ᐅ Air Source Heat Pump vs. Geothermal System for a New 4,300 sq ft House
Created on: 16 Jul 2016 14:55
M
markus-db
Hello forum members,
I have already read quite a bit in the discussions about ground-source and air-source heat pumps, but most of the projects were in the range of 150–200 m2 (1600–2150 ft2). Since we are planning to build somewhat larger, I am wondering if this changes the considerations, so I created this thread.
Here’s a brief summary of the key data for the building project:
- Project on the outskirts of Berlin
- New single-family house, not KFW 55 standard (because according to the energy consultant, a ventilation system costing over 20,000 € would be required, which we do not want)
- Exterior walls made of aerated concrete 36 cm (14 inches) thick, no additional insulation
- No fireplace planned or desired
- About 300 m2 (3200 ft2) heated living area from ground floor to attic (GF, 1F, attic)
- About 100 m2 (1075 ft2) heated basement area
- Total heated area: 400 m2 (4305 ft2)
- Heating exclusively via underfloor heating (basement to attic, fully planned)
We decided against a gas heating system because although this might currently be economically more favorable than a heat pump (of any type), sustainability is important to us and we are willing to accept higher costs (especially investment costs).
With our energy consultant, we developed two options:
Always included is a photovoltaic system (nominal capacity about 4.5 kWp) with a buffer tank (probably around 7.5 kWh) – the "idea" is to generate electricity for the heat pump ourselves. Of course, this will not cover the full heat pump demand (see below), but even meeting about 50% annually would help. (Excess electricity produced in summer would be sold.)
Option a) Ground-source heat pump: According to the heating load calculation, we need a system with about 18 kW output and boreholes totaling 440 m (1443 ft) depth in this area. This is divided into 5 boreholes of 88 m (288 ft) each.
We have an offer, but unfortunately, no itemized pricing. Without going into too much detail (since the forum is not for assessing my offer), the total cost for the boring works, a Vaillant heat pump with 19.7 kW / COP 4.7 (standard not specified), 300 l buffer tank, 500 l domestic hot water storage, plus all additional costs comes to 42,500 € gross.
Option b) Air-source heat pump: Here, obviously, no drilling is required. The heat pump is a Heliotherm model with 18.5 kW / COP 4.14 (A2/W35) in fully modulating operation, hydraulically decoupled connection, buffer tank and domestic hot water storage similar to option a). Total cost: 25,000 € gross.
(There are additional costs of about 18,000 € for around 400 m2 (4305 ft2) of underfloor heating, and 15,000 € for the photovoltaic system, but these are the same for both options and are therefore excluded.)
Summary of the heating system costs:
- Ground-source heat pump: 42,500 €
- Air-source heat pump: 25,000 €
- Difference: Ground-source heat pump is 17,500 € more expensive upfront
- Air-source heat pump is less efficient than ground-source, especially in winter when it is needed most, so ground-source is cheaper in ongoing energy use
I have the following questions for the experts:
- Do these considerations make sense overall?
- Is it a good idea to partially cover the heat pump’s electricity demand with a self-generated photovoltaic system?
- From your perspective, is the extra investment of 17,500 € for the ground-source heat pump option worthwhile (also considering the heating load of about 18 kW)?
- What other factors should I be paying attention to?
The overall goal is to implement a sensible but as sustainable as possible energy system for the future, without incurring unnecessary costs (unfortunately, I have not won the lottery and funds are limited).
Sorry for the long post, but I wanted to be as precise as possible. If you need more information, please let me know.
Thank you!
I have already read quite a bit in the discussions about ground-source and air-source heat pumps, but most of the projects were in the range of 150–200 m2 (1600–2150 ft2). Since we are planning to build somewhat larger, I am wondering if this changes the considerations, so I created this thread.
