'Double piped' well?

@ gsmith22,
I would like to know more about your design modeling and data for your project please.
Eric

 

I'm in somerset county, NJ. I bought a house last year that is ~30 years old with original propane furnaces and a/c. They need to be replaced, propane is ridiculously expensive, and I don't have natural gas service. Geo was a natural option and I'm in the process of having it converted. I'm on a hill with rock relatively close to the surface so horizontal piping was out and vertical was my only real option. Only considered closed loop, local geology would prevent open loop. As you know, its expensive to drill vertical wells so I wanted to make sure the heat exchanger side of the work was optimized. I also wanted it to work well the first time, especially given the current 30% fed credit (If I have to spend money, its best to do it now). Being an engineer and having some back ground in heat transfer and fluid dynamics, I took the opportunity to research how systems are designed. So I got IGSHPA's design and installation guide for residential and light commercial and mainly just followed its design process chapter to chapter. Its very well written, easy to understand, and provides a nice roadmap for anyone looking to design a complete geo system. Part of the loop design process is to figure out if there is an imbalance in heating and cooling. They have you do what is called a weather bin analysis that ultimately provides an estimate for the amount of heating and cooling runtime as well as Btus pulled from/rejected into the ground. Runtime and heating/cooling imbalance gets factored into the equation that provides the minimum amount of borehole length. In my case, it caused the overall length of boreholes to be ~20% greater than I would need if heating and cooling were balanced.

As it turns out, I ultimately asked for my ground loop to be increased over what the mechanical contractor was going to specify. He came up with the loop design with what appeared to me to be a simplified computer algorithm created by Waterfurnace - GeoLink I think its called. It seemed to have alot of generic inputs that weren't necesserily specific to my location/geology/usage and didn't seem to be readily understood by the contractor. One of the things that stuck out to me in the design manual's discussion of loop design is that for unbalanced heating and cooling situations, the min/max water temps of the loop will change over time if unbalanced loading isn't taken into account. Essentially, you can make up for the imbalance by either putting more vertical borehole in the ground and/or separating the vertical wells further apart. I used a combination - added about 50ft to each hole and spaced them at 20ft vs standard 15ft spacing. Going to 25ft spacing really didn't help nearly as much as going from 15 to 20 so the rest of the imbalance I made up with borehole length. Hope this helps.

 

Thank you for your response. It was very helpful.
Eric

 

What you were describing with unbalanced loads having a memory effect in vertical loop fields is only really relevant in larger commercial systems, where the inside bores (in the middle of the loopfield) are more isolated from undisturbed ground, thus heat (or coolness) traveling through the ground reaches the inner loops much more difficult.

It is not an issue in smaller vertical loopfield with only a few holes.

 

The physics of heat transfer remain the same regardless of a residential or commercial installation. It is true that a larger loopfield (more often found in a commercial application) is more prone to an unbalanced load memory because of the grid effect of many bores reducing the ability of the interior bores in the grid to be replenished by ground heat/cool. But, unbalanced load is unbalanced load. Did it massively affect my borefield? No, (3 bores in a line) but the effect wasn't zero either as you seem to indicate. The IGSHPA's manual that I listed above goes through the calculations and it is clear that unbalanced load needs to be accounted for. Spacing at 20ft vs 15ft had a very distinct effect for me and I would have ended up with a more longer bore length than I used.

With regards to the OP's original issue, he has a single bore so no overlapping heat cone effect from adjacent bores. I was merely pointing out that data from one (mild) winter is probably not sufficient to declare victory and to keep an eye on entering water temps especially given his climate (same as mine) with much more heating than cooling.

 

One thing is a memory effect in a large loop field, the other thing is an unbalanced load between heating and cooling. 2 different things!

A loop field needs to be designed to cover its dominant load.

But with a single loop for example, you do not have a memory effect when loads are unbalanced. The memory effect stems from inner boring which heat (or more likely coolness from the outside does not reach since the outer boreholes shield the inner boreholes.

IGSHPA has a good job in setting standards for the geo industry, but there are many factors out there they do not cover either. The IGSHPA manual are guidances ...nothing more.

 

I'm going to reply only to correct what I think is a mixing of terminology/concepts. I don't want to have someone reading this coming away with the wrong idea about thermodynamic behavior.

memory effect (ground temp changing over time because of heat extraction/rejection due to geosystem) is driven by unbalanced loading. They may not be the same but they are certainly interrelated and definitely not "2 different things!".

If you pulled as much heat from the ground during the winter as you put into the ground during the summer (balanced loading), the ground temp will not change over time (have a memory). Math/logic says this must be true and the ground wouldn't care if you were a small residential bore system or a large commercial building with a huge grid of vertical bores. The amount of heat into the ground equal to the amount of heat pulled from the ground nets a zero energy transfer to the ground when looked at on a yearly timeframe.

Only a small sliver of people live where in any year hot and cold are balanced (or nearly so), so almost everyone has to consider unbalanced loading to some degree and possibly in different directions (canada mostly pulls heat; florida mostly rejects heat). If you don't consider unbalanced loading and just design the bore system length for the design heating (or cooling, whichever controls) load, you will lower (or raise) the ground temp around the bores. This is because if the geosystem can't rebalance the ground temp (via heat pump over both heating and cooling seasons), the earth has to rebalance the ground temp by heat moving through the ground toward the bores for a heating dominated climate (or away for cooling). If you only provide enough bore length to transfer the heat or cooling seasonal loading and don't consider how the rebalance occurs, the earth can't rebalance the bore quick enough between seasons so you end up with changing ground temps over time. Providing more bore length and/or spacing the bores further apart spreads the temperature change of a bore over a larger volume of earth so that the earth can make up the difference more easily.

But you don't have to believe me, look at the bore length design equation in the IGSHPA's design manual and you will see that 1) minimum bore length is based on the amount of heat pulled (or rejected) into the ground by the system and 2) a multiplicative factor is applied to the minimum bore length to increase it. The factor is directly dependent on the value of the unbalanced loading and bore hole spacing/grid layout. The math mirrors the physics. This manual is far more than just guidance. At a minimum, its the standard of care for the proper design of a ground source heat pump system - deviate/ignore it at your peril. It does provide suggestions regarding installation methods so on that point only, maybe it could be considered guidance. But for design, no way. I challenge you to point out a more thorough design guide for a residential ground source heat pump system.

In a large commercial system with a grid of bores, it becomes even harder for the temp of the interior bores to get replenished by the earth. So here especially unbalanced loading is critical (and maybe gets accommodated for in a residential system by the lack of large grid systems and/or general oversizing due to lack of engineered design). But that commercial system should also be designed using ASHRAE's geo design manual and not the IGSHPA's manual. However, I would point out that the ASHRAE and IGSHPA's manuals have nearly identical bore hole length design equations indicating the fundamental concepts are the same.

For the OP's issue, that single loop can absolutely have a memory affect if that single loop isn't long enough to deal with both the system capacity and the earth's ability to replenish the ground around it before the next heating season since the cooling season won't reject as much heat as was removed (for northern NJ). What it won't have is the compounding affect of multiple adjacent bores further degrading the earth's ability to replenish it. Again, my point to the OP was to be cautiously optimistic given only one (mild) winter under its belt.

 

Nope, this is not how it is.

In general, you want the entering source temperature to be around 30F, and incoming not above 90F, as a rule of thumb. So it is within the operating range of the geo heat pump.

As an extreme of an unbalanced load, you have systems which are heating only for example. They drop to the temperature you design to, like 30F, and if you turn them of, the heat traveling through the ground replenishes the ground usually within a month, back to the average ground temperature.

Here is an example. Heating and domestic hot water only, drops down to 30F in mid winter, uses phase change to not drop below 30F, and within 1 month of turning it of is back to 50F which is the deep ground temp in our area. No effect of unbalanced load. July 1st, it is always 50F, +/- 1-2 F.

http://welserver.com/WEL0383/

The 3rd graph from the bottom.

We monitor about 200 of our systems. their behavior is always the same.

A larger commercial loop field is different, since the center bores cannot be reached by heat and coldness as easily and are shielded by the outer bores. That one typically heats up with unbalanced load, typically in cooling dominated applications.

This is simply not linear heat flux as the IGSHPA formulas assume, this is heat moving through the ground which none of the formulas account for. Or many other factors which are not accounted for. You design your loop to support a certain EWT, the grounds are always replenished quickly within a month in a residential system with 1 or 2 boreholes.
Again, the formulas are only as good as the factors the include and account for. Neither ASHRAE or IGSHPA account for certain aspects, I stopped using any of their formulas, or even design programs. After a while you need to throw them out of the window and make manual adjustment, other wise you have badly performing systems.

Most people don't check their system performance over years, we do. That way you learn that boreholes and fields behave very different than any software or formula suggests to you.

 

I have very little experience watching ground loop performance, but your explanation is exactly what physics and common sense suggested to me. The ground is continuously trying to keep any and all loops at average ground temperature. As soon as a system is turned off, the ground is changing it's temp back to average.

 

The case study that docjenser provided is being helped significantly by ground water movement that moves heat to the bores which is why they recover so nicely/quickly negating the unbalanced heat extraction only use of this particular geosystem. Nothing like an outlier to assume as a standard! If you look closely at the yearly EWT graph, you can see almost daily fluctuations of 3 to 5 degrees in the middle of winter when that system is definitely running at its max! I'll take good ground water movement around my bores 6 days a week and twice on sunday over the best possible rock that exists for heat transfer. Which is why open loops have little variability in their EWT over the course of a year - the heat is constantly being moved away by ground water. You can't plan to utilize ground water movement for heat transfer unless you do drilling/testing prior to the system installation to verify its presence and how much heat it can move. Its a specific site property and there would be no way to quantify its presence and ability to move heat without the testing. Alternatively, you could do the testing and then incorporate the heat movement which for the above case would likely show heat transfer values that are orders of magnitude higher than dry rock by itself (what is used by IGSHPA in their soil and rock field manual). IGSHPA and ASHRAE produced a procedure that works everywhere and the suggestion that it is garbage because of specific geology at this site/local region is a joke. Rather than throwing IGSHPA and ASHRAE research and data out the window, it might be better to figure out why your results differ (in a positive way too) from what their standards suggest and attempt to determine when these helpful factor(s) may not be present so you don't end up on the wrong side of an under-looped system.

 

That is all true. But it is still also true that the ground is trying to average temp all the time. Basic laws of thermodynamics state this. I am not experienced enough to know for sure, but I would expect that if unbalanced loads are a problem, the loop field is probably undersized all around and will have poor performance and extreme EWT.

 

It is not a case study, it is short term and long term monitoring in real time.

You make a lot of statements here, without having and good data to back it up!

First, it is not an outlier, the loops all behave that way, at least the ones we monitor. Here are a few more, they might not all be live, due to network issues, but most are, and give you a more representable sample.

http://buffalogeothermalheating.com/live-systems/

Second, the 3-5 F degrees fluctuation stems from daily changes in load and specifically solar gains, when a loop rests. If the sun is shining, we sometimes see a system not running for 8 hours or more, and the loop goes to it s equilibrium with the ground, which usually is about 5F warmer. Only if running, the loop temp drops, only then it can create a temperature difference between the water in the loop and the ground, which then draws heat from the ground.

Third, the system "definitely" never running at max just because it is the middle of the winter. It does so on the coldest day of the year at around 0F, but average temp is about 32F in the middle of the winter, meaning it does run at 50% load most of its time. If it would be running full out, you would see a line without fluctuations.

IGSHPA developed installation standards which work with a good margin of error, but the field has emerged and we now are building different systems using different materials.
The formulas and software are based on the linear heat transfer model, they don't account for any internal and external gains, don't account for domestic hot water making in the summer, and many other factors not accounted for. I remember the time where the IGSHPA manual told us to size a system for A/C loads, and supplement everything else with electric supplement heat. Boy, would I would have many unhappy customers if I would do it that way.

The point being that unbalanced loads are not an issue in smaller loop fields, since they don't have outer bores which shield the inner bores from heat or coldness of the ground, so the inner bores have more of a memory effect, which typically only occurs in a cooling dominated larger multifamily or commercial application (which have to have larger bore fields).

I install and design geo systems in different geo graphic and geologic areas, never have seen a memory effect (meaning the loop field heating up over the years) or heard of one in single family homes. Most of our single family homes have a single borehole. If you insist they exists please provide the data to show that, and don't simply refer to the IGSHPA or ASHRAE formulas which we know do not account for many factors playing a role here.

