# Question on long distance soil temp testing

Status
Not open for further replies.

#### fastline

##### Member
I am designing a hybrid or slightly different horizontal geothermal HVAC loop and will need or like to monitor soil temps are various points for research purposes. The field is large and if I can find an economical way to set probes, I would like to have several. I will worry about how I want to display them later. I need only be concerned with adding the probes during installation of the geo loop right now.

Now, I had a thought to make the system really simple and just add several PVC risers down to the depths I need to test so I could simply walk over and drop a probe in there. However, it would be nice to have permanent probes and since I am excavating the area anyway, I can throw wiring in there.

Just not sure what probe system would be best for this? That seems like a LOT of K-TC wire. Maybe thermistors would be best or would there be an accuracy problem with the long runs of wire?

#### Les Jones

##### Well-Known Member
If you could have some electroncs at each probe site (Possibly powered by a solar cell and battey for nightime readings) you could use a 20mA loop system or convert the analog signal to a digital signal. Depending on how large is large you may be able to use RF to send the signals. Possibly ESP8266 modules or HC-12 modules.

Les.

#### AnalogKid

##### Well-Known Member
I like thermistors. They are weirdly non-linear, but can be 0.01% accurate for reasonable cost. How long is the longest wire run? 10 ohms of wire resistance is 0.1% error for a 10 K thermistor at +25C. With something like an LM4040 reference and a reasonably decent opamp, a relatively simple linearizing circuit produces good results. If the perfect output is a plot of output voltage versus temperature that is a straight line, then a plot of the error superimposed looks like a long strung out S. Calculated error is below perfect at low temps, crosses to above perfect, heads down again crossing dead-on at 35C, is below perfect for a while, then cross again to above perfect. Calculated peak-to-peak error is +/-0.5C from 20C to 50C, and gets worse outside that range.

One circuit, trimmed and tweaked, with a big analog mux in front.

ak

#### fastline

##### Member
I think what I am mostly looking to do is target a method or plan just so I can prewire or drop conduits to revisit this after it is installed. If you can imagine a geothermal tube in the ground, I would like to see temps near the tube and outward to about 3ft or so. I think I could possibly set this up to keep all thermal probes within a 3ft x1ft area in which I could then set a common box there for a digital converter or RF unit?

So if I simply install risers right at the depth and spacing that I need, any idea how I might design the probes to go in there?

We have worked with thermistors before and I think we simply programmed the response graph per the manufacturer into a PIC uC and was very accurate. I am trying not to make this all too complicated though.

the temps should stay between 40-80F

#### AnalogKid

##### Well-Known Member
Let me back up. A negative temperature coefficient (NTC) thermistor has an approximately logarithmic change in resistance as a function of temperature. The warmer it gets, the lower the resistance. Devices are specified at their 25C resistance. My fav is a 10K part made by Thermometrics. If you start with a constant voltage source and put a resistor in series with the NTC, the voltage at the node is *approximately* linear with temperature. If you have a temperature range of interest, and bias the thermistor with a fixed resistor equivalent to the NTC value in the center of the range, the error band will be centered about the response curve. This can be compensated with a couple of opamps, but the easiest way is to digitize the voltage and use it to index a look-up table.

One tweaked stable low power voltage regulator
One precision bias resistor
One analog switch or reed relay array to mux the thermistors into the circuit
one A/D or uC with an A/D built in

The mux keeps the overall system simple, and keeps the bias current from heating the thermistors and introducing a self-heating error.

As for the sensors, you can buy them pre-packaged in a waterproof capsule with long leads, or make your own with a soda straw and some RTV.

I see you just posted some additional info. spounds like you're already experienced in these topics.

ak

#### JonSea

##### Well-Known Member
Use One-Wire DS 18B20 temperature sensors. You can run many of them on a 3 wire network of a thousand feet or more without problems.

#### tcmtech

##### Banned
A standard platinum wire RTD device like a PT-1000 tends to work very well for longer distance measurements and can usually be connected through standard twisted pair type wire such as ethernet cabling as well without having much effect on it plus can be interfaced with through a cheap PT1000 to voltage or converter unit.

Also with cat5 cable being ~2.5 ohms per 1000 feet the cable resistance is negligible in the overall sensor loop.

Looking at generic PT1000 devices and converters you could put together 7 channel single converter sensor arrays for about $30 -$35 apiece plus the cost of cable.

That would be my choice for the design.

#### fastline

##### Member
One other consideration here is the soil may have a rather low PH. Anything that I might bury would need to be a long life solution. Would there be a consideration regarding probe types if the life of the probe and system is considered?

#### fastline

##### Member
OK, in thinking about all the possibilities, i think the wisest thing here might be to install my proposed risers out of the ground at designated heights and locations, then while the field is being built, just throw in a 1" conduit line back to the mother ship so later I can simply pull wire out to the testing site and drop some sensors down the holes. I have concerns that if I direct bury my stuff, if for whatever reason I have a failure, I don't have a backup. With my proposed setup, I guess I can always pull the sensors back up and reinstall something else if needed.

#### tcmtech

##### Banned
So why do you feel you need to monitor the field soil temperatures in different or any locations anyway?

Is this for personal research or actual academic work and what makes it a hybrid system?

#### fastline

##### Member
Mostly for personal and green energy research. the geo loop is being placed under a new septic lateral field. I want to determine how the saturated zone affects the thermal conduction in the area and if there are any adverse affects. I am not the first to do this but it seems this type of installation is not very well documented thus not much support by authorities.

##### Active Member
I second JonSea's suggestion. This is done with grain elevators and deep-well water monitoring, often 3000+ feet into the ground. E

#### tcmtech

##### Banned
Mostly for personal and green energy research. the geo loop is being placed under a new septic lateral field. I want to determine how the saturated zone affects the thermal conduction in the area and if there are any adverse affects. I am not the first to do this but it seems this type of installation is not very well documented thus not much support by authorities.
You won't gain a huge amount I can tell you that much.

Dry soil has a natural specific heat of ~ .2 and wet soil is around .35

http://www.engineeringtoolbox.com/specific-heat-capacity-d_391.html

Also the relation of heat pump system operation Vs soil conditions has been extensively studied and documented as well.

Realistically being under your sewer drain field runs will help with the thermal mass and some thermal conductivity (but not a huge gain on either) but if you are using the system for winter time heating you could run the very real possibility of freezing your sewer drain field solid rendering it unuseable in the dead of winter.

I haven't worked with geothermal much myself but I have talked to enough people who have dealt with it and they overwhelming view is that you take whatever your contractor says is sufficient for your ground loop and at least double it.
Reason being very few places ever have exactly dead equal summer/winter heat sink/source ratios let alone a natural ground temperature that's favorable to both so at some point after a few seasons of regular use you will end up with either a thermal saturation or thermal deficiency issue in your effective ground loop assuming nature itself has not already put a heavy bias on one end to begin with causing your system to stall at one or the other early.

Where I live our natural ground temperature is ~40 degrees which makes geothermal here useless for anything but summer time air conditioning no matter how big of ground loop is used.
Many here have tried and many contractors got rich and keep getting richer from those attempts but so far no one who has had a system for any length of tine says it was as good as the sales and contractors said it would be for heating.

http://www.greencastonline.com/tools/soiltempmaps.aspx

#### Nigel Goodwin

##### Super Moderator
One other consideration here is the soil may have a rather low PH. Anything that I might bury would need to be a long life solution. Would there be a consideration regarding probe types if the life of the probe and system is considered?
You can get water proof versions of the DS 18B20 for very little money, which should be ideal for your requirement.

#### spec

##### Well-Known Member
Just a point about transporting an analogue voltage value over a long distance. The best option these days is to convert the voltage to digital and transport using a digital link.

But if you must stay with analog you should use a current link where the analog voltage is converted into a current. Then the voltage drops caused by wire and contact resistance do not affect the accuracy.

By the way, converting from voltage to current and back again is no big deal.

spec

#### fastline

##### Member
You won't gain a huge amount I can tell you that much.

Dry soil has a natural specific heat of ~ .2 and wet soil is around .35

http://www.engineeringtoolbox.com/specific-heat-capacity-d_391.html

Also the relation of heat pump system operation Vs soil conditions has been extensively studied and documented as well.

Realistically being under your sewer drain field runs will help with the thermal mass and some thermal conductivity (but not a huge gain on either) but if you are using the system for winter time heating you could run the very real possibility of freezing your sewer drain field solid rendering it unuseable in the dead of winter.

I haven't worked with geothermal much myself but I have talked to enough people who have dealt with it and they overwhelming view is that you take whatever your contractor says is sufficient for your ground loop and at least double it.
Reason being very few places ever have exactly dead equal summer/winter heat sink/source ratios let alone a natural ground temperature that's favorable to both so at some point after a few seasons of regular use you will end up with either a thermal saturation or thermal deficiency issue in your effective ground loop assuming nature itself has not already put a heavy bias on one end to begin with causing your system to stall at one or the other early.

