Geothermal heat pump to preheat propane water heater water?
I have an older house that was built in the early 60s. The first floor has poured concrete foundation walls on the rear and sides of the house. One side and the back of the house is 5′ into a soil… one side is about 1.5′ into the soil and the front is at grade. When we bought the house it had a heat pump that ran the primary cooling and heating needs. It runs forced air through ducts in the attic and down to the first floor using ducting through some closets. This system is on its last leg. It also had a propane boiler that circulated water through 2 zones, an upstairs and downstairs zone. The water is ran through baseboard heaters. It also had a weird wood stove system that forced air through pipes in the flue hood on the first floor for heat. We have replaced this with an energy efficient wood stove.
I have ran some analysis of the existing structure and the future structure as we update insulation. We need to be able to deliver 36,000 btu/hr – first floor and 18,000 btu/ hr – second floor during the coldest days. The second floor demand could go up as well since we have enclosed a porch on the second floor that was above the basement family room. We may decide to heat it during the winter. If we do, it would require an additional 13000 btu/hr (It has a lot of windows).
So right now I need to make a decision on the HVAC equipment that should be installed. My current plan is to install a geothermal heat pump with loops through a large pond on the property. I will run water through this system and into a propane water heater to add the additional heat needed for baseboard heater. Based on the hydronic calcs I ran it looks like water needs to be 160F during the coldest days and 120F on 35F days. I will use the wood stove when possible and turn off that zone when it is running. During the summer the geothermal should perform well since the water at the bottom of the pond is pretty cold… I will send the heated water into the water heater and potable hot water will run through an internal heat exchanger… Is this overly complicated? Or is this a good approach?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
At what outdoor temp is the whole-house load ~54,000 BTU/hr? How confident are you in that number?
That would be a very high load number at 0F outdoors, 68F indoors for a reasonably tight, reasonably insulated 2000 square foot 2x4 framed house, perhaps not super-high for a 3000 square foot house. But at a design temp of 25F it would be high even for a 4000' house.
For reference, my circa 1923 2x4 framed sub-code ~2400' house + 1600' of insulated conditioned basement (~4000 square feet total, 2400' fully above grade) doesn't see a heat loss of 54K until it's about -23F/-30C outside.
It's probably simpler, cheaper, and more effective to just use a heat pump water heater than cobbling up a heat exchanger to save on propane use.
User-6999468,
First of all, can you tell us your name? (I'm Martin.)
Ground-source heat pumps need to be individually engineered and site-built by a variety of contractors cooperating on site. As a result, these systems are very prone to design and installation problems. A few years back, a homeowner with a ground-source heat pump connected to a heat-exchange loop in a pond reported to GBA that they suffered years of problems, with no resolution from contractors.
Moreover, ground-source heat pumps are expensive. More information here: Are Affordable Ground-Source Heat Pumps On the Horizon?
My advice is to focus an air sealing work and insulation improvements. Hire an independent energy consultant to perform an accurate Manual J -- your estimates seem high, as Dana wrote.
You should be able to design a simple heating system using miniplits (a combination of ducted and ductless units) and you should be able to heat your domestic hot water with a heat-pump water heater or a conventional electric water heater. This approach will save many thousands of dollars compared to a ground-source heat pump.
In this case when earth coupled heat exchanger is a pond loop (as opposed to drilling wells or trenching) the economics of GSHP can be favorable. But that doesn't mean you can just oversize everything by 2x. The Manual-J and plucking any low hanging fruit on the building envelope upgrade path will still pay for itself. Hooking up a 5 ton heat pump to serve a 2.5 ton peak load won't run nearly as efficiently as a more appropriately sized heat pump, and it'll cost more up front.
Vetting the contractor's competence is still necessary to limit the financial risk, but that's true with mini-split solutions as well. In some locations there are multiple GSHP contractors and good support within a half hour drive, but that's not the norm, which makes the homeowner much more dependent upon the individual installer for support.