Here’s a brief summary of the key data for the building project:
- Project on the outskirts of Berlin
- New single-family house, not KFW 55 standard (because according to the energy consultant, a ventilation system costing over 20,000 € would be required, which we do not want)
- Exterior walls made of aerated concrete 36 cm (14 inches) thick, no additional insulation
- No fireplace planned or desired
- About 300 m2 (3200 ft2) heated living area from ground floor to attic (GF, 1F, attic)
- About 100 m2 (1075 ft2) heated basement area
- Total heated area: 400 m2 (4305 ft2)
- Heating exclusively via underfloor heating (basement to attic, fully planned)
We decided against a gas heating system because although this might currently be economically more favorable than a heat pump (of any type), sustainability is important to us and we are willing to accept higher costs (especially investment costs).
With our energy consultant, we developed two options:
Always included is a photovoltaic system (nominal capacity about 4.5 kWp) with a buffer tank (probably around 7.5 kWh) – the "idea" is to generate electricity for the heat pump ourselves. Of course, this will not cover the full heat pump demand (see below), but even meeting about 50% annually would help. (Excess electricity produced in summer would be sold.)
Option a) Ground-source heat pump: According to the heating load calculation, we need a system with about 18 kW output and boreholes totaling 440 m (1443 ft) depth in this area. This is divided into 5 boreholes of 88 m (288 ft) each.
We have an offer, but unfortunately, no itemized pricing. Without going into too much detail (since the forum is not for assessing my offer), the total cost for the boring works, a Vaillant heat pump with 19.7 kW / COP 4.7 (standard not specified), 300 l buffer tank, 500 l domestic hot water storage, plus all additional costs comes to 42,500 € gross.
Option b) Air-source heat pump: Here, obviously, no drilling is required. The heat pump is a Heliotherm model with 18.5 kW / COP 4.14 (A2/W35) in fully modulating operation, hydraulically decoupled connection, buffer tank and domestic hot water storage similar to option a). Total cost: 25,000 € gross.
(There are additional costs of about 18,000 € for around 400 m2 (4305 ft2) of underfloor heating, and 15,000 € for the photovoltaic system, but these are the same for both options and are therefore excluded.)
Summary of the heating system costs:
- Ground-source heat pump: 42,500 €
- Air-source heat pump: 25,000 €
- Difference: Ground-source heat pump is 17,500 € more expensive upfront
- Air-source heat pump is less efficient than ground-source, especially in winter when it is needed most, so ground-source is cheaper in ongoing energy use
I have the following questions for the experts:
- Do these considerations make sense overall?
- Is it a good idea to partially cover the heat pump’s electricity demand with a self-generated photovoltaic system?
- From your perspective, is the extra investment of 17,500 € for the ground-source heat pump option worthwhile (also considering the heating load of about 18 kW)?
- What other factors should I be paying attention to?
The overall goal is to implement a sensible but as sustainable as possible energy system for the future, without incurring unnecessary costs (unfortunately, I have not won the lottery and funds are limited).
Sorry for the long post, but I wanted to be as precise as possible. If you need more information, please let me know.
Thank you!
M
markus-db3 Aug 2016 17:43Saruss schrieb:
I still consider 18 kW to be too high. If you can provide some details about the aerated concrete block, window areas, and geometry, I can roughly estimate whether that could be accurate.Thanks for the offer.
Here are some more detailed specifications I can provide now:
- All exterior walls are masonry made with Ytong, 36.5 cm (14 inches) thick, including in the basement (black tank waterproofing)
- The geometry is mostly cubic, except for the ground floor where the house has an L-shape (about 28 m² (300 sq ft) added to a cube of approx. 130 m² (1,400 sq ft))
- The house is now fully basemented, also with the L-shape
- On the upper floor, there is a roof terrace in the L area with approximately 28 m² (300 sq ft)
- The attic is purely cubic with a hipped roof
I have now obtained the corresponding U-values and calculated the envelope areas myself (which almost match the heating load calculation, so that is at least OK):
- 154.40 m² (1,661 sq ft) with U-value 0.21: basement floor slab
- 150.80 m² (1,623 sq ft) with U-value 0.18: exterior basement walls (net, excluding windows and doors)
- 227.40 m² (2,448 sq ft) with U-value 0.21: exterior walls ground floor and upper floor (net)
- 174.66 m² (1,880 sq ft) with U-value 0.17: attic / roof (net)
- 118.30 m² (1,273 sq ft) with U-value 0.86: window areas (mostly safety glass)
- 10.30 m² (111 sq ft) with U-value 0.7: exterior doors (U-value estimated here, no data available)
- 28.02 m² (302 sq ft) with U-value 0.24: roof terrace areas on the upper floor
In total, this results in roughly 860 m² (9,260 sq ft) of building envelope and a naive calculation yields an average U-value of about 0.29 W/(m²·K) (the heating load calculation states 0.34 W/(m²·K))—a difference of about 15% (compared to the envelope area). Unfortunately, the heating load calculation only provides an average U-value, so I currently don’t know how that difference arises.
Do you think the above U-values are standard, or does any of them fall outside the typical range? I have no experience with this so far...
However, I’m unsure how to get from the envelope area and U-values to the heating load. As far as I know, you also need the air exchange rate, which is assumed as 0.6/h in the heating calculation, as well as the minimum outside temperature, assumed to be –14°C (7°F). The indoor temperature is 20°C (68°F), and the heated net volume is 1230 m³ (43,435 ft³), gross volume 1620 m³ (57,200 ft³).
Lots of numbers to keep track of...
The U-values are generally acceptable. In my case, the windows and walls are better insulated, while the walls and floor slab in the basement have poorer insulation.
Using these data, the transmission heat losses amount to about 7.5 kW at the design temperature (plus a bit more for thermal bridges, depending on build quality, so roughly 8 to 8.5 kW maximum). The calculation is straightforward: area × U-value × temperature difference (the ground temperature stays above the design temperature). With a building air exchange rate of 0.6/hr and the mentioned volume and design temperature, the ventilation heat loss is about 8.5 kW without heat recovery. However, I consider this value unrealistic since ventilation does not happen based on volume, but on actual air consumption, i.e., the number of occupants. If 5–6 people live inside, I would estimate the ventilation heat loss to be a maximum of around 4 kW (just think about how much hot air a 4 kW forced-air heater produces... that’s roughly how much air you need to exchange—and when it’s cold, even a small air exchange quickly reduces humidity).
For domestic hot water, you can assume a power demand of 1–2 kW. The design temperature does not apply continuously, and with a total around 14 kW, you can quickly heat plenty of water (and if more heat is needed or it’s colder, an electric heating element in the heat pump covers the difference for those few hours).
In my opinion, the calculation you have is already reasonable. However, for the ground-source heat pump, I would choose one in the range of 14–15 kW. For the air-source heat pump, I would be less certain and would try to find a model that provides slightly higher capacity but can modulate (so that efficiency and performance remain good at low temperatures, but no excess power is produced at higher temperatures, which would lead to inefficient operation).
Using these data, the transmission heat losses amount to about 7.5 kW at the design temperature (plus a bit more for thermal bridges, depending on build quality, so roughly 8 to 8.5 kW maximum). The calculation is straightforward: area × U-value × temperature difference (the ground temperature stays above the design temperature). With a building air exchange rate of 0.6/hr and the mentioned volume and design temperature, the ventilation heat loss is about 8.5 kW without heat recovery. However, I consider this value unrealistic since ventilation does not happen based on volume, but on actual air consumption, i.e., the number of occupants. If 5–6 people live inside, I would estimate the ventilation heat loss to be a maximum of around 4 kW (just think about how much hot air a 4 kW forced-air heater produces... that’s roughly how much air you need to exchange—and when it’s cold, even a small air exchange quickly reduces humidity).
For domestic hot water, you can assume a power demand of 1–2 kW. The design temperature does not apply continuously, and with a total around 14 kW, you can quickly heat plenty of water (and if more heat is needed or it’s colder, an electric heating element in the heat pump covers the difference for those few hours).
In my opinion, the calculation you have is already reasonable. However, for the ground-source heat pump, I would choose one in the range of 14–15 kW. For the air-source heat pump, I would be less certain and would try to find a model that provides slightly higher capacity but can modulate (so that efficiency and performance remain good at low temperatures, but no excess power is produced at higher temperatures, which would lead to inefficient operation).
M
markus-db3 Aug 2016 20:43Thank you for the calculation. To make sure I understand it correctly:
So in my case:
560 * 0.29 * 37 = 6 kW (through the ground, temperature difference of 37 resulting from 20°C to -14°C (68°F to 6.8°F))
300 * 0.29 * 20 = 1.7 kW (under the ground, temperature difference of 20 resulting from 20°C to 0°C (68°F to 32°F) soil?)
In total, about the 7.5 kW transmission losses you mentioned. Is that the calculation?
Also here, so I can follow:
Are you assuming an air heat capacity of 1.2 kJ/(m³·K) (which is about 0.28 W/(m³·K))? Using the net volume of 1230 m³ (43436 ft³) and again the temperature difference of 37°, you get:
0.28 * 1230 * 37 = approximately 12.8 kW of energy needed to heat the air.
Multiplied by the building air exchange rate of 0.6 gives just under 7.5 kW per hour. You calculate 8.5 kW, so where am I mistaken? Is the heat capacity different?
And these two figures are roughly added together (including a buffer for thermal bridges), plus some extra for domestic hot water, and that is the "simplified method"?
So overall, there is little to do about the transmission heat loss, but a large margin in the loss due to air exchange rate (as you also mention).
I will ask the energy consultant where the 0.6/h value comes from.
If the brine heat pump really tends toward 14-15 kW, that could possibly save one of the five boreholes, and the price difference to the air heat pump would be smaller. That would be something and would simplify the decision significantly.
The air heat pump offer we have also includes a fully modulating device — so that is already planned.
Saruss schrieb:
Area * U-value * Temperature difference
So in my case:
560 * 0.29 * 37 = 6 kW (through the ground, temperature difference of 37 resulting from 20°C to -14°C (68°F to 6.8°F))
300 * 0.29 * 20 = 1.7 kW (under the ground, temperature difference of 20 resulting from 20°C to 0°C (68°F to 32°F) soil?)
In total, about the 7.5 kW transmission losses you mentioned. Is that the calculation?
Saruss schrieb:
A building air exchange rate of 0.6 per hour results, given the stated volume and design temperature, in a ventilation heat loss of 8.5 kW without heat recovery. However, I consider this value unrealistic because ventilation is based not on volume but on actual air consumption, i.e., the number of occupants. If 5-6 people live there, I would estimate the ventilation heat loss at a maximum of about 4 kW
Also here, so I can follow:
Are you assuming an air heat capacity of 1.2 kJ/(m³·K) (which is about 0.28 W/(m³·K))? Using the net volume of 1230 m³ (43436 ft³) and again the temperature difference of 37°, you get:
0.28 * 1230 * 37 = approximately 12.8 kW of energy needed to heat the air.
Multiplied by the building air exchange rate of 0.6 gives just under 7.5 kW per hour. You calculate 8.5 kW, so where am I mistaken? Is the heat capacity different?
And these two figures are roughly added together (including a buffer for thermal bridges), plus some extra for domestic hot water, and that is the "simplified method"?
So overall, there is little to do about the transmission heat loss, but a large margin in the loss due to air exchange rate (as you also mention).
I will ask the energy consultant where the 0.6/h value comes from.
Saruss schrieb:
In my opinion, the calculation you have is reasonable, but for the brine heat pump I would choose one in the 14-15 kW range, for the air heat pump I would be less certain and try to get one that offers slightly more capacity but can modulate
If the brine heat pump really tends toward 14-15 kW, that could possibly save one of the five boreholes, and the price difference to the air heat pump would be smaller. That would be something and would simplify the decision significantly.
The air heat pump offer we have also includes a fully modulating device — so that is already planned.
