Geo vs. High efficiency electric furnace

Joe,
1. I generally have one supply register per room, if I need a 7" duct instead of a 6" duct to carry the extra cfm for the next sized unit, then the cost for that run is only a few dollars more and no more labor. There may be one or two more supply registers needed for the whole house. Returns; I do centralized returns for open areas and one per room that can be closed off (bedrooms for instance) so again no real increase in labor cost for returns.

2. I am confused on your loop sizing strategy, but I see that here you are saying the next larger unit for the same house will need longer loops. So I assume that means that the 80% unit will have shorter total loop than if the next larger unit is used. Shorter loop for the same load means colder ewt during 100% of heating, which means lower heating capacity and lower COP. There is a loop cost increase then between the 80% and 100% strategy

3. There is only a $50 list price difference between the ND038 and the ND049

4. So all in all, the significant price difference is in the cost of the looping so I think the argument is more about loop size than unit size. (perhaps not) But longer loops always mean lower operating cost.

5. When the load reduction is taken into account, and the increased production the longer loop length has on satisfying the same load, then my statement is not contradictory. ( I am not talking about larger units per se, but meeting 80% of the load or 100% of the load.) By reducing the load, the unit and ductwork become smaller to satisfy 100% of the load. The loop length might be the same as others would use for 80% of the NON reduced load, but with the reduced load there is less btu per foot of loop extracted from the ground so the ewt is higher. Look at the heating capacity changes based on ewt for the same unit.

6. In the NE of this Country, the rater community has come to the conclusion that geo does not produce the efficiency promised and they are actively trying to exclude geo from Energy Star rated homes and are down rating the performance stated by the manufacturer. The reason is that most geo units are being designed and installed to meet 80% of the load, have 30% propylene glycol antifreeze, and low loop temperature of 20F at peak load. This creates a system with high pumping energy and low heating capacity and large electric supplemental heat. The raters started looking at geo performance because the Energy Star rated homes were not performing to the predicted level of efficiency because these systems were operating at COPs close to or less than 1/2 the COP in the published AHRI data. This is my biggest motivator to spend my time challenging you and others on the 80% of load is the proper sizing philosophy. Consumer confidence is low when the respected rater tells the builder and homeowner that the reason their house is not performing as expected is because the geo system is not really energy efficient at all and they wasted their money. These same raters push the new air source heat pumps as the real high efficiency heat pump systems. It is our own fault if we continue to push low first cost over operational cost as the design criteria when the reason the customer wants geo in the first place is to have the lowest operating cost possible.

7. I do look at the comments with an open mind. However I believe I am justified in challenging the accepted wisdom that 20% undersizing of geo systems is the proper sizing. If you took a unit that was sized to meet 80% of the heating load and increased the loop length, then you could get the same unit meeting 100% of the load . I have been trying to address this question of proper sizing, should it be 80% or 100%. I have not seen any demonstration of the efficacy of the 80% of load sizing strategy, with the accompanying lower loop length; but, rather statements about big trucks vs small cars and only a $100 difference in operating cost that sound good but do not deepen the discussion.

 

Yes heat extraction is determined by measuring delta-T and flow of the loop fluid, but I am confused by your statement that the capacity of a given fluid to transport btus is enhanced by adding methanol (I could get sidetracked on the wisdom of using such a poisonous chemical as methanol for antifreeze, what is your mitigation protocol when all of it ends up in the aquifer from a catastrophic failure, not likely but possible, but will restrain myself from further comment on it since from a strictly performance stand point it is the best of the antifreezes used.) It will allow for a greater total of btus removed over time from the same loop, but at reduced efficiency and extended operating time for the same load. The shorter the loop, the worse the efficiency and capacity of the unit becomes.

When antifreeze of any type is added to the loop fluid at least two things happen, viscosity goes up and density goes down when compared to water. This means that more energy is required to move the same amount of fluid, less btus are transported by the liquid and significantly greater flow rates are required to obtain the same turbulence in the pipe.

