Since there are no currently active contests, we have switched Climate CoLab to read-only mode.
Learn more at
Skip navigation
Share via:


We use technology to decentralize fertilizer production, enabling rural farmers restore their soil health/carbon and improve harvest yield.



Most fertilizers today are produced in large-scale, centralized, and capital-intensive facilities, and then shipped via long-distance transport to rural markets. Due to this logistical mark-up, rural farmers often pay 2-3 times the world price for fertilizers. Due to their limited income, these farmers can often only afford the cheapest, synthetic, one-size-fits-all varieties that over the long term may acidify and degrade their soil due to over-dependence. Farmers, as a result, often see their post-harvest yields decrease by up to 30-40% in the last 20 years. This is a significant concern for local food security. In the world, by 2030, this will be a $30 billion/year problem.


We use a combination of MIT-developed, patent-pending technology and trade secret to decentralize fertilizer production, such that it can be carried out profitably on a village-level scale using locally available resources/labor. Imagine small-scale, low-cost, mobile systems that can be latched onto the back of tractors or donkey carts, and be moved from farm to farm to enable localized conversion of crop residues into a fertilizer base under 30 min without external energy (100 times faster than our competitors). This base is then mixed with our proprietary recipes to complete the nutrient as standalone fertilizer. Our process reduces the long-distance logistics needed to deliver fertilizers to farmers by producing the bulk locally. In addition, our proprietary process can custom-tailor the fertilizer composition in different types of soil conditions and crop needs at an almost single-farm granularity. Our carbon-negative product restores local soil health and improves farmers’ yield by up to 30% for the same cost that they pay for conventional inputs. Also, as our fertilizer is rich in inert carbon, applying a ton of it sequesters 1.7 tons of CO2e. Thus, we promulgate a new way of fertilizer production that both restores soil health and makes smallholder farming net carbon-negative.

Is this proposal for a practice or a project?

Not sure

What actions do you propose?

We have already successfully tested the reactor prototype, and so far, more than 150 systems have been successfully deployed in Kenya. The prototype utilizes a new chemical concept developed at MIT called oxygen-lean torrefaction as a rapid way to break down crop residues into constituent nutrient components 100 times faster than current alternatives (e.g. composting). MIT has filed a patent on the core technology, and the company has taken out an exclusive option/license from MIT to commercialize further. The company has also developed a suite of our own in-house proprietary recipes that allow the technology to be applicable for different types of crop and soil conditions, allowing us to custom-tailor the fertilizer blend almost at a single-farm granularity. Further details about how the technology works can be found in a series of peer-reviewed articles cited in the "references" section, as well as in co-founder Kevin Kung's MIT PhD thesis.


Initially, we set up what we call a “MiniPlant” model, which reflects a localized, village-based fertilizer production unit serving a farming community of a radius of 10 miles, or around 1000 farmers. This MiniPlant purchases the crop residues from the local farmers, operates the equipment to convert the residues into fertilizer, packages the final product, and distributes the Safi Sarvi fertilizer back to the same community via an existing local network of agricultural input suppliers to the farmers. So far we have already established one such MiniPlant in Mwea, Kenya, and we have shown that it can support a team of 10 full-time employees while being financially profitable. We also have around 3,000 paying customers subscribing to our Safi Sarvi fertilizer. Below, we dissect the MiniPlant production model in greater detail.


In order to scale up, we must set up MiniPlants in other communities. We have already identified candidate communities in which we know the prospective local partners to which we can expand. From our prior experience, the most effective way to set up a new MiniPlant is that we begin by involving the wealthier, older farmers in a village, who not only have larger risk-taking appetite but also serves as the role model for aspiring younger farmers. In a typical trial, these farmers set aside a small fraction of their land for using our product, such that they can compare our product’s results to the status quo after one season. When they noticed the immediate benefits even after one season, they started giving raving reviews, which spread by word-of-mouth in the tightly knit local farming community. To help visibility we also put up signposts near trial plots, attend local agricultural trade fairs, and collaborate with the government’s agricultural initiatives for training smallholder farmers. In the initial phase, we will own and operate the first 5 of these MiniPlants ourselves.


