What if the heat that cryptocurrency mining generates was used for growing vegetables? Imagine being able to grow food in cold climates! The possibilities are endless. A couple of renowned Swedish institutes and a cryptocurrency mining company joined forces to bring mining and farming together. This Greenhouse - Data center project means there is potential to make nordic communities self-sufficient. How amazing is that?
Warmth - the by-product of not-just-bitcoin-mining but any type of data center, is scarce in Northern countries. The missing component. And this precious resource has gone to waste. Well, not anymore!
Bitcoin mining is a growing industry in the Nordics, where cheap, green energy is widely accessible. While computers are using up this energy to mine bitcoin, something else is produced on the side. Something that has been known as just an externality: heat. The main by-product of bitcoin mining is the tremendous amount of heat the computers generate. That is, until now. Soon, this heat won’t just be unintended waste. It will be a valuable resource.
The partners who collaborated on this initiative got together to share what you should know about The Greenhouse Project. They talk about the initial goals, give some background information, and provide a glimpse into the technical details too.
Just a 1 MW facility in a sub-arctic climate could heat up to a 3 000 m2 greenhouse, or partially heat a 10 000 m2 greenhouse which would increase the local self-sufficiency with 8%.
The initial visualisations for the planning phase
Assembly of one of the containers, followed by its installation in Sweden
I already built a lot of “conventional” crypto data centers in many countries for GM. Every single one was unique and a challenge. But in the end they all were kind of the same. This project with the interconnection and a greenhouse involved was something completely new and exciting. And I always like to come up with new ideas.
I often thought about how to use the excess heat of our datacenter. From heating homes to drying fruits I thought about many use cases. But none of them really worked. To finally have a use for all this energy is really great and excites me probably the most about this project. In 2019 we were contacted by our Swedish partners about the idea. And we were really glad to start working on a solution together with them!
As we have never done something like this before and the entire project is a proof of concept everything needs to be customized and fitted for this purpose. There is no off the shelf solution. In most cases you use water or oil to transport heat, because a cubic meter of air can transport much less energy than those fluids. But these solutions are expensive and not suitable for large scale operations. Therefore we decided to use water as a medium to transport the heat. Then again we had to go with a very high mass flow compared to fluids. This caused a lot of headache for us but we are glad we could solve this problem.
I hope this project will be used as a display to show that greenhouse data centers are working. From my point of view this is only the first step into the right direction. In the future I hope it will become a norm to use the massive amount of excess energy form data centers for something like greenhouses.
For the near future I hope to expand the area of the greenhouse further and utilize more of the hot air. All the heat will come from the 600 kW container you see in the pictures above ( at least in the first stage for the 300m2 greenhouse). If expanded to 5.000 m² we will need 1 MW, so another 400 kW data center container will be added.
I started my work as a data center researcher with focus on the energy perspective and a data center´s role in the energy system. I have carried with me a vision of using the data center excess heat to farm vegetables for elementary school meals. By working with grassroot policy and making an awareness from childhood where the vegetable comes from, we can encourage local and ecological shopping.
I was involved in planning and writing the project proposal DC-farming. This became a Vinnova project focusing on finding the possibilities and challenges in datacenter greenhouse farming. I have had the responsibility for this as a project coordinator, and that’s how I became a part of this current collaboration.
I have had a continuous dialog with Boden Business Park and Genesis Mining together with my colleagues at Luleå University of Technology, contributing with our knowledge and being responsible for the research perspective of the collaboration.
The real challenge is the coordination and wideness of the project covering both soft to hard aspects. Soft aspects involving the human perspective as farmer and data center owner, and hard aspects like data center and green house operation, and make this all work together.
What also excites me is the preliminary results we generated showing that there is a great potential of using the data center excess heat for greenhouse farming. A 1 MW data center would have the ability to strengthen the local self-sufficiency up to 8% with products that are competitive on the market.