Colder loop temperatures allowed by antifreeze, which is what I assume you are referring to when you talk about greater delta T between the brine and soil, reduce the heating capacity of the geo unit, reduce the COP of the unit, increase the pumping power required to move the same quantity of fluid (or the same pump will move less fluid, lowering the heating capacity and COP even more)

If a person is taking out a 30 year mortgage, then the extra cost for the extra looping is being paid off over 30 years, so the calculations I have done for clients show the energy savings exceed the extra mortgage cost from DAY ONE. This means that the shorter loop is actually costing the homeowner money every day they own the house. If they use their own money to invest in the looping instead of another investment the dividends exceed anything on Wall Street and are guaranteed, going up over time.

 

I was referring to the heat gathering capacity of the loop -- and the
delta-T between the loop fluid (brine) and the rock/soil surrounding the
underground pipe. Heat flow into the loop is directly proportional to the
temperature difference between the brine and the soil -- and without
antifreeze, that difference can be no larger than 15°F to 20°F (for an
undisturbed ground temperature in the low-mid 50's).

Mitigation policy: Same as "War Of The Worlds" -- bacteria in soil
or groundwater croak the evil MeOH invaders in a matter of days.

Some true, some not -- but none of the adverse effects are large
enough to justify the cost of extending the loop enough to avoid
antifreeze.

1) The added length increases pumping power requirements
more than the small viscosity-related penalty due to methanol.

2) With antifreeze, more BTUs (per gallon and per minute) can
be transported -- because the wider range of brine temperatures
allows a larger temperature rise (EWT - LWT) in the loop. Just
do the math -- the larger delta-T absolutely swamps the <5%
reduction of brine factor (a.k.a. volumetric heat capacity).

Quite possibly true, even with a damn-the-expense Mercedes-Benz
approach to design.

Even truer with a reasoned cost/benefit approach with appropriate
compromises to optimize the total life-cycle cost of ownership.

 

Looby,
The ground around the loop gets colder as the energy is extracted to heat the house, the difference in the ground temp and loop temp is a factor of the resistance to the flow of heat. The delta T between the ground and the loop fluid does not go up as the loop temp drops, except during a heating call. In between the loop and ground delta T stabilizes. The temperature effect to the ground reduces as the distance to the loop increases and heat flows from further away to try to equalize the temperatures, but under high loads the draw out is faster than the flow in and the loop field temp decreases. When multiple wells are involved, they interfere with each other and the closer the wells are to each other the colder the loop field will be for any given load. (I know this is basic IGSHPA training but less trained people are also reading our posts) The conductivity of the soil is also a factor governing how fast the heat moves in from the surrounding soil.

I agree that the antifreeze allows more heat to be extracted from a well, but there is no question that as the loop temps drop less heat per minute is extracted.

I would love to see the studies showing that existing ground bacteria quickly convert methanol. Is this in all ground conditions? I have avoided using it when I need to use antifreeze because of its toxicity.

The added length might increase the pumping power, but not if the pipe size is adjusted to reduce pressure drop. I use 1" pipe for 300' wells and 1 1/4" for 500' wells. And if more wells are added to get the longer loop, then the flow is split and pressure drop decreases so the pumping power would go down.

The btu capacity of the loop fluid goes down as antifreeze is added, on a gpm and time basis. The btu calculation is 500 x gpm x delta T for water and 450 x gpm x delta T for antifreeze. (yes the 450 is average but the number is always less for antifreeze than for water) For a given flow rate and ground temp the delta T for water will be greater than for antifreeze due to much greater turbulence for water over antifreeze.

Each difference may be small, but added together the difference is much bigger than you are thinking.

 

The amplitude of short-term, near-field temperature fluctuations is
highly dependent on soil conductivity, heat capacity (diffusivity),
and groundwater movement (advection).

So, there may or may not be a significant temperature recovery
"in between" compressor cycles -- just as there may or may not
be rapid near-field temperature draw-down during calls for heat.
Of course, these also depend on the frequency of compressor
cycles and the duration of "in between" times.

