Distributed Farming with unsteady Renewable Energy

First of all, thank you for welcoming me to this community. I just started learning about this project, so please be understanding if I there are some things I do not yet fully comprehend.

I come from a computer science background and I work in the renewable energy sector for a company that owns and operates hundreds of wind turbines as well as some solar plants spread across Germany and other parts of Europe. We are interested in potentially contributing to this project by operating 3Nodes as Farmers, that will be powered exclusively by renewable energy.

Understanding the problem
As most of you are probably aware, renewable energy, aside from its many advantages, has one major disadvantage compared to carbon based forms of energy. This disadvantage lays in its limited availability. Even though there is potentially an endless amount of wind and solar power to be harvested, this power is not available at all times. Depending on the region, our wind turbines are able to produce electricity between 70% to 90% percent of the year. When combining wind power with solar power, the availability can be further increased, as a lot of times during the daytime the sun is shining when wind is scarce.
Still, reaching even close to 100% availability in an economical and sustainable way is not achievable.

Understanding the Opportunities
While renewable energy is not available at all times, there are other times where renewable energy production exceeds the demand. This circumstance is further intensified by electricity grids often being unable to carry these peak energy production volumes all the way across the country to areas with demand for electricity. When the production of electricity exceeds the demand, the price per kWh sold at the energy exchange can drop significantly and even become negative. From this arrises the opportunity to make use of this overproduction of energy in an economical way.

Forming a solution
There are quite a few ways to make use of excess electricity. It can be stored in batteries, converted to heat (warm water) or to gas (hydrogen). Another possible solution is to ramp up energy intensive operations when supply of energy is high (and prices low) and to restrict those tasks to a minimum when supply is low (and prices high). Energy intensive IT-Applications can be a suitable candidate for this.
The problem of course here is, that we as a society are used to all of our IT-Services being constantly available to us, no matter when and where. In the last one and a half years we have started to use some of our cheap green energy to power ASICs to mine some Bitcoin on a small scale. The advantage with this is, that our crypto miners can be turned on or off at any time, depending on energy availability and prices. I do not want to discuss the pros and cons of proof of work mining in this context, but we are eager to find alternative IT-Applications that would enable us to make use of our unsteady renewable energy availability. There are some fields of use that I could think of that do not necessarily need constant availability, such as machine learning or 3D-Rendering. But setting up the entire infrastructure and acquiring paying customers for our services will be completely new challenge by itself. That’s where I see an opportunity for us to join the ThreeFold Grid as Farmers.

Sustainable Computing
From our perspective the future of computing will be 100% sustainable (it has to be). The German government even set itself the goal that all new data centers from 2027 on shall be carbon neutral. In order to achieve this, energy must be consumed right where it is produced and when it is available. Technologies and concepts like ThreeFold could be a key component in this transition.
As a producer of green wind and solar power we want to be part of the solution. Obviously we have to be able to turn a profit with this in the long term, but there is also the research component in this all that interests us.

Distributed Farming with unsteady Renewable Energy
From my current understanding Famers on the ThreeFold grid are only getting paid if their 3Nodes are available at least 97% of the time. Since this is not achievable using renewable energy alone, I would like to start a discussion here to determine whether or not there are or will be opportunities to contribute to the ThreeFold grid as Farmers with a lower availability.
From my humble viewpoint, a highly distributed and redundant system such as the ThreeFold grid should be a perfect candidate to support the needs and requirements of carbon neutral IT-Applications by enabling them to stay up and running, even when some parts and subsystems of the entire grid are unavailable due to a lack of sustainable energy.
Maybe Farmers could be enabled to clone their 3Nodes and set them up in multiple different locations to increase availability and redundancy on a node by node basis (multiple nodes are treated as a single node by the grid). Of course profitability would be greatly limited by this, if the reward payout scheme is not adapted accordingly. In general, from what I understand, the Grid already takes care of the required redundancy itself so it should not matter too much if one of our nodes is unavailable from time to time.
Nevertheless, when uptime requirements for all nodes decrease, so could the quality of service of the entire grid. I don’t know what approach would be best suitable, but if there is any potential solution for sustainable energy companies like us to contribute to this project, I would be glad to hear about it.

Any input, ideas or remarks are on this are appreciated. Thank you for joining me in this discussion.



Hi Jakub, and welcome to the community :wave: I’ve been following your posts in the main Telegram chat and really appreciate the perspective you’re bringing. Indeed, I think it’s important to discuss how the Grid could become more compatible with the reality of renewable energy sources like wind and solar.

