(This post has received a lot of feedback as expected, I’ve posted some update information at the bottom of the post from Sam Johnson’s team and from Kelix themselves)

Last month, I published an interview about the Kelix ground coupling system and there was interesting response. Here’s why I did it: I always want to speak with anyone who is actively working on making geothermal more kick ass. By kick ass, I mean profitable and increasing adoption of the technology.

The main item that struck me after I published the post is the few nasty emails I received. Instead of fellow geothermal pros supporting one another and figuring out the best places for new technologies, I noticed it seemed the industry was trying to show where a new technology WOULD NOT work, instead of where it would work. This is childish and we shouldn’t tolerate it.

After the piece, Sam Johnston approached me and offered to share some AMAZING DATA on a comparison study they did between a closed loop geothermal system, a DX system and the Kelix technology.

A few more notes before Sam begins. 

First, The main issue I’ve noticed, and the Kelix article highlighted, is that the geothermal industry seems (someone correct me if I’m wrong) to not have a standardized testing process and way of communicating different products to the market. What this causes is a “she said/he said, manufacture vs manufacturer debate”. This tends to increase the risk associated with using the technology both for property owners and EPC contractors who want to get into the business. Anything we can do to create a more standardized way of communicating and testing new products will DRASTICALLY improve the industry.

Second, please don’t email me any comments about the article. Put all of your comments and recommendations publicly in the comment sections. The reason I’m stating this is simple, I’m assuming I’m going to get some angry emails from EPC contractors, engineers, architects, and manufactures about something that isn’t exactly correct in this analysis and that it’s not 100% perfect. What we need to take away from this case study is the intent behind the individuals who performed it. What are they trying to accomplish and WHY?  If we believe that they were trying to make the industry better then we should figure out what makes sense and is useful about their analysis and where it could be improved.

The reason I asked Ryan Carda to provide his feedback is that I know his goal is never to bad mouth other people but to move the industry forward. His comments about the data are amazing and will be useful to anyone who is going to do further testing and analysis on ground source technologies.

If you have any questions, comments, etc about these case studies please don’t hide, leave them publicity in the comments and they will be addressed. If you have any serious problems with the information, give me a call 617 702 2676, but if you’re goal is simple to cut people down, I’m likely not going to listen to you.

Enter Sam Johnston from JTI Energy

During my experience with Environmental Compliance Services, Inc. (ECS) I had the privilege to be part of a team that was to demonstrate geothermal alternative energy in a very special way.   This effort was to become central to a branded division named Terraclime.  What follows is my article distilled from many individuals who participated in this comparative systems effort. The main purpose of this article is to promote alternative energy, especially geothermal by telling this story and revealing lessons learned.

As a member in a partnership, ECS acquired the Mill, some 150,000 square feet in 2007.  The Mill was seriously contaminated and considered a “brown-field” property.    ECS’ core business is remediation; remediating equals changing  “dirty-dirt and dirty-water”.  The Mill became a great “brown-field to green-field” story.

The Mill had its beginning in the early 1830’s when modern industrialists of that time were attracted to the nearly 30’ water fall drop of the adjacent Mill River.  This hydro power would help develop their silk plantation and needed silk processing.

After the nationwide financial collapse in the 1830’s, the Northampton Association of Education and Industry (NAEI) purchased the Mill.  This utopian group was trying to counter the evils of slavery in our largely agricultural society that was fast becoming an industrial nation.  The NAEI was an important stop on the Underground Railroad and major participant in Suffrage efforts.

Anyone working in the ground source geothermal trades or for that matter, any of the refrigeration “heat-pump” technologies knows there are tremendous successes and some awful failures.  I have both great successes and a few notable failures to my name…  Many of the participating trades and practitioners offer sound advice based on experience but, unfortunately, much conjecture is usually stated to explain what is not well understood.  The Mill under the guidance of ECS’ Senior VP, Dan Felten (PE, CGD, +) was a great step toward demonstrating truths about ground sourced geothermal and dispelling some of the misinformation.