However, regardless of what soil conditions you postulate,
a lower minimum brine temperature will always extend the
capacity of a given loop to deliver BTUs, over the long haul.

Moving the goal posts doesn't change the final score. Pipe diameter
adjustments have the same effect -- with or without antifreeze.

FWIW, my ND038 system with a single-circuit vertical loop has about
1000' of 1.25" pipe underground (450' bore plus 50' trench) filled with
20% V/V methanol -- with a single UP26-99 pumping about 8 gpm,
and yielding a very comfortable Reynolds number of ~5890.

It's kinda hard to undercut a single UP26-99 on pumping costs.

BTW, switching to pure water would allow me to add another 145'
of pipe to the loop (for the same p-drop across the loop field). How
much do you estimate that would raise my LWT? How many feet
of loop would I need in order to safely eliminate antifreeze?

BTW #2, WF suggests a brine factor of 485 for methanol. That's a
3% penalty compared to water -- so my "<5%" claim was generous.

Please show us the numbers.

.

 

The guy in NE Ohio went with

propane. LOL.

 

Exceptional!

 

Masoud,

Your wit is exceeded only by your great thinking and ability to deal with Joe Hardin. Thanks for being here.

 

Yes, you are right, the antifreeze allows lower operating conditions and for a given loop will allow greater btu withdrawal annually. It does not increase minute by minute capacity of the system to deliver heat. The colder the ewt the lower the COP and heating capacity of the unit. I am sure everyone agrees with this.

Resistance to flow is based on pipe, fluid, and gpm. Again, we can all agree to this statement. For the same fluid and gpm larger pipe will reduce resistance and turbulence goes down. For the same fluid, gpm and pipe size, more parallel loops in the system will reduce total flow resistance and lower turbulence due to lower gpm in each loop. For the same pipe and gpm adding antifreeze will increase flow resistance and turbulence will go down and thermal carrying capacity will drop due to the change of viscosity and density of the fluid. The colder the loop fluid, the more the flow resistance with antifreeze becoming more viscus and flow rates will go down and ability of the loop to deliver btus to the geo unit will go down.

All of these numbers can be found in the performance charts provided by manufacturers.

 

I have enjoyed the give and take here, but I think that we should agree to disagree.

I think the primary point of contention is whether to size for 80% of the design load or 100%. We have all stated our opinions on this and have gotten into deep discussions on it. I doubt that we will change each others minds on this. Your opinion is shared by many, perhaps most designers. I have a different opinion.

I believe that people buy geo to save money on heating and cooling their homes and I have found my clients willing to spend more money on loops to save operational cost.

I think the real difference comes down to design priorities. I design the most efficient system I can. Others design functioning systems that contain installed cost. These are both defensible positions. I suggest asking your customer if they want to spend the extra money to enhance the performance, most will perceive the value. Of course if you do not see any value in this, then it would be pointless to suggest it to your customer, but another contractor might and win the job.

The way I contain cost is to reduce load. When a potential client contacts me about geo, I first talk to them about the house and how to reduce the load. In this way I reduce the cost of the geo while meeting 100% of the load. Generally the unit size is reduced significantly, a 2 ton unit in a 3500 sq ft house is common, then I put a 500' loop in and use water. The final result is a less expensive system to install and operate. The improved comfort and indoor air quality are bonuses that many appreciate more than the energy efficiency.

 

Mark,

Thank you for the compliment. I’m not used to hearing compliments, hence the blush! Knowing Joe has been a pleasure. Good to be here.

Regards,

Masoud

 

.
Agree to disagree, aye.
.

 

Yea !!! a peace treaty has been accepted.

Lloyd...welcome to the forums. Your expertise is a welcome addition.

 

So it seems that you were off on the first house by a factor 2. With 98% of the load I mean total load, not peak load. It is not just an opinion, and not just for poor performing buildings. True, tight houses have a different sensible and latent cooling load than leaky once. But the disadvantages of oversizing do not go away. Welcome to the forum.

 

You do not seem to understand that a system designed for 80-90% of the peak load will be designed for 99% of the total load. If a loop is correctly designed for a min temp of 30F, even with a one week below normal cold period, loop temps should not drop below 27-28F, and loops do not freeze. Plus the system will run much more efficient the remainder of the season, if the equipment is smaller. Most of our systems use $20 worth of supplement heat per year, but save about $50-70 in running costs compared to the next size. And did I mention the $3000 less in upfront costs...

 

This is getting funny now. No antifreeze in NY state, no wonder your loops freeze up as you mentioned before! But even if you you double you loop size, and you would be able to increase your EWT by lets say 5 degrees, that would increase your heat output by maybe 5%, probably less. I am running a 10 ton loopfield on a 6 ton heat pump at my house, the EWT is about 4-5 degrees higher, but LWT is 28 degrees now. It would be frozen shut!
I can see your point that a few percent more heat output can mean a lot for a close to passive house, I also see that they benefit so much more from solar gain, etc etc,but sizing loops to use no antifreeze in NY State, beats everything everyone here knows about geosystems...! The increase in pumping power would wipe away any benefit one might have from a higher EWT. Don't get me wrong, a had good discussions with Joe about loop sizing, and I size systems larger than he does, but I am also paying 15 cents/kwh. You have a point that larger units with more cycles might allow ground loops to recover better (but more data is needed to show this).

 

I think peak load has to be defined. I made a mistake in using it. I should have stayed with design load, an used peak load instead of design load, Design load is based on calculations made using temperature bin data for the location and 98 to 99% of the recorded temperature bins are used in manual J to calculate design load. For those who are even more confused, the historical temperatures recorded at a site, such as an airport, are divided into temp ranges of 5F, each being a bin. Then the number of hours the outdoor temperature is within each of these 5F bins is calculated and this bin data becomes the basis on which heating and cooling loads are calculated.

My loops never freeze up. I was perhaps not clear enough. I use 40F as my low loop ewt calculation point. I would not do this with a standard built house as the loads below design temp would likely freeze up the coil when the unit was running 100% of the time.

I have years of experience dealing with high performance homes and it was this experience that led me to use no antifreeze. The two big differences in heat flux and load with high performance homes is the balance point, the outside temperature where a building starts to need heat, is lower (a "standard house has a balance point of about 60F and a real high performance home has a balance point <50F) and the <design load peak is smaller. The pumping power is less, not more than when using antifreeze and smaller loops. I get great flow rates and system performance.

There is a performance gain with longer loops because the ewt will always be higher in heating the longer the loops are. There is no such thing as a loop that is too long, just too short.

What I think people are falling to understand is that an even more important consideration than COP is heating capacity. Look at the difference in heating capacity between 20F and 40F for any geo unit. As the ewt gets colder the capacity of the system drop, as the heating capacity drops the unit has to run longer to deliver the same number of btus. One of the interesting things that happens during a cold spell is the loop temps drop, which means the system delivers less heat per minute, so it runs longer to meet the same load.

The house with the 2X oversized geo unit has GREAT efficiency, the geo system has its own electric meters and we know it uses less electricity than calculated . It was this house that allowed me to understand how easy it is to over estimate load with manual J and a high performance home. I adjusted the infiltration rate by using blower door data, which helped, and then made other corrections in the calculations.

There is a real misunderstanding of efficiency and larger (over) sized geo units. They are not oil boilers or 10 SEER A/C units and do not perform like them. The use of two stage units changes the picture even more, adding dehumidification control on the ECM fan speed changes the picture even more. Then add in the vastly different way a true high performance buildng works and everything you think you know about "over sizing" changes. I honestly have extensive experience with many net zero houses and so have field experience to back me up. The house with the 4 ton geo and 500' of vertical loop is just as efficient as the house with the 2 ton geo and 500' of looping. The difference in cost between the 2 ton and 4 ton system, installed, was about $500. The loop length has more to do with system efficiency than the size of the unit. Change the ewt and you change the heating capacity and run time, it is as simple as that.