First let me clarify that the Grid is really just a raw capacity provider. It is still the responsibility of anyone deploying workloads on the Grid to provide their own solutions for redundancy. We do develop a distributed storage system called Quantum Safe Storage in house and the Grid has native support for the most popular tooling for making compute workloads redundant, Kubernetes. Fully self healing architectures have always been a part of the vision, but there’s still a lot of work to be done before this is the general case reality.

That said, I think with some creativity, capacity available on a limited basis has many use cases. A Kubernetes cluster, for example, really just needs one node with high availability. As long as there’s sufficient overlap in the availability of other nodes, it can reprovision the services it’s responsible for to maintain a given state. How that would work in practice given the availability cycles of wind and solar is beyond my scope.

We’ve always planned to include special accommodations in the model for farmers in regions where power and data infrastructure simply can’t meet high uptime requirements (currently 95% for DIY farmers). I think it’s reasonable that we could consider some similar accommodation for farmers using exclusively renewables. This could even come along with a different kind of certification status, which would indicate both the renewable sourcing and the fact that uptime would be limited.

As far as considering different pieces of hardware as duplicates of the same node, this is probably not practical. The individual hardware nodes would need to share state in a way that would be complex to maintain and would depend on enough simultaneous uptime to handoff that state to a different box. Rather than try to bake this in at the Grid level, I think it’s better handled by higher layers in the stack.

Thanks for your interest in the project and for sharing your thoughts. I think there’s a great opportunity here and hope we can find a fair way to get you farming :slight_smile:


Hi Jakub,

Thank you for sharing. What a great vision and yes - this would help the world of IT / the internet a lot. I have spent the last two decades building traditional (managed services) datacenters. This is where I have seen the following thing happen over and over and over: for a successful online service, the difference in usage (bandwidth and server load) can be 100x (maybe even more these days). Most online services have close to 100% of that peak capacity on all the time and take the high operational costs to not have to scale things up and down too much (with change comes risk).

Anyway - long story cut short - Content Delivery Networks (CDN’s) are helping to manage peak loads and make the core site scale down while the brunt of the web (service) delivery is being done through the CDN.

I can think of numerous things that we can build together where the “compute” workload follows the sustainable energy surplus and gets done where there is electricity available. The smart contract for IT can be used to scale things up (and down) but also move workloads around. I am very excited to see how we can make this work and create truly sustainable compute (and storage I am sure, archiving being one thought but this has to be looked at in a little bit more detail) and in the long run also commercially sustainable.

Looking forward to engaging, I believe we can find a way to reward farms that are not on 100% of the time, based on geography (nonreliable power) or green energy.


Hi Scott, thank you for your input on this so far!

Could you elaborate on how capacity providers would be able to provide their own solutions for redundancy? From my current understanding Farmers have very little control over their 3Nodes running Zero OS. There may be some key info I am missing here. What else can Farmers do other than setting up a decent environment and infrastructure for their machines, installing Zero OS and connecting everything to the Grid? Could we setup and manage our own Kubernetes cluster and integrate it into the Grid somehow?

The availability cycles of solar and especially wind can be predicted up to a certain degree with pretty good accuracy. It’s not like power will suddenly drop out of the blue. Our systems already offer some degree of prediction and can make this data available to external applications. So I am sure there will be ways to handle outages in an elegant way.

An adapted accommodation model for farmers exclusively using renewables would be interesting. I would have to run the numbers to be completely sure, but by combining solar and wind power we could reach an availability of 90% or more, depending on the exact location and total power consumption of the farm.

A different kind of certification for this could be interesting. I am sure some companies or individuals looking for available capacity on the grid would appreciate the possibility of purchasing storage or computing that is powered by renewables only. Of course the certification process must be solid so that I won’t be exploited by farmers trying to cheat the system.

Hi weynandkuijpers, thanks to you as well for your thoughts on this.

I can see why many companies would not like to take any risks on this and have all their needed capacity available to them at all times, while being in full control. In order to have these companies consider different solutions they would have to be 100% solid, scalable and secure. Evaluating how good of a job the ThreeFold grid can or will do in this regard is currently beyond my abilities. CDNs already tackle part of the problem, thanks for pointing this out. How exactly would the Threefold Grid be superior in this regard?

Yes I totally agree that there should be plenty of computing workloads that can be processed location independent, and some even independent of time (meaning it does not matter if the task is finished today or tomorrow). Though I may have to do some further research on how smart contracts can help with accomplishing this as I have no experience in this regard yet.

CDN was only an example that dealt with scaling issues, but beyond that there is a world of opportunities. One of the things that we have been brainstorming in the not do distant past is to have smart PDU’s that can turn devices on and off based on workload requirements. This can be done for normal grid powered equipment but would be even nicer for wind and solar powered devices.

I’d like to brainstorm this a little further if okay for you and see what we can come up with. Our dream is to be able to have compute and storage generators everywhere and start building new ways and locations to create compute and storage capacity. The model as it stands today is not sustainable and not scalable. This hardware still needs to sit (or be attached) to some form of “real estate” and were entering in very existing discussions with property builders and al sorts of other real world infrastructure builders (roads, bridges, towers, etc, etc).

Love to set the first to look into how we could build a model around “using surplus energy” and fueling temp compute and storage capacity. Online vid meeting?

1 Like

Yes smart PDUs certainly could make setting up these systems easier. In a way we’re already using (semi)-smart PDUs for our mining operations. At least for monitoring power consumption and turning devices on/off. Anyway, dynamically balancing workloads according to renewable energy availability on the grid can also be implemented on the software side, so I don’t see this as a top priority atm.

You say the model as it stands today is not sustainable and not scalable. From the vision that Threefold has set out I believe that sustainability and scalability have been primary goals from the very beginning. I’d be glad to give you time to brainstorm on this. And yes, we can also have an online video meeting to better communicate our different viewpoints and talk about this.

Thank you Jakub. In order not to loose the contact and information in this group, shall we do and record the online meeting. I’d love to see what we can do to bridge and make these two worlds met.

Does this week / next week work for you?

Next week should work. Let’s discuss the details via PM to not clutter the forum space.

I’m a professional environmental engineer if I can ever be of help to anyone on this. Only licensed in FL though.


To have continuation here, Jakub and I am meeting in a zoon next week and with Jakub;s permission we will record the talk and publish for the forum to keep up. @FLnelson that sounds very interesting, let’s kick this one off and then continue here. :pray:

Sure we can record the zoom meeting. Not sure how structured this talk will be and whether it will be worth watching for others but let’s see.

@FLnelson your input could be valuable. As a software developer my environmental engineering skills are rather limited. Environmental engineering is a versatile one. Can you tell a little more about your skillset?

Did you guys manage to get it recorded?

I think I forgot… But the meeting was really good. Give us some time to cook the broth and we will come back… :slight_smile: with some news and updates.


really looking forward to news and updates on this topic. This is such a great opportunity. I would love to participate in this project. I have studied industrial engineering and am currently working for a real estate developer and also run a small company providing IT-system-administration-services. We have just build up a pretty solid farm in Germany. Let me know if i could join you and help in any way!


Hi Dany,

Glad to hear you are interested in this topic. There is still a lot of brainstorming to be done, but I believe we are heading into the right direction. The discussion has somewhat shifted to private chats and meetings but when there is anything substantial I will do my best to post an update here.

How large ist your farm and may I ask where you are located at?
If possible at all I would love to get the opportunity to check out your farm sometime.

Greetings from Berlin


Hey Jakub,

I live in Mainz, real estate development office is located in Wiesbaden, projects in Frankfurt, Wiesbaden and all around the Rhein-Main area.

Our farm is located in Wiesbaden at the office of one of my sysadmin clients (dedicated power line and internet connection). Currently we are running 11 3nodes (9 servers, 2 PCs) providing 176 cores / 352 vCores, 2816GB RAM and 50TB SSD storage to the ThreeFold grid.

We are looking for suitable locations with fiber optic connection and sufficient energy supply to build up additional farms.

You are very welcome to check out our farm. I’m glad to share/exchange experience with you. Feel free to get in touch!


Wow! I truly believe that we should spearhead the grid expansion through datacenterless (no pun intended to data based farms) deployments: real estate projects, wind and solar farms, universities etc. All of the mentioned “locations” are in flight with pilots and projects.

So proud :upside_down_face:


Hey @weynandkuijpers ,

@jakubprogramming and I had an online meeting about this project. We are trying to figure out how we can come together to build up a farm at one of the wind-energy-plant sites as a pilot projekt / feasibility study. It’s not quite easy because there are so many things to be taken into account but there is also so much potential in this… it has to be adressed.

Maybe you can do me a favor. As mentioned here (ThreeFold | Hewlett Packard Enterprise) HPE is partner of ThreeFold and “wants to help ThreeFold to offer hardware and IT services to the community.” I would like to get in touch with HPE in relation to this partner programm in order to check if they could provide us with hardware for this particular project.

1.) Is it ok for ThreeFold if I would contact HPE in regard of this partnership?
2.) Can you provide me with any contacts at HPE that I could talk to?

That would be great. Thanks!


Salam @Dany. Yes we can and yes I will. Let’s connect privately to send that information. @weynandkuijpers on Telegram.

1 Like