I calculated the heat losses as follows:
qm U heat loss
Exterior wall ground floor and upper floor 227.4 0.21 1623.636
Exterior wall basement 150.8 0.18 434.304
Base slab kg 154.4 0.21 518.784
Roof surfaces 174.6 0.17 1009.188
Windows 118.3 0.86 3459.092
Doors 10.3 0.7 245.14
Terrace area 28.02 0.24 228.6432
Each is A * U * ΔT. According to the standard, an additional factor of 0.05 to 0.1 for thermal bridges should be added, which would result in a total of about 8.7 kW. However, the standard does not always reflect reality, and I think the larger the building, the lower the surface proportion of thermal bridges (for example, where does a large window have a thermal bridge?). The temperature difference in my case is only from –14°C (7°F) to +34°C (93°F) to 20°C (68°F). You can of course use 23°C (73°F) as the indoor temperature (but we are quite comfortable with around 20°C (68°F); 23°C (73°F) feels too warm to me).
For the air, I calculated it this way (actually transformed and paying attention to units… you can remember it as a simple rule of thumb):
1 m³/h = 1/3600 m³/s
Density: 1/3600 m³/s * 1.21 kg/m³ = 0.0003361 kg/s
Specific heat capacity: 0.0003361 kg/s * 1010 J/(kg·K) = 0.3394 J/(s·K) = 0.34 W/K
So, 0.34 W ventilation power per degree Kelvin per 1 m³/h of ventilation loss.
1230 m³ volume * 0.6 = 738 m³ air exchange per hour, with ΔT of 34 K results in:
34 K * 738 m³ * 0.34 W/(K·m³) = 8531 W = 8.5 kW.
The 0.6 comes from a standard. The whole procedure is more complex overall (especially for the basement area, where depending on depth, you can calculate ground temperature and possibly specific heat transfer coefficients for the basement slab, since usually less heat is lost to the ground, or the heat partly remains ‘stored’ near the basement area in the soil, simply put. As a physicist, you know about nonlinear differential equations with boundary conditions, which could make this calculation more precise). The estimate I’m making here with temperature is still “rough,” and actually, less heat loss should occur.
Otherwise, you have basically repeated it correctly.
qm U heat loss
Exterior wall ground floor and upper floor 227.4 0.21 1623.636
Exterior wall basement 150.8 0.18 434.304
Base slab kg 154.4 0.21 518.784
Roof surfaces 174.6 0.17 1009.188
Windows 118.3 0.86 3459.092
Doors 10.3 0.7 245.14
Terrace area 28.02 0.24 228.6432
Each is A * U * ΔT. According to the standard, an additional factor of 0.05 to 0.1 for thermal bridges should be added, which would result in a total of about 8.7 kW. However, the standard does not always reflect reality, and I think the larger the building, the lower the surface proportion of thermal bridges (for example, where does a large window have a thermal bridge?). The temperature difference in my case is only from –14°C (7°F) to +34°C (93°F) to 20°C (68°F). You can of course use 23°C (73°F) as the indoor temperature (but we are quite comfortable with around 20°C (68°F); 23°C (73°F) feels too warm to me).
For the air, I calculated it this way (actually transformed and paying attention to units… you can remember it as a simple rule of thumb):
1 m³/h = 1/3600 m³/s
Density: 1/3600 m³/s * 1.21 kg/m³ = 0.0003361 kg/s
Specific heat capacity: 0.0003361 kg/s * 1010 J/(kg·K) = 0.3394 J/(s·K) = 0.34 W/K
So, 0.34 W ventilation power per degree Kelvin per 1 m³/h of ventilation loss.
1230 m³ volume * 0.6 = 738 m³ air exchange per hour, with ΔT of 34 K results in:
34 K * 738 m³ * 0.34 W/(K·m³) = 8531 W = 8.5 kW.
The 0.6 comes from a standard. The whole procedure is more complex overall (especially for the basement area, where depending on depth, you can calculate ground temperature and possibly specific heat transfer coefficients for the basement slab, since usually less heat is lost to the ground, or the heat partly remains ‘stored’ near the basement area in the soil, simply put. As a physicist, you know about nonlinear differential equations with boundary conditions, which could make this calculation more precise). The estimate I’m making here with temperature is still “rough,” and actually, less heat loss should occur.
Otherwise, you have basically repeated it correctly.
Similar topics