Furthermore, when putting this in a broader perspective and having data center as driver in urban and industrial symbiosis it becomes exciting! For example, using CO2 from the local biogas plant for farming boosting or aquaponic farming where fishes and vegetables are grown in harmony.
Something we are currently working on is farming of mealworms based on data center excess heat. Mealworms will generate protein for poultry farming and strong fertilizer and can be sold on the market. By feeding the hens with living mealworms the import of protein from China can be eliminated and at the same time farming becomes ecological. The hens will strengthen the local self-sufficiency in eggs or chicken meat, the fertilizer is suitable to use as nutrition of greenhouse farming which also is heated with data center excess heat.
Our data center research focusing on cooling has showed that by using direct fresh air for cooling can generate up to 40°C excess heat for a traditional data center. In case of a High Performance Data center, which is more power dense, the excess heat temperature can reach up to 55°C. By going for liquid cooling, as immersion or on-chip cooling, excess heat liquid temperatures up to 70°C has been achieved and for stable conditions about 50°C, this is ongoing research with the objective to reach about 80°C in stable conditions.
We have shown that by using a traditional data center with direct fresh air for cooling different applications can be attached to e.g. fruit drying, wood chip drying, greenhouse heating, mealworm farming. By using liquid cooling other applications will be possible e.g. district heating, fish or algae farming, penciling production, furthermore liquid is a better energy carrier and easier to transport compared to air.
By combining all these alternatives, we have mapped the different options where data centers can act as driver for Industrial and Urban Symbiosis, in the areas of: Heat & Power, Food & Health Care, Marine and Recreation.
I see students as an important and great resource in all of my projects. Often they are very creative and give a new perspective. Also by working with students you spread the awareness and knowledge about the ongoing research.
A group of students from building construction and architecture has developed a modular data center greenhouse in conjunction with the planned exploration of Genesis Mining, which offers a space for meetings and offices next to the farming area. The model will be used in the coming workshops as an example of how a data center greenhouse could look, by having examples to show will encourage a dialog and discussion.
Currently there is an ongoing work with three students from the program of sustainable energy technology where they are using Computational Fluid Dynamics for analyzing temperature and flow distribution in the data center greenhouse that is planned in this project. The study will show how the air generated from the data center should be distributed in the greenhouse to achieve a good plant and working environment. The result will be used for planning of the air handling, both for evaluation of suggested layout and recommendations for changes or new layouts.
Once the data center and the greenhouse is installed and built, we first need to develop a control system that optimizes the use of available excess heat in aspects of both temperature and humidity within the greenhouse. This we see as really challenging, but the goal is to develop algorithms that are scalable and adaptive to other data center greenhouses both regarding sizes and outdoor climates.
A medium term goal is making a showcase to demonstrate the possibility of using data center excess heat for competitive farming in a sub-arctic climate, which have not been possible before. We need a knowledge platform too that can be used as support and help for replications at other locations.
At a later stage I would like to see further development of the urban and industrial symbiosis area where data centers act as driver by attaching more applications, being the pioneer in showing possibilities and benefits generated by this symbiosis both from the humanitarian and technology perspective.
Boden Business Agency is the initiator of the project, and the main coordinator for solving the heat transfer solution between the Datacenter and the Greenhouse. This also includes seeking the funding/financing of the project.
There is a great deal of national interest in the Energy Symbiosis, and from the Swedish Energy Agency. Moreover the national government has publicly expressed that we need to be more self-sufficient in food production rather than rely on imports. So what excites me with this project is that there is an opportunity to contribute in scaling up the food industry and at the same time meet the national energy efficiency targets. This project is the first of many projects within the Energy Symbiosis, and hopefully the results will show that it is very possible to scale up into larger commercial production.