Disaster stories of grotesque oversizing, poor performance and worse support aren't hard to find for any type of HVAC equipment, including standard split AC + gas furnaces or hydronic boilers. eg:
Earlier this year I talked a guy in Texas out of moving forward with a 60KBTU furnace + 100KBTU of propane furnace on a 2400' 2-story (one fore each floor) and 2-3 tons of AC per floor in a house (less than 2 decades old as the "solution" to his comfort issues with the pre-existing merely ridiculously oversized resistance electric hot air + AC equipment. The estimated heat load for the whole house was under 40,000 BTU/hr, and the cooling load about 2 - 2.5 tons based on a cursory napkin analysis of his utility bills. (And getting the ducts better balanced &/or out of the attic probably would have brought it under 30K heating/ 2 tons.) This place screamed for a a single 3 ton- 2-stage heat pump solution using the original ducts (or a pair of 3/4-1-ton ducted mini-splits with some duct modifications). I'm not sure what was eventually installed, or where the numbers would have come out of a formal Manual-J (assuming he followed up on that advice), but I'm pretty sure he didn't stick with the original proposal. This fell out of a question on how to provide enough make-up air to the ludicrously oversized propane furnaces. The financial hit for that looming disaster wouldn't be as large as oversizing a pond loop GSHP by 2x, but it's the same order of magnitude.
Hydronic boiler system design and oversizing or micro-zoning disasters also abound, sometimes costing as much to rectify as fixing a lousy pond-loop GSHP implementation.
120F-160F is not a good fit for a GSHP. It doesn't matter much that you are proposing to post-heat with propane to get it that high - it's a closed system and the return water will likely be too hot for good GSHP efficiency.
As Martin said, air seal and insulate and then consider ASHPs plus some supplemental heat from your existing systems.
JON R: Good point, though I suspect replacing the baseboards with cheap high-output low-temp tempc convectors could be cheap enough. Somehow I missed the 160F peak temp requirement, and was assuming 120F max (which can be done, but not at great efficiency.)
A cheap electric boiler backup utilizing the existing baseboard and a few mini-splits, getting rid of the fossil burners altogether would work.
Please do your math very carefully without the now expired geo thermal tax credit most people cannot get the numbers to work for them.
When you do you math make sure you account for the power to run the water pump as the manufactures seem to leave that power out of the calculations. They often do not make that part of the system and it is sight specific so they feel justified in pretending no one will pay for that electricity.
Walta
Walter: The tax credit for ground source heat pumps got restored in the 2018 budget deal. See the bottom paragraphs under the heading "Section 25D Tax Credit for Power Production Property"
http://nahbnow.com/2018/02/energy-and-home-owner-tax-credits-included-in-budget-accord/
Pumping power is indeed a significant part of the efficiency equation that the system designer needs to pay attention to. ECM drive pumps help, but pumping head at the required flow rate is the major factor, and is in the control of the designer. There are many GSHP systems out there running at lower seasonal efficiency than cold climate ductless mini-splits.
Thank you all for the feedback, I am going back through and looking at my calculations for the heat load. One of the difficulties with my first floor is determining the loss of heat through the soil. My ideal situation would be to run baseboard heaters via just a geothermal heat pump and not utilize the a secondary water heater. I need to double check the return temperatures on the hydronic calcs that I was running as well.
You wrote, "My ideal situation would be to run baseboard heaters via just a geothermal heat pump."
To me, that approach sounds far from "ideal." But it's your house, of course.
Hi, Thanks again for all of the feed back.
I went back and looked at my old calcs and realized I was looking at the initial house conditions spreadsheet instead of the sheet after the remodel. that is why the downstairs numbers were so high.
It had concrete walls, furring strips and paneling. This has all been remodeled except in one area that is currently being remodeled. The new insulation is 1.5" of Polyisocyanurate, and 3.5" of fiberglass and drywall.
In the remodel scenario which is what I want the HVAC designed for I have the following calcs:
Manual J says I need 31,000 BTU/hr at 5F
My Calcs gave me: 40,000 BTU/hr at 5F (I may have picked too conservative soil temperatures)
I calculated an additional 17,000 BTU/hr if I decided to run heat through my 4 season room. I likely won't heat this area with the main system and instead may install a small wood stove.
