Exercise on sensitivities in borefield design

In this exercise, we will investigate the design of a geothermal borefield for an office building. The aim is to gain insight into the effects of the geothermal temperature gradient and to understand which parameters are particularly important when designing a borefield for a building with a high cooling demand.

The exercise

The case for this exercise is based on a real-life office building located in the city of Ghent (Belgium). To design a suitable borefield for this building, you’ll need to draw on knowledge from a range of previous articles (which we’ll reference where necessary). Through this exercise, you’ll explore the influence of the geothermal temperature gradient on the design, compare the impact of using MPG versus water, assess the choice between single and double U-tube configurations, and gain general insights into designing borefields for buildings with a high cooling demand.

!Hint
To get the most out of this exercise, we strongly recommend attempting the design questions below before reading the provided solution. Borefield design is far from straightforward, and the best way to master its complexities is through hands-on experience.

Input parameters

General input parameters

• Minimum average fluid temperature threshold: 2°C
• Maximum average fluid temperature threshold: 17°C

!Hinweis
These temperatures are selected to ensure the cooling can operate in a passive or free mode (hence the upper limit of 17°C). The lower limit of 2°C is set to avoid negative temperatures within the borehole. An average fluid temperature of 2°C typically corresponds to a supply temperature of 0°C and a return temperature of 4°C, assuming a temperature difference ($\Delta T$) of 4°C.

• Simulation period: 40 years
• First month of the simulation: January

Ground input parameters

The geological conditions of the ground at the project location is the following:

!Hinweis
Notice the thick, poorly conducting clay layer in the lithology. This will play an important role in this exercise.

In order to save time, the average ground thermal conductivities are calculated to be:

• 1.6 W/(mK) @ 150 m borehole depth
• 1.7 W/(mK) @ 100 m borehole depth

The other ground parameters are:

• Volumetric heat capacity: 2.4 MJ/(m³K)
• Location: ‘Bel-Gent’

Borefield input parameters

All the borefield in this exercise are rectangular, with an equal borehole spacing in length and width of 6 m. The buried depth is 1m and the initial, starting configuration is 15 x 13 boreholes with a borehole depth of 150 m.

Borehole resistance input parameters

The parameters for the pipe are:

• Double DN32 PN16 pipe (i.e. a wall thickness of 3mm and an outer diameter of 32mm)
• Borehole diameter: 140mm
• Distance from pipe to borehole center: 40mm
• Grout: 1.5 W/(mK)

The fluid is 30 v/v% MPG with a 0.2 l/s flow rate per borehole.

Thermal load input parameters

• Heating power: 306 kW
• Yearly heating energy: 398 MWh
• Cooling power: 336 kW
• Yearly cooling energy: 269 MWh
• SCOP: 4.5
• SEER: 20 (passive cooling)

Design questions

For this exercise, you are invited to answer the following design questions while tracking the total borehole length for each step. This will help you assess the cost and performance implications of various design changes.

!Hint
To keep your work well-organised, it is recommended to use a separate scenario for each design question.

1. Given the original borefield design of 15×13 boreholes @150 m, is this a good design?
2. How many boreholes should we drill extra if we decrease the borehole depth to 100 m? Try to think about this first before you start simulating.
3. What happens if we update the thermal conductivity of the ground to the correct value? Can we change the design?
4. What happens to the temperature profile if we change the fluid to water?
5. How can we redesign our borefield to be more cost-efficient?
6. What happens to our design if we switch from a double U tube to a single U tube?

Solution

Below you’ll find the answers to the design questions outlined earlier. It is important to emphasise that there is no single correct answer. The value of this exercise lies in understanding the reasoning behind each decision rather than strictly agreeing with every assumption.

Each geothermal project is unique, and the choices you make—regarding parameters, configurations, and thresholds—depend heavily on project-specific constraints, design priorities, and practical considerations. Use these answers as a guide, but don’t hesitate to challenge the assumptions and explore alternatives.

Question 1

The temperature profile from the original simulation shows a maximum average fluid temperature of 17.17°C, which is slightly above the design threshold of 17°C. At first glance, this might suggest that the borefield is undersized. But is that really the case?

Not necessarily.

In borefield design, there are always underlying assumptions—both in the input data (e.g. load estimations, operational conditions) and in the model itself (e.g. neglecting groundwater flow, as discussed hier). These simplifications can introduce small deviations from actual system behavior.