 

Okay Lloyd, again this is just too much you are now suggesting the installed cost difference between a 2 and a 4 ton unit is $500?....and 500' of vertical loop is okay for a 4 ton?
Lemme ask ya,
since your end point was "the 4 ton was just as efficient as the 2 ton" why buy the more expensive unit?

You are percieving these as attacks, but your assertions are boardering on ridiculous. A 4 ton in my neck of the woods would have ~ 600' of vertical bore or 1,200 feet of loop. Since I am amongst the unwashed without a CGD perhaps you can enlighten me as to what is so mild in your climate that you can use 40% of the loops we would employ in MI and be significantly oversized ("designed for 40* EWT").

j

 

If the only difference between two jobs is the unit (notice that both have 500' of vertical well) then the only cost difference is the very little increase in price between the two units and the up size of the ductwork (no increase in number of branch runs) then the cost difference is $500. Also it appears that I am using a different meaning, when I say 500' of loop I mean that the vertical well is 500' deep. I apologize for that. I should be more clear.

What I see as attacks is when you say things like my "assertions are boardering [sic] on ridiculous." But I can take it, I just see this type of statement as unproductive and think that the public forum needs to be free from this type of statement, but do as you see fit.

As you know, the conductivity of the soil is the key to sizing the loop field. Our loops go into dense rock. The loop has to match the load of the house and the unit, working together as a system. In the first instance 15% of ethanol is used, and the unit was sized for 100% of the design load at 30F ewt. In the second water alone was used and the unit sized for 100% of the design load at 40F. The close monitoring of the first (and second, third , fourth) house allowed us to adapt the system design to the houses as they were performing, not as the manual J for the first house predicted. I was able to adjust the manual J so its prediction was closer to the actual. Changing the infiltration rate to the blower door measured rate of the completed and tested houses helped a lot.

Remember, these are functioning houses and geo systems and we have several years of operation to look at. Also, because they are functioning net zero energy houses, the value in never needing electric heat as supplemental is a smaller solar PV system. If you think goe is expensive, look at the cost of PV. If my systems were using electric supplement to heat then they would not be net zero buildings.

As I said in an earlier post, I have a different design concept than you and most others do. I am not saying yours create non functioning systems. I am saying my designs work.

I do know that the only time a client complains about the cost to operate the geo system is when the house has a higher load than they expected, due to failures in the air and thermal control layers. In these instances the geo has enough reserve capacity to still perform reasonably well, but they do have to have 10-15% ethanol added to extend the range and supplemental heat is used a little. You could say that the system ends up where yours start.

What do you do when the house performs worse than you predicted? I would think that when the system is designed so the geo runs 100% of the time when the outside temp is 20F above your design temp there is no reserve for a poorly functioning house and the unit might even lock out on low temp during intense cold spells.

 

Sorry, I tried to observe the armistice, but this is just too much.

Please explain: how it is possible for two systems with identical
loops and identical loads to be "designed" for different EWTs?

If, as you claim, the 4-ton system is "more efficient," that clearly
implies that, over a heating season, it is extracting more BTUs
from the loop field, and thus, it must depress the far-field ground
temperatures more than the "less efficient" 2-ton unit.

If, as you have claimed elsewhere, the 4-ton system will have
shorter run times than the 2-ton system, this clearly implies
that BTUs must be extracted more rapidly during the shorter
cycles, and thus, the 4-ton unit must also depress near-field
temperatures more rapidly and more deeply than a 2-ton unit.

Also, since the larger unit extracts BTUs more rapidly, you
must increase loop flow rates (and pumping costs!) and/or
accept a lower LWT ...but remember, NO ANTIFREEZE!

Sorry, but your alleged "design" is over-constrained. Once you
specified the loops and loads as being identical, you have no
remaining degrees of freedom with which to control EWT --
except for efficiency. However, contrary to your description,
the more efficient unit must produce lower EWTs.

...not just a good idea, IT'S THE LAW!

Looby
.