A large portion of the Mill’s third floor location nearly 6400SF was to be occupied by ECS engineers, project managers and support staff and was my home as the Alternative Energy Program Manager (which became Terraclime).  This space was divided into three sections:

An Eastern 1/3 of third floor using an EarthLinked Technologies Inc. (ETI) 5 Ton DX, refrigerant to forced air.  Initial system configuration used 407C but was downgraded to R22 due to 407C performance problems most notably fractionation.   This means the chemical components in the 407C refrigerant changed from a gas to a liquid at different rates which severely affected the refrigerant’s ability to process heat both in the ground and at the fan coil.  Five 100’ copper U-Loops were installed in 3.5” diameter bores at approximately 30⁰ from vertical.  Bores were grouted with SuperGrout.  This grout was a “first” for the Mass DEP/UIC.  Because of concerns about copper expansion and contraction in the SuperGrout, a 6” high by 3.5” diameter soup-can shape was poured at the top of each U-Loop so copper could be observed in the grout across a wide range of possible temperatures ranging below zero⁰F to over 150⁰F.

A South central 1/3 of the third floor used a Carrier (ClimateMaster) 5 Ton, 2 stage water-to-air system installed with 25% EtOH (ethanol) in 306’ concentric tube Kelix.  Kelix inventor was introduced to our group at a Boston NESEA show while visiting with Massachusetts DEP, UIC folks who was expressing unofficial interest in exploring ground heat exchanger alternatives.  Informally, we were assured we would measure between 5 and 6 tons due to its unique concentric tube, turbulent flow design.  If correct, Kelix could prove to be as much as 60% less expensive than traditional U-Loop.

And a North central 1/3rd of the floor used a WaterFurnace 5 Ton, 2 stage water-to-air installed with 25% EtOH in 2 x 500’ x 1.25” HDPE U-Loops.  Both water based systems were interchangeable with either outside ground heat exchanger using a manifold piping system.  From initial design considerations, it was decided to put a triple pump configuration on the U-Loop and a double pump on the Kelix.

The Kelix was installed in West parking lot, 306’ by 5.5” bore, 6” on initial 12’ to accommodate sleeve and then grouted with SuperGrout.  Two 500’ x 1.25” HDPE U-Loops were installed using standard thermally enhance geothermal grout sand mix.

Data Acquisition used Measurement Computing system initially then Web Energy Logger using a One Wire loop system to facilitate a Web accessible database (WWW.WELServer.com/WEL0201/).

Over 75 data points polled at as little as 60 second intervals have been installed.  Types of sensors include transducers for amperage, pressure and flow and many thermocouples for temperature.  Real-time display and archived data have been made available for public consumption. Many educational institutions have acquired this data.  Delimited “flat files” containing a month’s worth of data are about 16MB (zipped about 3MB).

What follows are sample analysis and assessment efforts derived from this wonderful, growing source of data for the Mill systems.

DX ELT/EER July analysis


Kelix/Thermocouple cooling analysis

Water-to-air COP cooling assessment comparing both the WaterFurnace and Carrier systems

DX-to-air COP heating/cooling assessment


Cooling – heat rejection analysis, three different outside heat exchanger fields

Snap-shot results

  • The water based systems achieved a heating COP of 3.5 to 4.1 and cooling COP 5.5 to 6.1
  • Mill DX refrigerant system achieved heating COP 3.0 to 3.6 and cooling 3.4 to 4.3
  • The SuperGrout has performed well and has not cracked nor separated from the copper U-Loops at their near manifold test points since 2009 installation.



  • Equipment was less expensive to install, drilling completed in one day, less grout, less diesel, less labor
  • Copper highly conductive and SuperGrout TC > 1.0
  • No pumping penalty
  • R407C was proven to be a design problem due to fractionation. Newer models from all U.S. manufacturers now using 410A
  • This Mill DX system is not designed for two stage nor variable speed compression therefore was more costly to operate across same bin temperature range
  • DX was not well suited for our third floor application; as much as 38’ rise from compressor to third floor and 12’ drop from the compressor to manifold field; too much refrigerant and too much work for compressor
  • DX HVAC talent not readily available


Note, the water based Carrier and WaterFurnace systems were comparable with slight variations in performance some of which was attributable to Southern and Northern controlled space exposures.

