Shining a Light on the Limits of Vertical Farming: The Cost of High-Calorie Crops

Revisiting the role of Vertical Farming in achieving food security

In our previous article, we presented calculations on vertical farming (VF) and highlighted that increasing productivity is crucial to reducing the cost of low-calorie plants. The cost of electricity and energy efficiency are important but not the deciding factors.

But high-calorie plants, like wheat and rice, require more time to grow as they accumulate carbohydrates, proteins, and fats for their seeds, which also necessitates greater light intensity. How would that change the applicability of VF for these crops? How will producing high-calorie crops in a VF affect their cost and what are the limitations of agriculture in a completely controlled environment? These are the questions we aim to address in this post.

The reported theoretical limit of VF productivity for wheat (which we’ve chosen as a model plant) is 19.5 kg/m2/year. While precise cost estimation associated with that level of productivity is tricky, we can assess how much electricity is required.

The energy efficiency of photosynthesis is ~2%, meaning that only 2% of the electric energy spent for lighting is converted into chemical energy in leaves (we are going to cover a detailed breakdown of energy losses in our future posts). Furthermore, wheat has a low harvest index, with only one-third of the plant by mass being edible, making light conversion efficiency into edible calories about 0.6%. Even with an optimistic outlook on potential agrotech improvements, with the high energy content of wheat grain (3,600 cal/kg), electricity alone costs around $19/kg (given current electricity prices in the US). Therefore, even a 65-fold reduction in electricity cost is insufficient to match the market price of wheat ($0.29/kg) due to other VF costs. 

However, controlled growth conditions might result in significant increases in productivity, so we agree with those who see a future in the combination of controlled indoor growth with natural sunlight. But, first, let’s explore how much energy is needed to grow crops in a controlled environment.

The Advantages of Controlled Environment Agriculture (CEA)

Our energy cost calculation starts by examining wheat productivity in CEA, which was a subject of extensive scientific research. Astronautics is the area where pushing crop productivity to its limits is a very important question. Notably, NASA has conducted research into how they could supply future bases on Mars or the Moon, with the first experiments involving LEDs to grow plants conducted in 1986. Companies such as Plenty, Bowery, and CSS Farms have leveraged NASA's plant-growth research to develop their VF businesses, with several Bowery employees having worked on NASA-funded plant tests in a simulated space habitat in Antarctica, as well as the chief engineer at Green Sense Farms. NASA plant scientist Gary Stutte, after his retirement, joined Eden Grow Systems as the director of research and development.

The findings of NASA-supported research indicate that maintaining wheat in optimal and stable growth conditions can increase the yield from the world average of 0.3 kg/m2/harvest to 1.5 kg/m2/harvest (per one layer of growth area). Furthermore, optimizing growth conditions reduce the growth cycle to 70 days, enabling five harvests per year and yielding 7.5 kg/m2/year. Theoretically, this yield can be further enhanced up to 19.5 kg/m2/year by augmenting CO2 concentration and increasing the intensity of light.

Credit: Senthold Asseng et al, PNAS 2020.

The photosynthesis efficiency

However, these impressive results come at a cost. In the experiment that resulted in 1.5 kg/m2/harvest, one must provide light at an energy equivalent rate of 22 MJ/m2/day. This means that, to our rough estimate, above 650 kWh of electric energy would be consumed for lighting per kilogram of wheat grown in controlled conditions (44% LED energy efficiency is assumed). Compare that to the energy value of wheat, which is 4.23 kWh/kg. The energy efficiency for wheat is therefore equal to 0.65%, whereas commercially grown leafy greens, such as lettuce, have a practical efficiency of about 2%. This difference can be attributed to lettuce having a high harvest index (i. e. the ratio of edible mass to the total plant mass), while wheat has a relatively low harvest index of nearly 0.34. We don’t eat the stems, leaves, and roots which contain a significant portion of nutrition, produced in photosynthesis (66%). Increasing the harvest index to make photosynthesis work for its primary function, namely the creation of nutrients for food, not chaff,  is a crucial aspect of crop breeding and genetic engineering to improve agriculture's sustainability.

This diagram explains why the energy efficiency of photosynthesis is so low. It takes into account the minimum theoretical loss at every stage of energy conversion, from powering the LED, to harvest. In reality, the losses will be greater, for example, due to the emission of LED light to the sides, or a low harvest index. We will cover this in more detail in the next article.

Currently, crop plants have a 0.5 to 2% efficiency of converting light into edible calories, with sugarcane being an exception with an efficiency of 8%. In the future, we can expect energy efficiency levels of 5% to be reasonable for some commercial crops, including wheat. In our opinion, this is the upper limit of a rather optimistic, but still reasonable assumption. This can be achieved, for example, by increasing the efficiency of LEDs from 44% to 55%, doubling the harvest index, and thorough selection and genetic engineering for improving the efficiency of light capture and photosynthesis in the leaf.

VF will stay niche

Let’s (just for fun) now calculate the amount of electric energy necessary to sustain global demand for wheat if grown in VF, assuming the increase in wheat energy efficiency from the current 0.65% to an optimistic 5%.

Based on the energy content of wheat grains (3,600 Cal/kg) and the assumed photosynthesis energy efficiency, the estimated minimal electricity consumption is 85 kWh per 1 kg of wheat grains. Additionally, roughly 28 kWh/kg should be allotted for dealing with heat, as a significant portion of energy consumption in modern vertical farming is dedicated to climate control. With 5% photosynthesis efficiency, 95% of photons emitted by LEDs are converted into heat (at various stages of light emission/absorption), which requires extra electric energy for a heat pump. In our estimation above, we assumed a heat pump coefficient of performance (COP) equal to 3. This means that for every 3 kWh of heat extracted from the growth area, an additional 1 kWh of electric energy is required to power the air conditioner.

Wheat is grown on a vertical farm, as represented by the Midjourney

The total energy consumption amounts to nearly 113 kWh/kg. If multiplied by the global annual consumption of wheat (782.7 million metric tons in 2022, according to USDA data), we arrive at 90,000 TWh/year for producing this quantity of wheat through vertical farming. When we add rice and corn, the total amount of electricity required increases to 240,000 TWh per year. To feed the Earth’s population using artificial lighting, the global annual electricity generation would need to increase by almost 10 times the current value, which was approximately 25,000 TWh in 2021. This would require a land area equivalent to that of Egypt to be covered entirely with solar panels. The point of this exercise is to give an idea of how much of the Sun’s energy is captured by agriculture directly. Pushing all this energy through grids does not make any sense, which is yet another perspective to assess the potential applicability of VF.

Let’s do a bit more fun math. To sustain an astronaut on Mars with underground vertical farming, we would require approximately 30,000 kWh annually. This estimate is based on the consumption of 700 g of wheat equivalent per day for a 70 kg human. To generate this amount of electricity, we would need 150 m2 of solar panels, taking into account energy losses and assuming insolation on Mars of 1,000 kWh/m2/year.

And, as we mentioned in the intro, with an energy consumption of 113 kWh per 1 kg of wheat, the energy cost of producing 1 kg would be nearly $19 (based on an electric energy price of $0.165/kWh, the average US wholesale price as of December 2022), while as of January 30, 2023, the market price of wheat is just $0.29/kg.

So, what does it all mean?

We hope it is clear that using virtually free sunlight is necessary to make CEA viable for nutrient-rich crops such as cereals and legumes. In its current form: indoors with artificial light, VF is not about food security or comprehensive agricultural transformation and LEDs powered by electricity from traditional power plants will harm the environment. Sunlight, absorbed by solar panels, undergoes an overly complicated and ineffective chain of transformations before being re-emitted by LEDs and processed by plants.

The idea of using natural sunlight to reduce electricity costs in VF is not new and is at the core of the VF 2.0 concept. However, existing proposals are well-suited for leafy greens and vegetables with high water content. Sunbathing cereals at the uppermost level of a VF, at the latest stage of growth, is insufficient for drastic cost reduction, and returning to a greenhouse concept (where there is only one layer of plants) implies inefficient use of space.

In the upcoming post, we will speculate about CEA for economically viable and sustainable food production and review how CEA could efficiently combine indoor farming, efficient land use, and sunlight. Stay tuned for some serious science!