Energy up in Smoke:
The Carbon Footprint of Indoor Cannabis Production



  1. Peer-reviewed study, including updated and expanded analysis

  2. Presentation

  3. Infographic

  4. Energy benchmarking survey.

Media Coverage


What kind of facility contains lighting as intense as that found in an operating room (500-times more than needed for reading), 6-times the air-change rate of a biotech laboratory and 60-times that of a home, and the electric power intensity of a datacenter?

The emergent industry of indoor Cannabis production results in prodigious energy use, costs, and greenhouse-gas pollution. Large-scale industrialized and highly energy-intensive indoor cultivation of Cannabis is driven by criminalization, pursuit of security, and the desire for greater process control and yields. The practice occurs across the United States and in many other countries.

The analysis performed in this study finds that indoor Cannabis production results in energy expenditures of $6 billion each year--6-times that of the entire U.S. pharmaceutical industry--with electricity use equivalent to that of 1.7 million average U.S. homes. This corresponds to 1% of national electricity consumption or 2% of that in households. The yearly greenhouse-gas pollution (carbon dioxide, CO2 ) from the electricity plus associated transportation fuels equals that of 3 million cars. Energy costs constitute a quarter of wholesale value.

In California, the top-producing state—and one of many states to allow cultivation for medical purposes—the practice is responsible for about 3% of all electricity use or 9% of household use. Due to higher electricity prices and cleaner fuels used to make electricity, California incurs 70% of national energy costs but only 20% of national CO2 emissions.

From the perspective of individual consumers, a single Cannabis cigarette represents about 10 pounds of CO2 emissions,* or an amount of electricity equal to running a 100-watt light bulb for 75 hours with average U.S. electricity (or 135 hours on California’s cleaner grid). Each four-by-four-foot production module doubles the electricity use of an average U.S. home and triples that of an average California home. The added electricity use is equivalent to running about 90 refrigerators. Processed Cannabis results in 4,600-times its weight in CO2 emissions. For off-grid production, it requires 70 gallons of diesel fuel to produce one indoor Cannabis plant, or 140 gallons with smaller, less-efficient gasoline generators.

Minimal information and producer consideration of energy use, coupled with adaptations for security and privacy, lead to particularly inefficient configurations and correspondingly large energy consumption and greenhouse-gas emissions. If improved practices applicable to commercial agricultural greenhouses are any indication, the energy use for indoor Cannabis production can be reduced dramatically. Cost-effective efficiency improvements of more than 75% are conceivable, which would yield energy savings of about $25,000/year for a generic 10-module growing room. Shifting cultivation outdoors eliminates most energy uses (aside from transport), although the practice can impose other environmental impacts (see this Google Maps tour of a small corner of Humboldt County), such as poisoning of animals by rodenticides and other chemicals used by growers.

  1. *Note: an earlier version of this equivalency calculation assumed the average “joint” weight of 0.33 grams.  A more typical value is 1.0 grams, which effectively triples the original estimate of 3.5 lbs CO2 per joint.


What is the purpose of this study?

This study simply aims to quantify a previously undocumented component of energy demand in the United States, to understand the underlying technical drivers, and to establish baseline impacts in terms of energy use, costs, and greenhouse-gas emissions.  This study does not pass judgement on the merits of Cannabis cultivation or make recommendations for how to reduce this energy use, but observes that many reversible inefficiencies are embedded in current practices.

Who is the author of this report and who sponsored it?

The research described in this report was conducted and published independently by Evan Mills, Ph.D., with no external sponsorship or institutional affiliation. Dr. Mills is a scientist specializing in energy analysis and the role of energy in climate change, with a focus on the efficiency of energy use as the number-one strategy for reducing climate change. He has been working in this field for 30 years and is currently a Senior Scientist at one of the national laboratories as well a Research Affiliate with the Energy and Resources Group at U.C. Berkeley.  However, this study was done on the author’s own time.  More information on the author is available here.

Has the media covered the story accurately and responsibly?

The problem is a lot more complicated and nuanced than most pundits in the blogosphere make it out to be, and most in the mainstream media have actually missed the real story (a roster of articles is provided at the bottom of this page). Among the (few) relatively thorough, non-selective, and un-spun stories are those in the New York Times, FastCompany, Southern California Public Radio, San Francisco Bay Guardian, Times Standard, and the Huffington Post.   First place for the most inane coverage is an editorial from the Calgary HeraldThe Week did a particularly inept job at providing an objective cross-section of the coverage.

What’s missed or bungled:

  1. A number of media reports inaccurately associate the work with Lawrence Berkeley National Lab. The work was actually conducted independently of LBNL on the author’s own time.

  2. This isn’t simply a California issue, let alone a “counter-culture” or “hippie” issue.  It’s mainstream and it’s pervasive. The UN says 10% of North American citizens use Cannabis on a regular basis, and most of the production is actually not even in California any more….  Only 1/5th of national Cannabis-related emissions come from California (and an unknown fraction of that is for products consumed outside the state).

  3. Some reporters have spun the story into a blame-game rather than a discussion of what to do about a collective problem. Responsibility is often dumped exclusively on the growers’ doorstep rather than the consumers and other intermediaries (e.g. dispensaries) who can powerfully influence the energy choices being made.

  4. Few reports recognize that there are solutions. There’s a lot of room for a more virtuous cycle…. Using good-old market forces (see final entry on this FAQ page). This is an energy-using sector that has almost uniquely been passed over by decades of efforts from many quarters to improve efficiency …. The time for that has come and there is a lot of low-hanging fruit, which is good news to everyone concerned.

  5. Legalization is a simplistic answer.  The reality is that people will seek to grow indoors irrespective of their legal status.  More has to be done if carbon emissions are to be curbed.  More on this below....

What information sources were used for this work?

Data and assumptions for building the energy-demand model were drawn from the open literature, trade press, equipment specifications, and interviews with horticultural equipment retailers.  A detailed list of sources and technical assumptions are provided at the end of the report and in the notes to the data tables.

Does this report pertain only to medical Cannabis?

No. A number of media reports have misreported this.  The analysis pertains to all indoor Cannabis production. 

Isn’t the solution to just grow outdoors?

Complicated.  Many of the issues, and emerging opportunities (including achieving comparable quality in vs out) are discussed here.

Why are you singling out Cannabis when so many other things use more energy?

The study hardly singles out Cannabis. On the contrary, there has been a de facto double standard for decades in which virtually every other energy using activity has been met with efforts to improve efficiency and reduce emissions.  Examples of these efforts include education, product labeling, financial incentives, mandatory efficiency standards, and R&D.  Cannabis should no more be singled out for attention than it should be overlooked.

How certain is the national estimate of 10,000 tonnes/year (indoor+outdoor) reliable?

There are very few estimates of overall domestic production. For this analysis, I used the lowest value I found (see footnotes in paper - gov't source), and it dates from many years back.  Keep in mind that production greater than consumption, as unknown proportions are lost during production (to problems like mold outbreaks, fires, disease), as well as due to interdiction by the authorities. Not all production is smoked; significant proportions are diverted to extracts and edibles. No effort is made to account for net imports/exports from the US. There are no doubt huge uncertainties about the level of production, but they do not materially effect the essential findings - energy use and associated carbon emissions are significant and heretofore unaddressed.

OK, how does the energy use of this sector compare to others?

  1. Normal indoor cultivation settings (i.e., greenhouses) - According to a paper by Larry Kinney, indoor greenhouses spend about $5 per square foot per year on energy (at average US energy prices).  At those same electricity tariffs, indoor Cannabis is 11-times that number at $55/square foot.  But, at the prices actually paid in that industry it’s way higher than that (4x higher in California), since tiered tariffs or diesel generators involve much higher energy costs. According to another study, the energy use for indoor horticulture (all crops) is 7% of total energy use in the Netherlands.  According to research from the Netherlands, the production of ornamental flowers is 2.2 GJ per 250 flowers.  This translates to about 1,200 kilowatt-hours (including indirect embodied energy for fertilizer, pesticides not counted in the Cannabis analysis).  For comparison, the annual electricity use of a 4x4x8-foot Cannabis cultivation module is estimated at about 13,000 kWh. It is hard to identify the right “apples-to-apples” comparison metrics. Another study for Belgium places national energy use for greenhouses at about 1000 MJ/m2, which is 1% that of indoor Cannabis production.

  2. Other industries - Defining “efficiency” as how much energy is required to generate economic value, Cannabis comes out the highest of all 21 industries (measured at the three-digit SIC level).  At ~19 kBtu’s per thousand dollars of shipment value (wholesale price), Cannabis is followed next by paper (~14), nonmetallic mineral products (~9), primary metals (~8), petroleum and coal products (~6), and then chemicals (~5).

  3. Pharmaceuticals - Energy represents 1% of the value of pharmaceutical shipments and 20-50% of the value of Cannabis wholesale prices.  The US "Pharma" sector uses $1billion/year of energy; Indoor Cannabis uses $6 billion.  The gap in actual energy use is far smaller given the relatively high tariffs paid by Cannabis growers versus cheaper industrial tariffs paid by Pharma operations.

  4. Greenhouse horticulture - Cannabis production requires 13-times as much energy per square foot as greenhouse horticulture in Belgium.

  5. Alcohol - The energy used to produce one "joint" would also produce 18 pints of beer.

  6. Other building types - Cannabis production requires 8-times as much energy per square foot as a typical US commercial building (4x that of a hospital and 20x that of a building for religious worship), and 18-times that of an average US home.

Certainly, Cannabis production is not the “main” use of energy in the world...but it is too important to overlook. For example, the approximately 22 billion kilowatt-hours/year estimated for indoor Cannabis production is about one/third that of US data centers or one-sixth that of US household refrigerators.  The shares would be much higher in states where Cannabis cultivation is concentrated. Automobiles are responsible for about about 33% of U.S. greenhouse-gas emissions, which is 100-times as much as indoor Cannabis production (0.3%). 

Doesn’t the mainstream pharmaceutical industry use more energy?

The traditional ‘Pharma’ sector has received much more attention than has Cannabis. Energy represents about 1% of the value of pharmaceutical products, versus 20-50% of that for Cannabis (depending on local energy prices).  Total energy expenditures in Pharma were just under $1 billion for the year 2002.  Our estimate for the Cannabis industry is $6 billion per year. Typical pharmaceutical facilities are likely less energy-intensive--in terms of energy use per square foot---than indoor Cannabis cultivation environments.  Arguably, the mainstream Pharma industry already does much more than the Cannabis industry to reduce its carbon footprint...    That said, from a "carbon-accounting" perspective. it would be challenging--if at all possible--to meaningfully go about a deeper comparison to other Pharma, since to do that you'd have to have an apples and apples comparison not only of the energy/carbon (relatively easy) but also the impact benefit at the macro level. It has been reported that the global Viagra market is about the same size as the US medical Cannabis market; i will leave it to others to debate the relative carbon footprints.

Um...., what about alcohol?  Yeah, why don’t you pick on alcohol?!

Great question.  The energy use required to produce one marijuana cigarette is equivalent to that of about 18 pints of beer. Much more effort has been placed on improving efficiency in alcohol-production facilities than in Cannabis-production facilities.

You’re making it hard for me to justify my efforts to diminish the importance of this...  OK - aren’t we barking up the wrong tree? Doesn’t law enforcement spent even more energy on suppression and eradication?

Not even close. While worthy of a study in and of itself, but per a quick look: Total federal government energy expenditure for vehicles for all purposes is less than half that of Cannabis production and the government’s jet fuel for aviation, including military (see page 96 of this report) is “only” $7 billion per year (including overseas operations) -- that was FY 2007, so maybe more now. Total federal government for aviation gasoline (small planes) is $10 million/year...  In any case, two wrongs don’t make a right - seriously.  We need to reduce emissions wherever they occur. If we do nothing but point to others who emit more we will never get out of the greenhouse.