Take the example of Mr. Kibuchi, a rice farmer in Mwea, Kenya who owns a 1-hectare plot. For the past 20 years, he has been dependent on urea and other synthetic, acidulated fertilizers that are imported, paying about $520/ton for such product. In recent years, he has seen the crop yield gradually decline. In 2015, our marketing team reached out to him and he used our fertilizer. Immediately in the next harvest season, he saw an improvement, and after 6 growing seasons, he now completely uses our product and has seen his income boosted by 60%. That is sufficient to send 2 children to school and buy a second hand tractor for his farm last year. Through the past few years, we have already impacted around 3,000 farmers similar to Mr. Kibuchi. Beyond anecdotal evidence, we have also begun collecting data with a small group (50) of farmers, ranging from soil pH changes to harvest yield data. While the data are still preliminary, we have observed a pH change of on average 0.7 points one harvest season immediately after using our product. While the extent of improvement is a function of how degraded the original soil condition is, on average, this group has reported an increased yield of around 25% and increased income of around 40%. In order to causally link the fertilizer’s mechanism to the impact, we have also begun a series of crop growth trials with research collaborators at MIT.


As this scaling is taking place, we will also need to muster the resources to provide both the hardware technology as well as the training and expansion in new communities. The hardware is designed to be low-cost and manufacturable using off-the-shelf components. We will work with local metal manufacturing partners to train them in the fabrication of the units. After initial scaling at a manufacturing partner, we intend to further scale by collaborating with and licensing to existing agricultural equipment companies who already an existing distribution and dealership network for their tractors and similar products. Our equipment will be complementary to their offering, and indeed can be latched onto the back of their tractors. We have already been actively approached by two agricultural equipment manufacturers about partnership. We will use their existing equipment distribution network and dealership to get product to market. In terms of the training, eventually, as we scale past 5 MiniPlants, we will start working with other field partners, who will eventually own and operate our solution in their own villages. These partners may include: local agricultural cooperatives (who are often subsidized by the government for hardware investment), local NGOs serving farmers, as well as local microentrepreneurs. We have already gathered letters of intent from at least 7 local partners, available upon request.


In order to ease adoption and financing barrier amongst prospective field partners, our technology not only costs 100 times less, but also has a uniquely tiered and flexible financing structure that avoids upfront investment risk. Before we roll out our high-performance, continuous systems in a community, we have an open-source, low-performance, and flawed version of the technology that nonetheless can prove many things about the localized fertilizer business. Community partners interested in our technology are encouraged to commit $20 in upfront investment to locally fabricate a unit of our preliminary technology—freely available on the internet—to produce the fertilizer base and engage local farmers to use it. This is also a way for us to gauge and screen for the most committed and capable community partners to work with. Our high-performance, continuous systems will only be rolled out in localized fertilizer operations that already have active farmer-customers and are nearing financial breakeven. We will provide the training to these organizations’ employees to run our systems in the field. We will also work with public players such as Kenya Agricultural and Livestock Research Organization in incorporating our process (already certified by the Kenyan government) into its community outreach workshops.


Ultimately, in order to scale impact and carbon sequestration, we use a market-driven approach, charging local farmers the maximum that they are willing to pay for our fertilizer (which is equivalent to the price that they currently pay for conventional fertilizers), because from our experience farmers tend to eye more cheaply priced products with suspicion, thinking they are of inferior quality. We have demonstrated that, from the village-based MiniPlant perspective, the sales revenue exceeds the cost of localized fertilizer production by a margin of 40% currently. Therefore, as a growing number of farmers become repeat customers, the local MiniPlant will also become financially profitable beyond a breakeven production scale of 1 ton/day. We estimate that, at full production scale, the net profit for a typical MiniPlant is about $60,000/year. This is sufficient to comfortably support a full-time local production team in terms of livelihood. In order to sustain ourselves financially, we will assess each operational and profitable MiniPlant a per-ton usage fee to continue to operate our system. This per-ton usage fee will be enforced by our proprietary automated control system capable of sensing/running the core reaction safely and stably without needing any external energy, maintaining a consistent fertilizer quality in spite of the highly variable biomass input characteristics. Every time our partner turns on the reactor, our control system will know, and will debit the usage fee from the partner’s account.


Furthermore, as other communities observe the MiniPlant as a profitable model, local profit-minded entrepreneurs and impact-minded community organizations express interest in working with us to replicate the model in their own communities, allowing us to pool together resources to bring the solution to additional communities. As more communities adopt our model due to a combination of profit and impact-driven motives, this will also allow our social impact to be scaled across greater number of farmers to benefit their soils as well. Finally, because our product is formulated as rich in inert carbon, as more communities adopt our process, more carbon will also be automatically sequestered into the soil, and less crop residues will be burned in the open air. This drives reduction in pollution and carbon, resulting in a carbon-negative agriculture that also restores soil health over large areas in the long term.


In terms of regulatory approvals, when the fertilizer trials and pilots are done at a small scale, from our past experience, such activities have been exempt from approvals. However, as we start commercially selling the product, especially by working with existing agricultural input distributors, it is necessary to obtain official fertilizer certification from the government. After two years of effort, we are glad to report hat our product has been officially certified to sell as a commercial fertilizer as of early 2018. As we expand this work to other countries, it will also be necessary to seek the regulatory approval of these other countries. This will take time and effort, but we plan to do so with the local implementation partners who are well-versed in the locally relevant agricultural regulations. On the other hand, as our MiniPlant has been demonstrated (both at MIT and in the field) to produce no harmful emissions, and as our MiniPlant often operates at a scale smaller than what would be normally considered an "industry", we have found that our MiniPlant activities have been exempt from any of the other regulatory approvals related to emissions or commercial activities, beyond duly incorporating and registering the company with the local authorities.


In terms of weather-related eventualities such as floods and droughts, they do and will have an impact on our work, because if farmers decide that it is not the appropriate time to plant, then our fertilizer blend will not be purchased. Initially, when we began our work with rice farmers, we only had orders when the rice farmers were planting (twice a year), and when there was a drought throughout Kenya in 2017, it indeed hit our business hard. Since then, we have expanded the suitability of our solution to other farmer segments, ranging from horticulture to tea. As the different farmers have different planting seasons and weather dependencies, the versatility of our solution has in effect diffused the weather-related risk by making our product have consistent demand year-round, in many types of weather conditions. Indeed, we have found it a convincing argument to make to the farmers that, in the event of a drought, it is in their interest to lean on our product, which helps retain moisture in the soil more effectively and thereby reduce the need for external irrigation/rain while achieving the same level of growth and yield.

Who will take these actions?

The actions will be driven by Safi Organics, in partnership with Takachar and other partners.


Safi Organics is a Kenyan fertilizer company whose mission is to enable rural farmers access to affordable, high-quality fertilizers. On the other hand, Takachar is a spinout company from Massachusetts Institute of Technology (MIT) to commercialize the decentralized biomass innovations developed in Professor Ahmed Ghoniem’s lab under the auspice of the MIT Tata Center for Technology and Design.


Safi Organics has partnered with Takachar since 2015 in setting up an initial field pilot based on a preliminary version of the MIT technology. Safi Organics will continue to lead the scale-up of the community-based fertilizer production model, with technical input from Takachar.


In order to scale the production beyond one pilot community, Safi Organics will partner with the Kenya Agriculture and Livestock Research Organization (KALRO), which serves as an extension program for many rural farmers, and conduct regular trainings in different farming networks. Through its network, additional village-based partner organizations such as local agricultural cooperatives and NGOs will be identified that can set up and operate the village-based fertilizer production for profit. We already have letters of intent from seven prospective organizations.


Massachusetts Institute of Technology (MIT), from which the core technology in this proposal was original developed, continues to be interested in and is pursuing further testing on the production process and improvements. We will also continue to learn and improve from the MIT experience. Furthermore, Takachar works with the Lawrence Berkeley National Laboratory under a Collaborative Research and Development Agreement (CRADA) for product analysis, and with UC Berkeley for impact and lifecycle assessment.


In the long term, as we scale up, we also plan to work with traditional fertilizer companies. While people may consider these companies our direct competitors, from our past experience interacting with representatives from these companies, we take a different view. Indeed, Incumbent, large-scale fertilizer producers (e.g. OCP, Urakali) are actually very interested in accessing smallholder farmers in emerging markets, but have historically had trouble getting their products to them due to the aforementioned logistical challenges. Recently we have spoken to one of these companies which is launching a pilot program focused on delivering packages of seeds/fertilizers to rural Kenyan smallholder farmers affordably. We believe that our work and local connections can help them achieve their mission more effectively, as some of our nutrient recipes do make use of their product (just in much smaller quantities and with greater balance). As we expand, these incumbents can provide us with an entry point to new countries/markets and helping us scale. Therefore, we see these incumbents as potential strategic partners that can enhance our impact. 

Where will these actions be taken?

Fertilizer is commonly used worldwide in different types of farming, and this project has high applicability in all regions in which soil degradation and acidification have affected local crop productivity and food security. Initially, we will specifically implement our project Kenya because of our active network of partners in Kenya. In a country such as Kenya, where we begin our work, there are 4 million farmers, spending $76 million/year on ineffective fertilizers that risk degrading their soil due to misuse. Our solution will initially target rural farming villages of fewer than 5,000 people. These villages have traditionally difficult road access, making any imported fertilizers expensive. We believe that such scenarios make our process and product the most competitive. Our initial pilot MiniPlant production has been set up in Mwea, Kenya, and currently we serve the local farmers who mostly are into rice-farming or horticulture. We already have plans in expanding the MiniPlant model to other parts of Kenya and identifying the local partners to facilitate this expansion. If successful, we will gradually expand to other regions around Kenya, and then the neighboring countries in East Africa (such as Uganda and Tanzania) facing similar problems. We also have intense partnership interest in India, in which we expect to set up a pilot soon.

In addition, specify the country or countries where these actions will be taken.


Country 2


Country 3

United States

Country 4


Country 5



What impact will these actions have on greenhouse gas emissions and/or adapting to climate change?

Firstly, our process converts crop residues into a carbon-rich fertilizer mix that stays inert in the soil for hundreds of years. Normally, when biomass dies, the carbon is released into the atmosphere again, in the form of CO2 within a matter of days if the biomass residue is burned, and/or in the form of CH4 (methane) within a matter of weeks if such residue is left to decompose naturally. When used as fertilizer, our process removes this carbon from the biomass into the soil semi-permanently (in the timescale of centuries). Therefore, our product mitigates greenhouse CO2 from the atmosphere at about 1.7 tons of CO2 equivalent per ton of product applied.


In addition, improving yields and crop performance also improves the health of the soil, and slows the rate from which carbon is released from the soil back into the atmosphere, as various organic processes in the soil hold onto the carbon atoms. This has been demonstrated in recent studies such as Song et al. (2016) focusing on the impact of soil carbon content on the carbon flux into/out of the soil. Much of this impact is crop- and soil-dependent, making this effect hard to estimate and generalize without knowing the specifics of the crop type and soil type in which the implementation is taking place. However, as a general rule of thumb, the mitigation is around 1.0 ton of CO2 equivalent per ton of product applied.


Finally, our process also avoids the burning and disposal of agricultural residues by making use of them and turning them into the carbon-rich basis for our fertilizer mix. Burning of biomass, while as mentioned above can release CO2 back into the soil, has some additional climate forcing effects beyond the CO2 mechanism. According to a recent Stanford study, the secondary climate forcing effects of open-field biomass burning such as soot, brown clouds, and heat fluxes is responsible for up to 18% of anthropogenic global warming. By putting crop residues to more profitable use, we can avoid large-scale, open-field biomass burning and many of the associated secondary climate forcing effects.


Ultimately, we believe that in 5 years, as we scale to 100 MiniPlants producing 90,000 tons/year of the fertilizer, we can mitigate around 170,000 tons/year of CO2 equivalent. When scaled worldwide to its full potential, our solution can mitigate around 500 million tons/year of CO2 equivalent. This is equivalent to the current annual total CO2 emissions from a mid-sized industrialized country.

What are other key benefits?

This project is expected to improve farmers’ yields and income, create additional rural livelihood through decentralized fertilizer production, and reduce waste and pollution.


Farmers are diverse, but take the example of Mr. Kibuchi, a rice farmer. He nets ~$200/year from a 1.5-acre rice field. Since 2017, he has been testing our fertilizer instead, and noticed that for the same price that he pays, he is getting around 25% more in harvest yield. Part of this increased yield he keeps to the family table, while part of this excess harvest he sells to the nearby rice mill. As a result, at the same input costs, his new net is now around $300/year, which is a 50% increase, as shown below.


Old gross income: $800/year

Old costs: $600/year

Old net income: $200/year

New gross income: $900/year

New costs: $600/year

New net income: $300/year


In 10 years, we can impact 120 million new male/female farmers like Mr. Kibuchi of different ages. The benefit is likely greatest for smallholder farmers with low to medium income levels.


Furthermore, our biomass processing systems can support new profitable village-based fertilizer production in rural communities. Another group of beneficiaries are our partnering local agricultural organizations and entrepreneurs who operate and profit from our systems, and the additional rural youths they employ to support the localized production. We estimate that each Takachar system creates $60,000/year of new job/income opportunities per village. Take the example of Mr. Japheth. He graduated from high school, and had difficulty finding a job in his ancestral village. He was about to move to a slum in Nairobi to find an urban job. Then he heard about our pilot localized operation, and became a production associate. He got to stay in his home village, and three years later, became promoted to production foreman. Now he is contemplating an MBA degree to further his career.


Finally, by providing a profitable output for crop residues, we also reduce widespread residue burning and health effects/deaths of smog upon 200 million people who reside within 100 km of these areas or in nearby cities. This is compatible with many regional governments concerned about the effect that open-air crop residue burning is having on the local air pollution and seeking to end this practice. As we scale up we will certain leverage this policy-based incentive to help us achieve greater impact.


What are the proposal’s projected costs?

As mentioned above, to initiate a small-scale, village-based fertilizer production requires only an upfront capital investment of $20, using one of our open-sourced, batch conversion process for small-scale experimentation with farmers. From our prior experience, running such a pilot with an initial group of 20 farmers over 2 growing seasons require a budget of around $2,000. Eventually, as the village-based fertilizer MiniPlant scales to full capacity, it will gradually require more and more equipment investment, resulting in another capital investment of $15,000 over the course of two years. Much of this investment is gradual, and can be financed from the existing production margin.


One potential threat is that, by localizing fertilizer production, we may reduce the demand for conventional fertilizers from large-scale companies, who may not be entirely happy about this displacement. However, we wish to point out that this will not really be the case. From our interviews with incumbent, large-scale fertilizer producers, many of them have historically had trouble getting their products to the rural farmers due to the aforementioned logistical challenges. As a result, they are actually very interested in accessing smallholder farmers in emerging markets, and our work and local connections can help them achieve this more effectively, as some of our nutrient recipes do make use of their product (just in much smaller quantities and with greater balance). As we expand, these incumbents can provide us with an entry point to new countries/markets and helping us scale. Therefore, we do not necessarily see these incumbents as competitors to fight against, but rather potential strategic partners that can enhance the number of smallholder farmers in unfamiliar markets that we can reach. 


Finally, farmers as well as prospective implementation partners can be risk averse. They must see proof before they change products. Through our initial pilot, we learned quite a few things. For example, in a new community, we found it most effective to first engage an older, more affluent farmer who often can afford new products and some risk-taking. This farmer is also someone that other younger farmers aspire to be. Once the product is demonstrated to produce benefits with the initial farmer’s test plot, he/she often becomes the influential champion who will multiply our customers rapidly. As we scale beyond this initial community, we may encounter challenges reaching out to other geographically disparate communities. We foresee the need to conduct additional localized pilots. Fortunately, our equipment is relatively low-cost and can be operational within a few weeks.  In the medium term, working with an existing equipment manufacturer with long-standing credibility in many communities is essential to our scaling.


Currently, with our pilot production, we are already impacting 3,000 local farmers. “One of the benefits of using Safi Sarvi is that I always have enough food in the house for my family,” said Mr. Mudhike, one of the rice farmers in Mwea County who participated in our intervention. “I sell the excess rice and use the money to educate my children and take care of my other household needs.” We are sequestering around 500 tons/year of CO2 equivalent.


By 2023, we expect to deploy around 500 systems in the field. This will involve working with 50 local agricultural organizations like Safi Organics and the KALRO network, which will operate our systems profitably and sell fertilizer to local farmers. Letters of intent from Safi Organics and at least 5 similar organizations are available upon request. This will enable us to impact 300,000 smallholder farmers with degraded soil. This assumes each system, at full operation capacity, produces around 300 tons/year of fertilizer blend that can serve the needs of a village and its 15 km radius, or around 500-1000 farmers.


