We are evaluating the feasibility of a geocooling solution for a mid-size data center in northern Vermont. There are lots of state-level financial incentives but I can't put my finger on the design specs of such a system... even rules-of-thumb for sizing a system (read: cost magnitudes) would be useful.

The interior design calls for replacing the primary building chiller water plant with a variety of heat sinks, including heating a roof-top greenhouse, heating the warehouse and office areas, pre-heating building potable water, and any other creative heat uses we can find. Evaporative chillers are the typical mechanical standby unit and we are trying to avoid refrigerated chillers, especially during the summer. 

The basic design metrics are 100 cabinets with around 1MW of heat. Current loads are in the 150-175kW range and growing as we are designing the current expansion phase... We've raised the cooling water setpoint by 10F to 53F chilled water feeds, making geothermal cooling options possible in northern climates.

From what data I found, Vermont has low equilibrium ground temperatures (7-8 degrees C or 44.5-46.5F) but low thermal conductivity (2.5W/m/K). A good-bad situation but not insurmountable. Data source: http://smu.edu/geothermal/

From the Earthlinked site referenced in other forums, up to 250' of water-filled plastic tubing (1" OD, I suspect) per ton/3.5kW is a design metric, in a 4-6" drilled well. Refrigerant isn't really an option since chilled water is a common denominator in most HVAC equipment. How does the hot side and cold side stay separated and not contaminate each other?

So what sort of spacing to you need between cooling wells to maintain equilibrium temperatures? Our building plot has 12 acres total with 4 greenfield acres for a geothermal application, if we can make it fit. If we tear up the car park (and re-pave it later), we can get 7 acres. 5x well diameter for spacing? 10x? I'm assuming vertical orientation to get the most cooling volume available.

Am I even close?

Ideas and constructive comments are welcome.

Cheers,
-Dave

Dave,

Your project is a good candidate for a geothermal system. Vertical wells on commercial cooling dominated systems are usually spaced in a grid pattern 25' each way.

I am not sure what you mean by this "Refrigerant isn't really an option since chilled water is a common denominator in most HVAC equipment. How does the hot side and cold side stay separated and not contaminate each other? "

I suspect that you might be a little confused about how heat pumps work.

However be assured that a geo system will take care of your cooling needs as well as using the extracted heat in what ever way you wish.

A system like this needs to be designed by a mechanical engineer familiar with geo systems.

I was thinking refrigerants sometimes replace water with R134a for more effective heat transfer carrier/higher heat capacity than water. I suppose this could be used in a closed loop system but there is the ground water contamination concerns from leaks. Still, it doesn't change the thermal conductivity of the earth around a borehole so the point is moot.

If heat source and returns are simply a U-loop within a borehole, then it's simple to install... but the desired heat flow of the system would be lessened by having the hot and cold lines right next to each other at the top of the well. The heat leaks from one to the other and not into the ground. Or are the pipes partially insulated (on the inflow side, not the outflow/moderated side) to help avoid this?

In any case, a 25' grid makes sense for a rough design. I've read 300-400 feet deep wells with 1" OD pipe as a typical test design for an individual well. http://www.tva.gov/products/business/TCStudy/index.htm

The pipe arrangement you describe above is typical for a vertical well system.  Sure it would be better if the pipes could be separated or insulated,  but there is no practical way to accomplish this.

How ever,  we have installed 1000's of vertical wells and I can tell you that even though the set up is slightly flawed,  they do a great job of exchanging heat with the ground.

R-134a is a refrigerant, though not typically applied to geo systems.

On first blush this app seems a good fit for geo since it needs cooling year round, plenty of land available, and the northern climate features low deep earth temperature.

A system of this magnitude (far above the <10 ton residential and light commercial projects typically discussed here) would almost certainly require test bores to determine site specific soil thermal conductivity. Basically a well is drilled, U-tube inserted, and a measured amount of heat and waterflow applied to it for a day or so and thermal dissa[ation measurements are taken.

Browse McQuay for some docs on larger systems.

Also browse internet for references to Standing Column Well systems - more common in the northeast than elsewhere from what I gather. Notable installations include a public library in Haverhill, MA and a Quaker Friends facility in Philadelphia.

I agree the application sounds like a good one, but you do need someone on the ground to direct your research- start with area contractors with commercial experience. The efficiency or lack there of, of a typical U-tube is not a starter discussion, as we've said before it's all about footprint and there are many, many ways to skin that cat.
Good luck,
Joe

Amorphis, Sounds like you might want to talk to these guys: http://www.northeastgeo.com

They are the distributer for Climatemaster commercial sized geo systems in New England and will consult with you on system design. They have done some of the Standing Water Column based projects Engineer mentioned. Also search for Geothermal at Harvard University. You will see they install larger SCW systems.

So how about this - brain spasm this morning on the way to work:

Why don't we use air-side economizers to swap out bulk quantities of outside air? Most of the year filtered air would be sufficient for keeping a data center cool, given our climate here. We're looking at buying a building so a potential landlord won't get testy with the structurally-intrusive design. If you could re-use two loading dock doors... hrm.

I envision the geothermal system as the chiller-replacement system for the few hot/muggy days of summer and to dehumidify a small portion of incoming air with re-blending to the main air stream before hitting the cabinets. The geo plant would be just big enough to dehumidify and/or cool a fraction of the air (half? quarter? I'm working on the calculations for peak usage) and would only run if needed.

Blend recirculated exhausting air (warm/dry), fresh filtered air (variable), geo-chiller (cool/moist), and adiabatic-cooled air (cool/saturated) for the right DC air/humidity that works for ASHRAE Class 2 spaces. Class 1 would be ideal if we can tune it that tightly without excess equipment. Both allow for some slack in ideal settings, enough where the overall HVAC solution would not require chillers for a more strict environmental policy.

Selection of HVAC hardware to match external air traits allows for the more power-efficient and effective hardware to condition the air, instead of hacking a refrigerant system to do everything at an expensive utility price. Did I mention redundancy?

Overall design like page 36; environmental specs like page 27:
http://www.thegreengrid.org/events/technical_forum/Day_1.Track_2._ASHRAEs_Building_Ratings_and_C.pdf

I scored a couple of ASHRAE Datacom-series books about data center design and specs - so far the "Best Practices for Datacom Facility Energy Efficiency" is worth every penny of $60. Hugely insightful and I wish I had it a year ago as we started these designs. Interestingly enough, they don't talk about geo systems or more recent air/air design tricks that make those systems much more efficient than old-school brute force HVAC airflow room-wide.

-Dave