Geothermal Series – Part 5

Introduction

In this post we are going to look at the robust sensor network that I installed in the geothermal system. I will give the good and the bad of the installation and design process.

If you have not read the previous posts in this series, please do so.
Part 1 – DIY Geothermal System Design
Part 2 – Geothermal Trenches – Digging for Energy
Part 3 – Ground Loops
Part 4 – Geothermal Mud

Purpose

When I decided to do a geothermal build, I wanted to try some unique solutions, like the geothermal mud. In order to see if those solutions would pay off, I decided to install a sensor network that would monitor soil and water temperatures at different locations. My goal is to better understand how the temperatures change along the ground loops. I wanted to measure the water temperatures at different points, as well as the ground temperatures.

The temperature sensors would do more than just provide valuable data. Temperature would be the trigger for the water pump that would provide the circulation of the fluid. In order to make this happen, I had to design a complicated system.

Design

The sensor network had to do a number of things. First, it needed to measure ground temperatures. Second, it needed to measure water temperatures. I also decided to add a soil moisture sensor and several above ground temperature and humidity sensors, along with a weather station.

Sensors

In order to get things started, I needed to know what kind of sensors to use. I went back and forth between several sensors before settling on the Dallas Temperatures DS18B20. Of course, I didn’t get the name brand ones, I got the knockoffs that were a fraction of the price and were able to be submerged in water. The DS18B20’s are just temperature sensors, so they don’t do anything else. But their biggest advantage is that they are one-wire capable. Meaning, that I can put all the sensors on a single signal wire.

For a while I looked at special water temperature sensors to use in my system. Unfortunately, water temperature sensors are either unreliable, expensive, or not available in a digital format that can be read at a distance. I decided to utilize the water proofing of the DS18B20 and make my own water temperature sensor. How hard could it be?

For the soil moisture sensor, I went with an Adafruit capacitance sensor. The more common type is the resistance sensor, which has two probes that measure the resistance between them. Unfortunately, the resistance sensors tend to corrode quickly and fail. Not a good solution for something buried 12 feet down.

The BME280 is my above ground temperature and humidity sensor. I also used a Sparkfun Weather Shield to run a small weather station to get the wind and rain data. The Weather Shield has a built in temperature and humidity sensor, but I wanted to get that information from a different location.

Design

The original design was to have five evenly spaced temperature sensors per ground loop. This would record the ground temperature at the loop depth, over the length of the loop. At the end closest to the tank on Loop 1 (East Loop), would be two arrays of sensors. One would go vertical, directly over the ground loop and would have a sensor every two feet from 12 feet down to 2 feet below the surface. The other array would extend horizontally at the depth of the ground loop to see how far out the heat would extend.

As a control, I placed a second vertical array in between the two loops at the far end. This measures the ground temperatures from the depth of the loops up to the surface in an area not effected by the loops. The location of the horizontal and vertical arrays are at the beginning of a loop, where it is the hottest, so that the effect is maximized.

Water temperature sensors would be at the entrance and exit to the tank, and then every 300 feet of ground loop under ground. This would give a little gradient over the length of the tubing instead of just measuring the input and output temps. The soil moisture sensor is in the middle of Loop 2 (West Loop). The soil moisture sensor runs on i²c so it had its own wiring.

I used 1/2″ conduit to connect all the sensors together. This protected the wires and connections from water. A trench connected the end of Loop 1 and Loop 2. This would be where the vertical control array would be. Unfortunately, during the digging of the trenches, the connecting trench collapsed. The ground was too unstable in that area to the the excavator back in to dig it out, so we had to go to plan B.

Plan B involved moving the control array to a new location. I moved it to the very end of the horizontal array to get it as far away from the loop as possible. This did give an added benefit in that we could reduce the number of sensors by one. The end of the horizontal array and the bottom of the vertical array would share the same sensor. One limitation that I see is that the new location is often shaded by the building and that may affect the ground temps close to the surface.

Ground Temp Adapters

The sensors that I used did not fit the conduit connectors. In order to make the two work together, I needed to design and make some adapters. The ground temperature adapters needed to hold the sensor and connect to a half inch conduit T-body. The water temperature adapters needed to hold the sensor inline with the water flow, while preventing leaks.

Ground Temperature Sensor Adapters

The ground temperature sensor adapters were the easiest of the the two main adapters. The adapter has a wide base that can fit over the outlet of the conduit body. The part that holds the sensor protrudes out to support the sensor. The entire piece was 3D printed and then sealed with epoxy resin to improve its waterproofing and durability.

Water Temp Adapters

The water temperature sensor adapters provided a unique problem. First, they needed to hold the sensor in the flow of water, which means that it had to enter the pipe without letting water out. Next it had to hold the sensor in a way that did not block the flow of water too much.

Design

I designed the adapter to fit inside a PEX T-fitting. The T-fitting would connect two sections of PEX, with the third opening being occupied by the sensor. Figure GI-5-7 shows the T-fitting from the point of view of the water. Notice how the tip of the sensor barely enters the flow, helping to reduce any restrictions, but still get good temperature readings.

The narrow end of the adapter fit inside the PEX fitting. The thicker end with the ribs duplicated a PEX fitting. A piece of PEX tubing clamped to both the adapter and the T-fitting held everything in place. I tested the setup on my water supply line in my shop. The water supply for my shop is close to 90 psi as it does not have a regulator.

Installation

Installation of the sensor network was difficult and very time consuming. Getting this right was the bulk of the time taken doing this project. There were 31 underground sensors that had to be wired, glued, and sealed prior to dropping the loops into the ground. The conduit was tied to the ground loops to hold everything together. Luckily, 1/2″ conduit is very flexible so it still allowed us to put one loop in at a time.

The vertical arrays were the most difficult to deal with during the installation. They had to be wired and glued in place at the last minute and then dropped into the trench. I also had to used pieces of conduit just to support them. But everything worked out eventually.

Thank You

Thank you for reading my post. If you find that you gained some knowledge and would like to read more, please consider giving us a like and subscribing. It’s free and it helps us get sponsors for future content. Also if you feel we are worthy of a small donation, you can leave a tip at the button below. Every little bit helps as the projects can be quite costly.