You say that the resulting electricity intensities are 200 watts per square foot.  This is absurd.

This point has somehow been mangled and misinterpreted as being the constant rate of energy use for lighting.  It's actually the installed power density of lighting plus all other equipment, and none of those "watts" are assumed to be running 24x7.  For example, for lighting this assumes 600W-MH (37.5 W/sf) for the vegetative stage and 1000W-HPS (62.5W/sf) for flowering (summing to ~100W per the definition of power density -- but for energy calculation purposes the 600W and 1000W devices are never drawing power simultaneously).  The rest of the demand is for all the other loads (AC, dehumidification, fans, ballasts, etc.). In any case, the 200W/sf value is a side note, and not integral to the energy and emissions calculations.

Is Cannabis production intrinsically and unavoidably polluting?

No.  Like virtually any other energy-using activity (driving, preparing food, making aluminum, heating a home) energy is commonly wasted and used inefficiently.  And, as observed in virtually every energy use domain, there are enormous prospects for improving efficiency and using low- or no-carbon energy sources.  Outdoor production involves particularly low energy inputs, although when mismanaged, the practice imposes other environmental impacts.

How significant are the energy costs for producers?

The answer varies widely depending on the production method, prevailing energy prices, and wholesale prices of the final product.  Averaged nationally, about one-quarter of the wholesale price is attributed to energy costs.  In regions with high electricity tariffs or the use of inefficient off-grid power generators, this value can approach half the total cost.

Wouldn’t just “going solar” solve this problem?

Nope.  Using solar to power inefficient systems is a waste.  A costly, mega-sized solar system to run a bunch of inefficient lights, fans, and chillers in a single house would just give solar a bad name.  For the same investment, many dozens of conventional houses could be made highly energy efficient and then be solar-powered all the way to zero carbon emissions.  This would provide much more carbon-bang for the buck. That said, once energy efficiency is maximized, then solar would be an excellent next step.  Given the particularly high electricity rates paid with ‘inverted-block’ tariff structures, solar applications in this domain could be particularly cost-effective. Once energy efficiency is maximized then, yes, it makes sense to introduce solar.  In fact, arguably the only truly acceptable indoor production approach is a zero-net-energy setup in which all electricity needs are met with solar.

Does this study enable or endorse illegal activity?

No. On the contrary, the analysis puts a bright light on certain adverse consequences of Cannabis cultivation, and highlights certain perverse and illegal outcomes such as the theft of power or the inappropriate use of low-income energy tariffs and abuse of low-income energy tariffs intended for the poor (not for well-heeled pot growers).

Does criminalization keep energy costs in check?

No. In fact, criminalization is an important driver of energy-intensive indoor production, as it is easier to conceal than outdoor production.  Criminalization also contributes to many of the energy inefficiencies of the process, including transport of small quantities of the product over long distances, noise and odor suppression measures that undercut ventilation efficiencies, and off-grid power production with diesel generators that are far less efficient and produce more greenhouse-gas emissions than many electric grids. The high prices made possible in a black market foster reduced concern about energy costs.  Moreover, decades of criminalization have resulted in this sector being passed over by otherwise massive efforts to incentivize and mandate efficiency improvements in virtually every other part of the economy. As a result there have, by definition, been no standards or regulatory interventions to clean up the production process. Facilities with zero net use of grid energy and zero greenhouse-gas emissions releases (via highly energy-efficient processes coupled with renewable power sources) are certainly feasible, but are more difficult to achieve under the status quo.

What effect will legalization have on energy use?

It remains to be seen.  People cultivate indoors for many reasons aside from evading law enforcement, e.g., climate control, quality control, pest control, and year-round production. Many legally sanctioned producers choose to grow indoors, and legal dispensaries deem Cannabis produced outdoors to be inferior.  It is not known whether legalization will reduce or increase the rate of indoor cultivation, or whether the new industrial-scale operations will be more or less energy intensive than the historically smaller-scale methods. However, thus far local authorities are requiring that cultivation be doing indoors, compounding the energy woes.  That said, in a scenario where production is legalized it is easier to address the energy issues through voluntary and/or mandatory energy policy measures such as codes, standards, incentives, and product labeling. If prices decline following legalization, the prodigious energy costs currently associated with indoor cultivation may become untenable.  Legalization is thus a necessary but not sufficient step in addressing the carbon footprint of indoor cannabis production. 

What are the take-aways for policymakers?

The unanticipated growth of energy use and greenhouse-gas emissions from indoor cannabis production threatens the attainment of energy and climate goals, particularly in areas where the industry is growing most rapidly. For example, energy use per household in Humboldt County (California) is 50% higher than the statewide average, and this is a relatively mild climate.  Another way to look at this is that the growth in this rogue source of energy is masking seven electric power plants worth of energy savings that are taking place thanks to energy-efficiency efforts elsewhere.  By under-counting the energy use of cannabis in official statistics, we are over-counting it elsewhere.

Perhaps the biggest elephant in the room is that the current objectives of drug and medical policy are not at odds with those of energy and environmental policy.  This conversation has to happen.  Meanwhile, many local ordinances are requiring cannabis production to be conducted indoors, without adequate regard to the energy implications.  The issue is particularly acute in extreme climates: imagine the energy required to maintain indoor temperatures of 80 degrees when it's 120 outside or -20 outside.  From an energy and environmental standpoint, not every state is well-suited to this practice.

Policymakers are faced with a vexing Catch-22 insofar as, in the absence of legalization, federal and state energy agencies are hampered from analyzing the issues and yet once legalization occurs, critical information will be lacking to direct the gold rush of investment in a sustainable direction.  In one example of how things are lagging, we lack building energy codes as well as LEED and ENERGY STAR designations that are relevant for facilities already being constructed.  

A strategic vision is needed. Even with very high levels of energy efficiency, the energy intensity of indoor cannabis facilities will eclipse that of more conventional uses such as office buildings.  The only tenable approach to indoor production is to set and strictly adhere to a zero-net-energy standard, i.e., to require on-site renewable energy production that completely offsets all energy requirements of the facility.  Performance has to be independently verified to ensure that promising design estimates are borne out and persist in practice.

What can be done with the results?

Carbon footprint could be reduced by perhaps 75% without even shifting production outdoors (which reduces emissions almost completely). Top-level pathways to reducing the carbon footprint of indoor production include:

  1. Growers selecting better, commercially available equipment.

  2. Equipment vendors developing even more efficient equipment, and educating their customers.

  3. Reducing the use of off-grid power generators fired with fossil fuels. The worst case is a gasoline-fueled generator, which results in 140 gallons of fuel burned to produce each plant.

  4. Applying science to understanding how to achieve necessary environmental conditions in a less energy-intensive manner.

The opportunity to influence a constructive, climate-friendly response rests with all involved parties.

  1. Producers lack information, motivation, and market or regulatory pressure to improve.

  1. Designers and manufacturers of the energy-using equipment could more precisely analyze and consider the issues from an engineering and market standpoint.

  2. Medical dispensaries could support more responsible consumer decision-making, by providing disclosure of product carbon content and other dimensions of environmental footprint.

  1. Planning and building officials at the city, state, and federal level may choose to seek better understanding of the energy and safety consequences of this activity in their localities. Some (Berkeley, Boulder, Fort Bragg) have already made steps in this direction.  Currently there are no coherent building codes, or LEED or ENERGY STAR standards that can be used in designing or retrofitting grow facilities.

  1. Electric and gas utilities have already begun to recognize legal producers, granting them lower (agricultural) tariffs in exchange for safety inspections.

  1. Energy providers, policymakers, and forecasters could better account for this particular driver of energy demand, and thus more accurately evaluate the effects of unrelated programs and policies on the consumption of energy at the macro scale. 

  1. Consumers and the general public can become informed about the carbon footprint associated with this practice and better consider the environmental consequences of their actions.