Given the scale of the project and the method used to estimate the cooling demand, a result of 17.17°C is still acceptable within the expected uncertainty margin. However, for projects with tighter thermal constraints or regulatory requirements, even a 0.17°C overshoot might be critical and should be addressed.

The total required borehole length for this configuration (15 × 13 boreholes at 150 m depth) comes to around 29000 m.

Question 2

When we lower the maximum allowed borehole depth to 100 m, we reduce the current depth by 33%. Instinctively, we would thus expect to increase the number of boreholes by the same 33% in order to end up with the same total borehole length. When we do the simulation, we end up with a borefield of 17 × 13 with a total borehole length of 21879 m, which is significantly less than in question 1. There are two main reasons for this:

• First of all, since we are now drilling less deep, the ground temperature is lower due to the thermal gradient. Since this borefield is limited by the maximum average fluid temperature, a lower ground temperature means that fewer boreholes are required. (If you have not read our article on the ground properties, you can find it hier).

• Secondly, since the boreholes are now shorter, the effective borehole thermal resistance is also slightly higher. This is due to the fact that there is less thermal short-circuiting between the legs in the U-tubes. (You can find more information about the borehole resistance in our article hier).

Question 3

If you recall, we changed the borehole depth but kept the ground properties constant (or perhaps you changed them yourself, which is even better!). Since we are now drilling less into the clay layer, the conductivity increases from 1.6 W/(mK) to 1.7 W/(mK). This lowers the temperature to 16.66 °C, due to a slightly better heat transfer in the ground and an even lower undisturbed ground temperature.

!Hinweis
At first sight, it may seem strange that the ground temperature changes when we don’t change the borehole depth. This is because the undisturbed ground temperature is calculated based on the geothermal heat flux (which remains constant) and the ground conductivity. In a poorly conducting layer, the ground temperature increases more rapidly than in a well-conducting one. Therefore, changing the average ground thermal conductivity to a higher value lowers the temperature gradient, resulting in a lower undisturbed ground temperature.

Due to this higher ground thermal conductivity, we can reduce the borefield size to 16 × 13 boreholes and still stay below our 17 °C threshold, resulting in a total borehole length of 20592 m.

!Caution
This question illustrates the importance of keeping track of the different ground layers and the potential danger of assuming a homogeneous ground.

Question 4

Since the fluid temperatures are relatively high, there is no real need to use antifreeze (and certainly not 30 v/v% MPG, which offers freeze protection down to −14 °C). We therefore set the percentage to zero and raise the minimum average fluid temperature threshold to 5.5 °C. As a result, the maximum average fluid temperature drops to 16.25 °C.

!Hinweis
This temperature depends on your specific heat pump requirements. Typically, a condenser outlet temperature of 4 °C is permitted. If you then assume a $\Delta T$ of 3 °C across the condenser, you end up with an average fluid temperature of 5.5 °C.

The Reynolds number, both for extraction and injection, now shifts out of the laminar region into the transient and even turbulent regime. This reduces the average borehole thermal resistance, leading to a lower ground temperature.

Question 5

Due to this improved heat transfer, the borefield size can be further reduced—to 14×13 boreholes, for example—resulting in a maximum average fluid temperature of 16.88 °C and a total borehole length of 18018 m, which is significantly less than the 29000 m we started with!

Question 6

As a final variation, we change the double U DN32 to a single DN32 pipe. This increases the effective borehole thermal resistance, resulting in a higher maximum average fluid temperature. To stay within our temperature boundaries, we need to increase the number of boreholes again to 15×13, giving us a final total borehole length of 19305 m.

!Hinweis
One might find it surprising that switching from a double U to a single U significantly increases the total borehole length. Normally, when the double U operates in a laminar flow regime, switching to a single U leads to turbulent flow, which improves heat transfer and keeps the total borehole length more or less the same. However, since the double U was already operating in a turbulent regime in this case, switching to a single U merely reduced the heat transfer area, resulting in a higher effective borehole thermal resistance.

Schlussfolgerung

This exercise demonstrated the various design strategies available when sizing a borefield for a building with a high cooling demand. By making informed decisions, we managed to reduce the total borehole length from 29055 m down to 18018 m, thanks to the lower undisturbed ground temperature and the improved borehole thermal resistance when using water instead of MPG. Switching to a single U configuration proved to be of no benefit in this project, as the fluid was already in a turbulent regime.

Literaturverzeichnis

• Sehen Sie sich unsere Videoerklärung auf unserer YouTube-Seite an, indem Sie klicken hier.