  • More cost effective operationally largely due to two stage control
  • HVAC talent readily available
  • Easier to pump water/EtOH mix some 50’


  • Greater turbulent flow with less pressure drop
  • Minimizes “cross talk” between inlet/outlet
  • SuperGrout TC > 1.0 vs 0.8
  • Fiberglass casing more conductive than HDPE
  • Kelix Equivalent Capacity (relative to U-Loop) = 0.376. Equivalent  to 120 feet/Ton vs U-Loop at 152 ft/Ton
  • Kelix is ~20% more efficient?


  • Easier to install
  • Well understood with very long track record

The Mill experience would not be complete with acknowledging a wonderful bevy of engineering and technical talent. Additionally, we had some very special team members that became known as the “Geo Team Seniors”.  The only criterion for membership was to meet AARP’s membership minimum age and have gray hair.  This ruled out the majority of our young, upstart (often irritable) engineering staff.  Although there were others, this article is pointing out Frank Biancardi.  Frank spent more than 40 years at United Technologies Research Center as the lead executive.  He was responsible for products and funding developed for P&W, Carrier and Hamilton Sundstrand.  Frank has over a dozen patents, dozens of papers and published articles to his name.

With great patience, Frank taught us about refrigeration thermodynamics from the “center” out.  Frank introduced us to working models for enthalpy which greatly explains this very misunderstood geothermal technology.  One session with Frank and his enthalpy curve and you knew there was a lot more to this science than most everyday practioners touted…. and why we sometimes do not get it right.

Most importantly, Frank introduced us to system central refrigeration cycle analysis.  Most of us are familiar with determining system performance in terms of ground heat exchanger measurements or inside air handler or water processing measurements.  Frank taught us how to use system compressor analysis programs.  We learned to measure the overall system performance from the compressor out.

I can assure you anyone who finally understood what Frank was teaching us has a much better understanding of this wonderful thermodynamics technology.

Let me address three questions of how I feel this data will impact the day to day operations of geothermal EPC contractors.

How does  the Mill data impact specifications for product?

  • Don’t trust conjecture of any kind: it’s all about “Q”, that is to say, quality of energy from or to the ground.  Calculating “Q” is equally important for air sourced systems.  “Q” is often used as a scientific symbol for describing energy quality.  Higher “Q” indicates lower loss of energy.
  • To date, there are no “magic” answers for any heat exchangers.  Innovation is important, but efficiency improvements tend to present themselves in small, incremental steps.
  • All manufactures’ claims need to be verified through test and measurement and testimonials.
  • The two big variables are always the ground (or air) heat exchangers and the customer’s energy usage habits.  My dear sister-in-law used to run the heat in her house with the kitchen door cracked open to the outside winter air!  She claimed she could not breathe without fresh air.  But that two inch door opening made a huge impact on the number of house air changes and therefore, the heat load!  And the poor possible geothermal installer would probably catch hell for not meeting customer expectations.
How does this data impact Production Estimates?
  • Manufactures specifications are very much like new cars sales estimated highway mileage…  your actual mileage will probably vary!

How does this data impact the type of customers geothermal is best for?

  • It’s still all about cost/BTUh…  the cost per BTU to install and the cost per BTU to operate (and maintain)  all compared to conventional heating.
  • Oil and propane customers are easiest to sell to…. But
  • Against any liquid fossil fuels, it’s not hard to demonstrate a 16-18% return on investment over a 20 year system investment cycle
  • Three arguments resonate with potential customer:
  1. Scottish – “what will be my return on investment?”
  2. Green – “we cannot keep burning stuff and polluting our atmosphere and oceans…”
  3. Geopolitical – “continuing to spend $50-80 Billion per day on foreign fossil fuels is untenable..”  Senator Bernie Sanders said, “… I’ve been to the Middle East oil producing countries, and they are doing just fine… They just don’t need any more of our money!”.
  • We still need a quality Cost/($BTUinstalled + $BTULifeCycleOpsAndMaintain) modeling system for all best practices alternative energy technologies…
  • Inexpensive DAC (Data Acquisition and Control) is needed with all systems to
  1. Help prove we are meeting customer expectations
  2. Optimal life cycle maintenance and intelligent dispatch such as “… the data says bring a clean air filter with you when you go…”.
  3. One customer referral is worth 100 cold calls


I asked Ryan Carda his opinion on the data and this is what he shared:

I’m always happy to see someone who takes the initiative to objectively evaluate the many options available when it comes to geothermal ground coupling options.  ECS went above and beyond the typical system owner to perform their own research. I think I speak for everyone in that their willingness to share their experience is greatly appreciated.

ECS has a good start in answering the age-old question of “which method is best”.  However, as a design engineer, there are a few things that jumped out at me which did not seem to be addressed in the article.

First, there appear to be a few variables that weren’t considered.  Direct comparisons of one ground coupling method to the next are hard to make unless you address them.  For example:

1)      The first is equipment run-time.  If they measured the actual run-time of each unit, the information would have been quite useful.  There is always the possibility that all three parts of the building weren’t used in exactly the same way resulting in different equipment operating patterns.  What if one part of the building was used as office space, another as a data center or manufacturing and the third as cold storage?  The energy requirements and subsequent ground loads would be different for each “5-ton” ground loop.  A 5-ton unit that runs 25% of the time will not work the ground loop nearly as hard as one that runs 75% of the time.

2)      Another variable that isn’t mentioned is the formation itself.  Do the soil conditions change from the 300 ft. bore used for the Kelix compared to the 500 ft. bores for the u-bend?  For example, there may be groundwater movement at 400 ft. that helped the performance of the u-bend which the Kelix and the DX systems didn’t see.

Second, in terms of ground head exchanger design, the two water source loops (Kelix vs. conventional u-bend) don’t appear to have been designed to produce the same water temperatures, so the head to head comparisons of “bore ft/ton required” wouldn’t necessarily apply.

1)      Loop length is a linear function of delta-t between the loop and the surrounding soil.  If the deep earth temp is 60F, a loop design that allows the maximum EWT to reach 70F will be twice as long as a loop design that allows the maximum EWT to reach 80F.

a.       Half the delta-t means twice the ground loop length requirement and vice versa.

2)      According to the graphs provided, the traditional u-bend appears to be providing a much lower EWT than the Kelix system.  That means the u-bend design could have been quite a bit shorter if they wanted to provide the same EWT’s (and thus the same performance from a ground loop perspective) as the Kelix.

3)      Also, as far the conventional u-bend design is concerned, there was no mention of spacing between loops, grout type/thermal conductivity, etc.  Two 500 ft. u-bends spaced 10 ft. on center would perform very differently compared to two u-bends of similar depth with thermally-enhanced, 1.20 Btu/hr-ft-F grout and 20 ft. spacing.

Third, all of the data in the report is from a single snapshot in time but doesn’t mention anything about long-term performance.

1)      Unless you live in the middle of the country, you will either need more heating than cooling or vice versa.  When one mode of operation is relied on more heavily than the other, the ground loop is exposed to unbalanced ground loads.  Especially in commercial applications, in a typical design, the long-term design lengths (sometimes referred to as the 10-year lengths) will be much different than the first year design lengths.  The increase in loop length requirements from the first year of operation to the 10th year of operation is simply our way to account for long-term draw-down or build-up of energy at the center of the loopfield.

2)      In extremely unbalanced ground load situations, soil volume becomes a critical factor to allow adequate energy storage in order to offset the negative effect of unbalance.  In these situations, bore length reductions (by way of extremely high grout thermal conductivity values, concentric heat exchanger design, double u-bends in a single bore, etc.) will have considerably less impact because thermal storage capacity is needed to counteract the energy draw-down or build-up in the field.  In other words, differences in design lengths from one construction method to the next will quickly approach zero as ground loads become more unbalanced as soil volume will be needed for energy storage.

a.       The Kelix system consists of considerably less soil volume than the conventional u-bend system.  What are the long-term performance implications of that?

Again, thank you ECS for sharing your story.  By digging even deeper to answer a few more questions, you will be doing us all a great service.

 Enter Chris Williams Again 

Thanks again to Sam Johnston and his team for gathering this data and thank you to Ryan Carda for providing some more insights. I’m looking forward to many other comments about what is really useful about this case study and where it could be improved further.

Updated Information and Commentary from Sam Johnston’s Team

Field conditions

Ryan’s comments about possible ground formation variations for each 5-ton system installation are certainly valid, but no extensive ground sampling was undertaken prior to drilling each of the ground systems. However, certain small differences were present as to where the ground fields were installed, and anomalies noted. Both water loop fields were located on the westerly side of the building, in the parking lot some 75-100 ft from the Mill river..  The DX ground loops were installed inside the Mill structure in a high bay area to demonstrate the use of compact drilling and hence lower cost equipment. The five 100′ loop DX system was arranged in an “eyebrow” pattern with bores originating in a 6′ distributor manifold area.  These DX bores were at least 250-300 ft from the nearest, much deeper water bores. These DX bores were drilled at 30 degrees from the vertical, and about 100 ft in length, as per manufacturer’s and driller’s guidelines.
The vertical U-tube and Kelix water loops were separated from each other by 20 ft.  During the drilling processes for these systems, no significant groundwater was encountered, nor was a detailed mapping or study of the ground conditions undertaken due to time and more significantly budget limitations. However, as Ryan has noted, if the ultimate installation were for a multi ton/multi bore commercial building, the added ground details and expenditures would be essential to provide  design/construction confidence. Even though the Mill is technically commercial space, these were residential modeled systems.
In retrospect, the U-loop was perhaps over designed (at 200 ft/ton, 2 x 500 Ft reverse return) to serve as a bench mark. In addition, ECS had access to highly qualified earth science professionals based on their 28 yrs experience in environmental remediation.
The three selected manufacturers were invited to participate in initial design issues and subsequent data analyses, and performance issues and some assistance was obtained and proved helpful. Early estimates of the building space heating and cooling requirements, based on the 1830’s construction of the Mill were offset by extensive insulation, new windows, and othe features during remediation efforts. As a result, the backup natural gas boiler/hydro fan coils never saw use.  Some three years into this process, the natural gas utility came to find out what was wrong with their meters since there was never any registered usage!
In presentations done by folks who participated in this endeavor, attempts were made to publish installation costs.  However, relative cost comparisons have some merit but at this juncture cannot be published as industry typical.

Some comments on the article from Matt Schaefer from Kelix


First of all, thank you for facilitating this review.  At Kelix we have experienced a lot of skepticism about the ability for our product to improve on existing methods, and dissecting actual test results is the best way to determine our capacity to do that.  I would just like to make a couple points from our perspective, based on what has been included above;

–          The lack of standardized testing and new product introduction methodology is a very real issue for the geothermal industry, and one that it needs to address ASAP in order to move forward.  As an example of how this can affect companies like ours trying to make improvements, we get asked “is your product IGSHPA approved?” all the time.  IGSHPA is the association everyone looks to right now for validation.  Unfortunately, when we have asked them to review the Kelix GHE and state that our materials and/or design is acceptable for ground source use, we have been told that they do not have any formal evaluation standards or processes.  This keeps us (and other potential innovators) from making progress and helping the industry to grow.

–          One source of skepticism about Kelix in particular is that, in the past, the company was making the claim of “5 tons out of a 300ft borehole” without properly understanding all the factors that affect geothermal performance.  We now know that we can shorten loop lengths by 20-50%, depending on the local ground conditions and the load balance of the building.  This is much more defendable because we can back if up with data and simulation results, and it acknowledges the fact that performance over time dictates field sizing.

–          The statement “the ground rules” is something we hear a lot.  In reality, the statement should probably be “the ground and the min/max water temperatures rule.”  Ryan points this out when he said the loop length would double by changing the max EWT from 80 to 70.  Note that there was nothing about the ground in that statement.  Kelix lowers the borehole resistance and makes it easier for energy to transfer between the fluid and the ground.  So when the ground is the limiting factor (e.g. when conductivity/diffusivity are low or the load is highly unbalanced) Kelix can’t improve performance as much – but when the water temps are the limiting factor (e.g. when conductivity/diffusivity are high and the load is fairly balanced) Kelix can make a big impact.

–          We know that our product isn’t applicable in all cases.  It has the best value proposition in areas where space is highly constrained or where drilling costs are high: typically $12/ft or above becomes worth looking at.

If anyone would like to contact me directly, I am happy to discuss these items in more detail.  We can also provide side-by-side test results showing our borehole resistance values, run simulations to show how much we can shorten the loop length on a particular project, and if warranted discuss pricing.  And you can also learn more at our website (www.kelix.com).