Within 15 years, we expect to deploy around 5,000 systems in the field. This will enable us to impact 30 million smallholder farmers with degraded soil. By producing around 4.5 million tons/year of Safi Sarvi fertilizer, our intervention can mitigate around 11 million tons/year of CO2 equivalent.


Within 50 years, we expect our growth to continue, as in many places around the world, millions of acres of farmlands that used to sustain agriculture are slowly being degraded over time, in part due to climate change, such that in the future, they may no longer support the same agricultural activities as today. Our solution can also help restore the soil health and food security in these regions. For example, according to an India-wide study by Kumar et al. (2014), the soil degradation problem will affect another 50 million hectares (30%). Therefore, at full scale, we expect our solution to impact 700 million farmers worldwide, mitigating around 500 million tons/year of CO2 equivalent.

About the author(s)

Samuel Rigu is the full-time Chief Executive Officer of Safi Organics. Currently he is based in Kenya, managing a team of eight full-time and four part-time employees on the pilot project and expanding the customer base. Samuel co-founded (with Kevin) a prior company in Kenya that sold more than one million mosquito coils to a large distributor. Samuel grew up in rural Kenya, where he witnessed first-hand the challenges of rural farmers such as his grandmother and neighbors for accessing affordable, high-yield fertilizer, and resolved to solve this problem. He studied agribusiness at the University of Nairobi, and worked at the Turning Point Trust Farm, where he implemented organic practices and turned the farm’s finances from loss to profit in 6 months.


Kevin Kung worked on the company’s core technology as his MIT PhD thesis in the field of biofuels and renewable energy, from which he graduated in June 2017. As a current Cyclotron Road entrepreneurial fellow, he is funded by the Department of Energy to continue the company’s ongoing R&D collaboration at the Lawrence Berkeley National Laboratory. Kevin has 10 years of experience with engineering design for resource constrained settings, including prior work on a low-cost hematocrit centrifuge design in Nigerian clinics, on water purification systems in Uganda, and on off-grid rechargeable solar lantern systems in Peru. He also has entrepreneurial experience in emerging markets, including selling over one million low-toxin mosquito coils in Kenya.


The following people are also on the executive team:

Joyce Kamande is the Marketing and Sales Director. She oversees the marketing team, makes regular visits to local farmers and partners, and builds the Safi Sarvi brand.


Justin Munene is the Accounting Manager. He oversees the financial aspect of the business and coordinates with suppliers and customers.


We have eight full-time and four part-time employees in Kenya to manage the pilot production.

Related Proposals

An earlier version of the technology was applied in a different field of clean cooking fuel provision, as detailed in the following proposal:

Takachar: turning organic waste into safe and affordable charcoal brique...


The underlying scientific principles of our technology have been demonstrated and published in company co-founder Dr Kevin Kung’s PhD thesis at MIT, initially with a focus on energy (solid fuel) applications rather than fertilizer: K.S. Kung, Design and validation of a decentralized biomass torrefaction system, PhD thesis, MIT, 2017. <>.


A featured MIT article about our work is here:


The website of our partner company, Safi Organics, is here:


Seven peer-reviewed articles based on the fundamental scientific work at MIT on the underlying technology have been published, and at least four more articles are under review or being planned in other journals. The seven published articles are available at the following DOI links: 10.1016/j.biortech.2012.07.018, 10.1016/j.biortech.2013.01.158, 10.1016/j.fuel.2014.07.047, 10.1007/978-3-319-20209-9_8, 10.1016/j.biombioe.2018.11.004, 10.1016/j.biombioe.2018.12.001, and 10.1016/


The following references have been cited in our proposal:


Abewal, A., Yitaferu, B., Selassie, Y.G., and Amare, T. The role of biochar on acid soil reclamation and yield of Teff (Eragrostic tef [Zucc] Trotter) in Northwestern Ethiopia. Journal of Agricultural Science 6(1): 2014, 1-12.


Kumar, R., Chatterjee, D., Kumawat, N., Pandey, A., Roy, A., and Kumar, M. Productivity, quality and soil health as influenced by lime in ricebean cultivars in foothills of northeastern India. The Crop Journal 2(5): 2014, 338-344.


Ulyett, J., Sakrabani, R., Kibblewhite, M., and Hann, M. Impact of biochar addition on water retention, nitrification and carbon dioxide evolution from two sandy loan soils. European Journal of Soil Science 65(1): 2014, 96-104.