My original thought was hydronic for winter since there is a system that was used prior and the system is much quieter and use the ductwork for ac. I was wanting an energy efficient system and with the pond close it seemed like a good oppurtunity to do the geothermal. What would you all recommend?
Fin tube baseboard has non-linear less predictable output at temps low enough to be efficient for a ground source heat pump even if you sized it right.
A modulating 2.5 ton cold climate air souce heat pump can cover the Manual-J load using standard types of duct systems, eg:
https://www.greenbuildingadvisor.com/sites/default/files/PVA-A30AA7___PUZ-HA30NHA5_Product_Data_Sheet-en-1.pdf
But there are numerous ductless and mini-ducted possibilities that could still work (possibly more efficiently than the big air-handler versions) at bigger modulation ranges/higher comfort.. A 1.5 ton Fujitsu ARU/AOU 18RLF mini-duct cassette will deliver about 2/3 of your design load at +5F, a pair of 1 ton ARU/AOU 12RLF mini-duct units would more than cover the load, good for 15,000 BTU/hr each even at -5F.
http://www.fujitsugeneral.com/us/resources/pdf/support/downloads/submittal-sheets/12RLFCD.pdf
But there are many ductless options that could potentially work too. What makes sense depends a lot on your floor plan, and how easy it is to run ducts. What DOESN'T make sense is the "mini-split head per room" approach, which leads to high up front cost, grotesque oversizing, lower comfort, lower efficiency.
The ducts are already present in the attic... Although I likely will add better insulation to them. I also have access to the floor joists currently so that is another option during this remodel. The upstairs has 4 rooms (3 bedrooms and a open living room) The downstairs has a large great room, a bathroom and utility room and a large open kitchen and dining area. With the mini splits would we have a totally separate unit in each location or is there just a refrigerant line ran to the unit? How does it handle condensation?
Thanks a lot!
Ducted and ductless mini-splits come in both modulating single zone systems (a single head/cassette per outdoor compressor unit) and non-modulating multi-zone systems (one larger outdoor unit serving multiple indoor heads.)
Since the heads don't really modulate when on a multi-zone system getting the sizing right is critically important for both comfort and efficiency. With modulating single zone systems there's some wiggle room, but you still want to size it so that it is modulating it's speed rather than cycling on/off most of the season. Modulation ranges are not infinite. What makes sense depends upon the floor plan, the room by room load numbers, etc.
I'm not sure what is meant by "How does it handle condensation?" Do you mean cooling season condensate at the cassettes? If yes, there are condensate drain ports at each indoor unit, that can sometimes be routed along the same path as the refrigerant lines, sometimes it has to be routed differently- it just depends. It's good to set up detectors to alert you when a drain port gets plugged (standard procedure in my area, but the detectors are not part of the original equipment.)
These are all part of the considerations to address when looking at ductless solutions.
Take the floor plan and mark it up with the heating and cooling loads of each room. A typical bedroom has design loads WELL below the nominal capacities of even half-ton ductless heads, but there are often ways to serve 2-4 adjacent rooms with a single 3/4 ton mini-duct cassette, with the ducts still inside the insulation & pressure boundary of the house (not in the attic, above the insulation.) The cassette can usually fit under the ceiling in a closet, with a very short plenum to duct serving nearby rooms, which could also be routed under the ceiling in a closet, or in soffits, etc. Fujitsu's units can even be mounted vertically (most can only work horizontally, due to the condensate management issue). Here are a couple shots of a vertically mounted Fujitsu 1.5 tonner serving a whole house in California with 4 ducts, in a vertical orientation with a "utility cabient/closet) taking up only about 7 square feet of space. Note the soffit that was added around the duct running down the hall. The big grille is the return plenum. Doored off rooms would need jump ducts or other return paths to the common hallway space to work without pressurizing/depressurizing different parts of the house: