Could CEA-Heavy “Farm Parks” Make Server Farms Better?

The Resource Innovation Institute explored the possibility of colocating data centers with large greenhouses to deal with waste heat, supply food and provide jobs in the U.S.

Turning-Server-Farms-into-Farm-Parks.jpg
(Lori Hays)

As artificial intelligence grows, data centers are popping up like weeds, and in many rural communities they are about as welcome.

But one group is looking into the possibility of “farm parks,” data centers colocated with greenhouses, cold chain logistics and other needful infrastructure, that could bring local food and jobs to areas where data centers are growing, as well as a much-needed PR boost for data centers.

During a Jan. 14 webinar by Resource Innovation Institute and Agritecture, study authors Derek Smith and Rob Eddy presented the findings of a feasibility study on colocating greenhouses with data centers in Virginia. The main idea is that the greenhouses can take advantage of the waste heat generated by the data centers. But the study concludes the model should be more.

“While a simple connection of a data center to an individual greenhouse via a heat exchange-only model is possible, it is economically limited,” the report reads.

“However, colocating data centers with multiple, large-scale greenhouses and other complementary industrial businesses would provide a number of economic and quality of life benefits for Virginia communities,” it reports. This is what Smith and Eddy called “farm parks.”

What Goes into a Farm Park?

Smith, RII’s executive director, says there are five ingredients that are needed to make farm parks work, based on the feasibility study:

  • “CEA-ready” data centers
  • Large, high-tech CEAs
  • A central resource hub
  • Cold chain and logistics
  • Other target sectors as needed

Colocation of data centers and greenhouses of course requires those two main elements. But because data centers are being built rapidly and CEAs generally are not, the farm park model calls for the building of “CEA-ready” data centers.

A CEA-ready data center is one with “stubbed off pipes that can connect to a later set of infrastructure that can then add greenhouses to it,” Smith explains. He also says the eventual greenhouses would also need to be big — really big.

“We’re talking 1,000-acre development and 200, 300, 400 acres under glass — something we have not seen here in the U.S.,” Smith says. “We’re going to have the scale opportunity to actually bring in a lot more than just lettuce. We can have a basket of goods, and we can fill trucks coming out of this business park to serve regional needs at a scale that we’ve never seen with CEA.”

At the heart of making the farm park model work is the central resources hub. Its importance is due to the mismatch of heat and energy between data centers and greenhouses, according to Eddy, RII’s technical director. The heat difference is the first and most immediate.

“A standard greenhouse needs 170-degree heating water,” Eddy says. While older data centers are often air-cooled, newer data centers, especially those needed for AI, are water cooled. But the waste heat water coming out of them is usually around 140°F, Eddy says.

“You could probably heat a lot of greenhouses across the country, certainly in the lower half of the U.S., with 140-degree water,” Eddy continues. But getting the heat up to 170°F to allow for greenhouses to be built anywhere a data center might be would require a heat pump. That would be a major component of the central resources hub, Eddy says.

Given the mismatch of needs and outputs of the data center and the greenhouses — for energy, heat/cooling, CO2 and more — the central resource hub could also connect to other components of the farm park model. Chilled water from the greenhouse could go to the data center but also to a refrigerated warehouse, for example.

“We can bring in the cold chain, the logistics and other key service components to really build out this park and this food production center,” Smith says. “In Virginia, they talk about, ‘Well, if we’re going to have tomatoes and other crops, we might as well have a salsa plant right next to it.’ So, food processing, et cetera.”

The et cetera could include a variety of efforts, depending upon what a community needs. Smith gave the examples of ag tech like robotics, university research, biomass processing and more.

“We can do this in a way that doesn’t impact the data center development pace,” Smith adds. “We can really build out multisector business clusters and the workforce talent pools because we have this concentration. And all of this can be adapted in both rural and urban settings.”

Potential Benefits of a Farm Park

Smith describes farm parks as a new type of business park that focuses on food, energy and data with benefits to the local communities where they, and the data centers they would surround, are located. However, he also highlights how “there’s hardly anyone that likes data centers.”

“We hear about rural communities pushing back. These are generally small rural farming communities that are economically challenged where these data centers come in and locate,” he says. Despite the impressive tax benefits that a data center can bring a location, he adds, “the data centers need a reputation improvement.”

Farm parks could bring that improvement, Smith says. According to the feasibility study, farm parks could bring more and more diversified jobs to a community than just a data center alone. They could also bring locally-produced food, improved food resilience and create economic drivers in the community.

“If we do all this, its $5+ billion dollars of economic development,” says Smith, referencing some of the largest models explored in Virginia. “I think that’s a pretty conservative number at this scale. This is creating 1,000 jobs or more. This is not just 50 or 100 jobs in a single data center, and we can really have it make a dent in feeding people through CEA.”

The Packer logo (567x120)
Related Stories
The company says the promotion of Lawrence Mallia to vice president of AI strategy and product solutions and addition of Manjusha Sunkavalli as a data scientist comes as its moves its AI-driven solutions from vision to measurable results.
Albertsons Cos. has launched the AI-powered Intelligent Quality Control tool that uses computer vision to help distribution center associates more accurately and consistently inspect fresh produce.
Great Lakes Tek Flex will tackle the unique challenges of Midwestern growers by connecting them with autonomous robotics and AI solutions to solve labor shortages and accelerate technology adoption across the region.
Read Next
The Canadian province looks to a massive acreage expansion to serve as an export-quality powerhouse while driving domestic sales at home.
Get Daily News
GET MARKET ALERTS
Get News & Markets App