Boden Business Agency have signed an agreement with the Swedish Energy Agency to become a strategic node for a resource efficient circular energy system between the IT sector and the food cluster. The purpose with this engagement is to create synergies between existing energy intensive industries such as a datacenter and to connect them with the food cluster, creating new material and circular energy flows, and to develop the conditions for further establishments of different types of demo- and pilot plants. We call it energy-driven business development. The Swedish Energy Agency’s goal is to replicate the innovations we develop here to other locations with similar infrastructure and businesses. Therefore this project fits very well into the larger energy system.
One of the main operational costs, at least in this climate, is the energy cost. So rather than use new energy generated from electricity, there is an opportunity to reuse waste energy from another process, and thereby reducing cost. For this demo project, the aim of the greenhouse owner is not to operate it as a large scale commercial business, but to 1) To operate the greenhouse as a social project to educate farming to people who are excluded from the labour market, and 2) to test what crops work best to grow in this climate. The lessons learnt from this project will serve as input to a large scale production facility.
The main purpose of this project to start with, is to understand and optimize the climatic conditions and airflow dynamics between the datacenter and the greenhouse. Further on into the project we would like to add CO2 captured from the biogas production process at the nearby biogas plant to boost the production capacity. The biomass waste from the greenhouse farming will be used as residual waste for the biogas production, creating a true circular economy within the energy system.
The idea of using a resource that would have otherwise gone to waste is very appealing to me and in particular also that this resource can be used to secure the food supply in northern Sweden. In this region, the subarctic climate is making it difficult for farmers to compete and the food production has decreased in the past decades. From a professional point of view, the complex dynamics that arise from combining the behavior of living plants with a climate control system is very interesting. One example is the nonlinear behavior due to evaporation of water from the plants depending on the air humidity. Another issue is the complexity of the control system itself, since we need a combination of several fans and dampers to accomplish the right conditions for both the plants and the DC. As a researcher in the field of automatic control, it is an interesting challenge to work with the control algorithms for these complex dynamics.
I have been making the calculations for designing the system for air flow from DC to GH. In practice, this means calculating what air flows of DC air and fresh outside air that are required to maintain a productive climate for the plants, in terms of temperature and moist, given variations in weather conditions. Also, I am involved in the development of control algorithms for achieving these air flows.
A major challenge is the varying conditions under which the system has to operate, from dark winter nights with -30 ℃ outside to warm summer days with sunlight almost around the clock. The moist produced by the plants will also vary greatly, depending on the sunlight and the moisture of the air. Under these varying conditions, the air flow system must then operate in completely different ways, sometimes using only DC air and sometimes only outside air and sometimes combining them in a specific way. Moreover, this must be done in a way that does not affect the DC, in the sense that the DC fans do not experience opposing pressure that makes it hard to push the necessary flow through the DC. In order to achieve this, a complex system connecting the DC and GH involving both fans and dampers needs to be used and the control of this system becomes a multivariable problem.
For the cold climate in the north of Sweden, our calculations show that a 300 m^2 greenhouse can easily be heated with a 550 kW container, even with outdoor temperatures reaching almost -30 ℃. But the potential is much bigger than that. The temperature difference over the greenhouse is in this calculation only 10 degrees since we assume a DC output temp of 35 ℃ and a GH temp of 25 ℃. If the DC output is increased to 55 ℃, then the GH area can be tripled to 900 m^2. Another possibility is to return some of the GH air into the DC which will also have the additional benefit of better moist level in the DC. Then our calculations show that we can increase the GH area to almost 2000 m^2 without too much moisture in the DC. The potential is of course even bigger in warmer climates but keep in mind that some of the heating is necessary to get rid of the moist produced by the plants and not only for overcoming low outdoor temperatures.
None at this point.
Once we have an installation up and running, we can start experimenting and evaluating the ideas. It will take some effort to make the control algorithms for all the flows, temperatures and pressures functional. As a next step, it would be interesting to build a setup with return of greenhouse air to the DC. The challenge is then to develop a control strategy that simultaneously optimizes the climate in the greenhouse and in the DC.
Contact us at firstname.lastname@example.org