Where I live our natural ground temperature is ~40 degrees which makes geothermal here useless for anything but summer time air conditioning no matter how big of ground loop is used.
Many here have tried and many contractors got rich and keep getting richer from those attempts but so far no one who has had a system for any length of tine says it was as good as the sales and contractors said it would be for heating.

http://www.greencastonline.com/tools/soiltempmaps.aspx

Thank you for the response. I have studied this design for about 3yrs now. My climate is quite different in the midwest as we have both brutal summers and winters.
Here are some design points for consideration.
1. The geothermal HVAC system was free to me. A nice Climate Master 5T unit!
2. I own an excavator so the digging comes cheap
3. The whole structure is also getting radiant floor tubing so there will be opportunity to add supplemental heat via other sources
4. Insulation is such that even though the system and loops are designed around 5T operation, 3T will be most common (low mode)

The main reason to install under the lateral field is I have to dig it to 4ft anyway so I will dig a little deeper and save a ton in digging time. It seems senseless to do all this digging and not throw tubing in it, even if I had to add more loops to get demand covered.

#### fastline

##### Member
You can get water proof versions of the DS 18B20 for very little money, which should be ideal for your requirement.

Am I correct that these are integrated devices with converters built on them so the output directly from the device is digital?

Assuming I wanted an economical device to crunch the numbers on all these thermal inputs, is there such a critter? Well, I know there is, but less than \$500?

As part of this system, I will also have to monitor a pile of thermal readings indoors as well.

#### Nigel Goodwin

##### Super Moderator

As you're doing this totally from scratch, it's something you want to get right from the planning stage.

#### JonSea

##### Well-Known Member
Each device on a One-Wire system has a unique address. A system can work with a pair of wires (power/data and geound) but a 3 wire system (data, power, and ground) has some advantages. Two factors limit the number of devices on a system and the maximum cable length. The first factor is the amount of capacitance of the system; each sensor adds capacitance, as does each foot of cable. Capacitance limits the rise time of the digital signal and at some point, the pulses are too rounded-off to work. The length of the cable is limited by the propagation though the cable.

Because of its capacitance, Cat-5 cable isn't the best choice from a network point of view. Years ago, I did some experiments - as I recall, with 5 or 10 devices, a network using 700' of cable provided reliable results. The usual warnings apply here; if you're planning on a network anywhere near that long, test it out first. The system I designed had a circuit board with a pair of RJ-45 jacks, a temperature sensor and a memory chip on the board and a 2' stub (which is not recommended) to an addition temperature sensor.

Arduino has modules to read One-Wire devices. It would be fairly simple to use an Arduino to read a One-Wire network and spit results out to a PC. The are also code modules for the ESP-8266 so you could make a wireless network. I don't know how well they would handle multiple devices.

Here's an application note from Maxim on designing reliable One-Wire systems. They have a ton of applications notes.

Here's an example of the waterproof probes Nigel mentioned, from ebay. I have used similar cheap probes from ebay and Chinese vendors and they work well. DS18B20s in TO-92 packages are also cheap on ebay.

#### tcmtech

##### Banned
Thank you for the response. I have studied this design for about 3yrs now. My climate is quite different in the midwest as we have both brutal summers and winters.
Here are some design points for consideration.
1. The geothermal HVAC system was free to me. A nice Climate Master 5T unit!
2. I own an excavator so the digging comes cheap
3. The whole structure is also getting radiant floor tubing so there will be opportunity to add supplemental heat via other sources
4. Insulation is such that even though the system and loops are designed around 5T operation, 3T will be most common (low mode)

The main reason to install under the lateral field is I have to dig it to 4ft anyway so I will dig a little deeper and save a ton in digging time. It seems senseless to do all this digging and not throw tubing in it, even if I had to add more loops to get demand covered.
Sounds like you're in the same boat as me. I to have a large commercial application heat pump unit I picked up cheap years ago and my own excavator equipment as well.

The only downside here is not needing high powered air conditioning given the system I have could easily cool two large houses plus my 32' x 64' shop and my 14' x 20' workshed without breaking a sweat.
Heating wise where I would need it it would have a hard time heating one house. unless I added huge solar thermal collector systems to heat load my ground loops all summer to get enough energy stored to run with any efficiency in the winter.
But for that the numbers even at the full on DIY level just don't add up in my favor being I have near unlimited and free fuel for heating at my disposal through collecting and burning used oil or wood which I have been doing for the last ~15 years now.

Status
Not open for further replies.