 

I looked at your "live systems" and most don't have data for EWT - I'm not debating they all behave the same because they are all in an ~30 mile radius of Buffalo. Western NY has very uniform geology similar to the mid west (because the rocks were laid down via a shallow sea hundreds of millions of years ago). Its nothing like where I am at (newark basin - failed rift valley during Pangaea breakup) where vastly different geologies exist within several miles of each other. I don't think it would be incorrect to conclude that all your vertical bore systems behave the same because they are all drilled into the same (or very similar geology). So from your perspective, it isn't an outlier and they all operate the same. Again, no argument given the uniform geology they are installed in.

The fact that you can see a 3-5 deg temp change in a day indicates heat is being moved toward the bore rapidly and replenishing what was removed by the heat pump. Water does that, rock alone has a very tough time doing that, dry soil won't do that.

ACCA Manual J (since we are talking residential) is based on accounting for heat gains/loss from both external and internal sources to determine building load. IGSHPA's standards/calcs have nothing to do with this calculation (and point to Manual J). You take that building load and use IGSHPA to determine equipment size and loop field. I don't know what IGSHPA standards you are talking about (mine are from 2009 so not that new). The ones I am using include domestic hot water making (its a load on the field just like heat and a/c) and don't suggest designing for a/c and making up any heat shortfall with electirc stip heat unless 1) you are in a heating dominated climate and 2) using single stage equipment. Otherwise they suggest designing for full load (heat or a/c) if using two stage or variable speed equipment. And the reason for the limit to a/c design with electric strip heat for single stage equip is/was two fold - so you get good dehumidification in the summer (limited cycling of equipment sized for heat) and you balance the heat and cooling load on the field (ie what we are talking about). If you heat more than you cool and limit the heating load on the filed to the cooling load (ie you install equipment for the a/c load), you get automatic balancing on the field. Use max load with two stage or variable stage equipment and the calcs include unbalanced load using a weather bin analysis. Its in the written text and the math of the IGSHPA design and installation book (at least the 2009 version).

To have unbalanced loading potentially show up as lowering/rising ground temp over time, monitored loop behavior would have to follow the following pattern: the EWT at the start of heating season (a/c just ended) would have to be well above the deep ground temp. The heating season and the surrounding ground would then have to lower the bore temp down over the full heating season. Similarly, the EWT at the start of cooling season (heat just ended) would have to be well below the deep ground temp with both the cooling season and the surrounding ground bringing the bore temp up over the full cooling season. That pattern of EWT could result in a system that works fine or one that might be subject to long term deep ground temp change. It would depend on the actual values. You have stated that none of your monitoring matches that pattern so i would expect that none of your systems could have an unbalanced load problem.

On the system you linked, it returns to deep ground temp very quickly without even a cooling season. This indicates the loop is way oversized for the load being applied. Shorten up the loop on that system to the point that the EWT at the start of the next heating season (its only heating) is just getting back to the deep ground temp and maybe you might have a problem. Maybe. But in reality it gets back to the deep ground temp before the heating season is barely even over which can only mean it has more bore length than it needs. I'm not debating this doesn't work, I'm sure the homeowners are thrilled it works so well, and you obviously don't have to worry about unbalanced load on that system. Everyone is happy. I'm also not suggesting you knew prior to drilling that it would work so well/could of had shorter bores. It takes a lot of pre-construction geotechnical work to figure that out and its cost prohibitive sometimes for even commercial work let alone residential work. But don't kid yourself, its not that "boreholes don't work that way", its because there is so much extra bore in that system (and maybe all of your systems), unbalanced loading won't ever be a problem given the combination of your geology and however you are sizing your field (since IGSHPA manuals were thrown out the window).

 

1)
The 3-5 deg temp change in a day stems from the a cold front moving through, or a warm front, and rapidly changing the entering water temperature depending the load on the loop. The system also takes temperature samples every 9 1/2 hours for a graph displaying the entire year, and filters out the times which the heat pump is not running.
It reflects pretty much the difference of a loop running at half load for the day, and full load.

2)
Manual J does not account for internal and external heat gains for heating, it only includes heat gains for cooling. You are wrong there.
"No credit is taken for solar gains or internal loads in calculating the heating load because the peak heat loss occurs at night during periods of occupant inactivity."
https://www.nrel.gov/docs/fy11osti/51603.pdf

3)
Obviously if a loop drops reliably to 30F and then stabilizes, it has the correct size. It is not oversized for the load as you incorrectly state. I am not kidding myself. I a following the data we generate.

4)
I have many unbalanced loads, but that does not matter on a residential loop field, since there is no memory effect on residential loop fields. They are too small.