ascurrent levels by diverse farming practices that will require moreuse of land, biological efficiencies, and human labor. Also,substitution of fuel crops in place of feed and food crops will havelittle effect since they have fairly similar yields under equivalentfarming practices.
Large increases in marketed outputscould be achieved by diverting cropland from supplemental animal feedto crops for direct human consumption. The potential increase islarge because slightly more than half of US crop production is fed toanimals.8 Since the feed conversion efficiency of animalsis only 10-20 percent on a weight or energy basis, each pound less ofanimal products derived from supplemental feed would permit an outputof 5-10 pounds of directly consumed crops substituted for thedisplaced feed crops.8 There is plenty of slack forreducing the consumption of animal products in our diet because theaverage American currently consumes twice the average minimum dailyprotein recommended by the international Food and AgricultureOrganization, and two-thirds of our daily protein intake is animalprotein.8,23
By these strategies for inputs andoutputs, mixed farms and US agriculture should be able to increasetheir energy ratios to 3, perhaps 4, the former figure alreadyachieved by Australia (Table2). An energy ratio of 4implies 4 units of marketed farm outputs in the numerator of theratio for every 1 unit of embodied energy in the manufacture ofpurchased farm inputs in the denominator. Provision of 1 unit ofembodied energy would require conversion of 2 units of farm outputsinto fuel and electricity because of the aforementioned conversionlosses of about 50 percent. Subtraction of the 2 converted units fromthe 4 units of marketed outputs would leave 2 units of farm outputs,or effectively an energy ratio of 2 after agriculture has met theembodied energy requirements of its own farm inputs. Considering thatthis value is double the above energy ratio of 1.0 for USagriculture, in absolute amounts this ratio should provide sufficientfood and maybe a small amount of energy for the nation's society.Hence, national agriculture must achieve at least an energy ratio ofroughly 4 to be a net source of commercial energy beyond the fooddemands of society and the energy requirements of its own farminputs. Otherwise, an energy ratio less than 4 means that manufactureof some agricultural inputs will require energy subsidy from societysuch as electricity from photovoltaic arrays or wind-electricturbines.
The above review also demonstratesthat nonrenewable energy sources, namely fossil fuels and nuclearpower, have energy ratios that are generally greater than theanalogous renewable energy sources. Furthermore, they former sourcesyield a lot of power from the little land area required for mining,fuel processing, and power production. In other words, fossil fuelsand nuclear power yield much greater power densities, or power peracre, than renewable energy sources because the latter are derivedfrom sunlight that is dispersed across the landscape.5Hence, fuel crops and solar technologies would require much more landto meet the current energy consumption of various USsectors.
An aggressive national program ofenergy conservation and efficiency will be required to sufficientlyreduce energy consumption such that the US economy could be poweredby renewable energy sources without using too much land. Absent suchmeasures, for example, if the nation's current transportation sectorwere to be fueled solely by the gross yield of ethanol from corngrain, then half of the entire US must be planted tocorn.24.25.26 This result assumes an average corn yield 90percent of the current Midwest corn yield and that the energy inputsfor corn production and ethanol processing in an all-solar futurewould be met by solar technologies. Since solar technologies havegreater power densities than fuel crops, they require less land thanfuel crops.5
We cannot look to America's forestsfor much woody biomass unless we increase the recycling of paper andwood products. The current annual net growth of America's forests isalready entirely accounted for by the national consumption of woodand paper products.8 Extra biomass could be gainedinitially by improvement of forest productivity through conventionalintensive practices such as fertilization and introduction ofnonnative species, but this would result in a loss of biodiversityand a concomitant, long-term decrease in biological efficiencies thatwould offset the initial extraproductivity.27,28,29
Nonetheless, some energy scholarsbelieve that energy conservation and efficiency will make it quitepossible to power our current standard of living with renewableenergy sources.30,31 Solar technologies would beparticularly important in meeting US energy needs since they havemuch greater energy ratios and power densities than renewable fuelsderived from agriculture. An all-solar future with solar technologieswould be radically different than one without them. The research andinfrastructure needed for an all-solar future should be developed nowwhile we have the luxury of high energy ratios from fossilfuels.