The Homemade Heat Pump Manifesto

[Daox]

Great progress!

How deep did you have to make your trenches?

[AC_Hacker]

Originally Posted by Daox

Great progress!

How deep did you have to make your trenches?

The frost line here was legally defined as 16 inches. It has been legally re-defined to 9 inches... Global warming?

I dug the trenches to about 2 feet.

By the way, the loop field is now complete.

Details to follow.

Regards,

-AC_Hacker

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

[GvilleRenovator]

That is GREAT!

Did you bury any one wire temp sensors?

[AC_Hacker]

Originally Posted by GvilleRenovator

That is GREAT!

Did you bury any one wire temp sensors?

That would have been an excellent idea, I thought about it often, but I really don't have a sufficcient understandiing of '1-Wire' technology at this time.

It would have been a very good idea to have a sensor on the 'IN' and 'OUT' of each loop and a sensor on the 'IN' and 'OUT' of the total loop field.

Also it would have been a good idea to have put a vertical string of sensors, say every two feet, on one or more loops.

I actually have a package with four 1-Wire temp sensors in my desk drawer. I'm schemeing to make a 15 to 17 foot plastic pipe with sensors, all filled with plastic to make it water proof, and burying it somewhere in the middle of the loop field.

Do you have a handle on 1-Wire good enough that you could explain it to me?

Regards,

-AC_Hacker

[AC_Hacker]

Here's a diagram of my Loop Field...

...as you can see, I planned for more loops than I actually put in.

If I need to, I can add more loops in the future. There's even room in the front yard for maybe 6 additional loops. During the summer, I measured the pavement of the street in front of my house, and it was 145 degrees F on the hottest day. Some serious heat under that street, and with my name on it too!

I arrayed the boreholes as I did to get them as close as possible, like how honey bees arrange their cells. It allowed me to pack more holes in the space I have, but it really made trenching more difficult. If I had more space to work with, I would definitely put them in a grid pattern.

The hook-up sequence was arrived at because I wasn't sure how much pressure would be reduced by stringing all the loops in series. As I have it, I can cut in after the 8th loop and re-connect the field as two branches, each with 8 loops.

It's much simpler with one loop, but if I can't get the flow rate I want, I may change it in the future.

[EDIT (2015-04-22) the idea to leave the possibility off future change turned out to be very important. My total loop field pipe length was 720 feet, way too long for 3/4" pipe and required almost 1/4HP pump. I subsequently cut the loop in the middle which gave me two loops and it cut my pump power requirement by a factor of 4! (1/16HP), because I was able to get the same flow quantity at half velocity, and flow resistance increases (& decreases) by the square of velocity.]

Here's the loop field after trenching, before connecting all the loops:

I found that it really pays to run the trencher as close as possible to the plastic pipe, without hitting the plastic. Making corrections for a wide space (I had many) has been very time consuming.

Also the trencher threw the dirt up just over the edge of the trench and quite a bit fell back into the trench. The trusty Rigid shop vac was the perfect tool for clearing out the trenches and reclaiming those precious extra inches.

Here's Rhonda-the-Grandma, demonstrating good trench-vacuuming technique:

As I stated previously, I though I'd need to use barbed connecctors in the trenches, but later realized that I could pull the pipe up tight, weld the connection, working just inside the trench and then push the pipe in and cover it while holding it down with anything heavy (like large stones or my foot).

Here's a pic of the welding jig as I used it in the trench. I put a cinder block under it to hold it up. By tightening the hose clamps on one side of the jig and leaving them a bit slack on the other side, I was able to free up a hand which I'd then use to hold the fusion paddle.

After each weld, I'd let it rest for about 5 minutes, then I'd fill it up with water and gradually introduce air pressure to 90 psi. This should be done gradually, because if you hit the water filled tube fast, it gets the water moving pretty quick which can exert much more than 90 psi on the branch under pressure. I didn't do it, but I'm pretty sure that it would be possible to rupture a water-filled branch by pressurizing it too fast. I think this is referred to as "Water Hammer". I tested every weld, every time. Leaks were very easy to spot, water was spraying out of the few bad welds. Using a water bath method and looking for air bubbles to find leaks was out of the question.

Here is a pic showing two welds:

The top weld is a nice looking weld, the bottom one was done in a hurry and looks pretty crappy.

However, when I filled them with water and pressurized them, they both held just fine.

That's what really counts.

Best Regards,

-AC_Hacker

(* more to follow *)

%%%%%%%%%%%%%%%%%%%%%%%%%%

[GvilleRenovator]

Originally Posted by AC_Hacker

Do you have a handle on 1-Wire good enough that you could explain it to me?

I WISH I had a clue. Searching for answers currently.
I agree on your idea for a string of sensors. Since my untested plan is to have loops 40 - 60 (or more) feet deep I was thinking 10 - 15 foot spacing.
I just ordered my Logomatic and several 1 wire temperature sensors. I will post my findings/experimenting on your DIY Data Logger thread.

[AC_Hacker]

Originally Posted by GvilleRenovator

... I just ordered my Logomatic and several 1 wire temperature sensors. I will post my findings/experimenting on your DIY Data Logger thread.

Well, if it helps, here's what I know about '1-Wire'

• 1-Wire actually requires at least two wires and probably even better, three wires. You gotta have at least a data wire (which has a small amount of voltage), and a ground wire. There's an operational mode (paricitic) wherein a small capacitor that is internal to the 1-Wire chip stores electricity that is 'borrowed' from the data line, and this charge is used to run the chip. Because it takes some amount of time for the capacitor to store enough power to execute a request for data and to write data, there is a limit to the frequency of data requests/writes. There is another mode wherein the 1-Wire chip uses a data wire, a ground wire and a dedicated power wire. in this mode, you can make data requests as often as you wish, limited only by the chip's ability to recognize and respond to a data request.
  
• Each 1-Wire device on a string is hard-coded at the factory with its own unique identifying number. When there is a request for data, all devices read that request and look to see if their unique ID number is contained in that request, if it is, then that chip will respond, if not then no response.
  
• To send a request to the chip, and to read the response from the chip, some device that can be programmed to provide the proper timing and instruction sequencing and response reading, is required. I looked to see if there was a set of chips that had these routines all pre-written and ready to go. I wasn't able to find such a thing, with the exception of a chip that would send out the instruction string with the correct timing and instruction sequence. This instruction sending and data recieving is the kind of thing that microprocessors like the Arduino (the Logomatic uses the ARM processor, which is also a good choice) are good at doing. I found websites that had examples of 1-Wire code that could be incorporated into a C source code that would then have to be compiled and loaded into the Arduino. There is an Arduino group here in town (there's a good chance that there's one in your town, too) that has a good bit of expertise in these matters, and they are friendly and ready to help. There are also on-line groups who are also friendly and ready to help.

... but I was so busy, busting my butt, trying to get the loop field in before the rains began, that I couldn't take the time to learn the fine points of the Arduino (or the ARM) design cycle.

Hope this helps...

Regards,

-AC_Hacker

&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

[AC_Hacker]

Please note that there was a proceedural detail I failed to include, so to keep things in approximate order, I appended the information to a previous post.

Please follow this link: apppended loop spacer tube details for that information.

Best Regards,

-AC_Hacker

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

[Daox]

Sounds like a good idea, but some of your pictures aren't showing.

[Hugh Jim Bissel]

Looking awesome!

Originally Posted by AC_Hacker

It would have been a very good idea to have a sensor on the 'IN' and 'OUT' of each loop and a sensor on the 'IN' and 'OUT' of the total loop field.

Also it would have been a good idea to have put a vertical string of sensors, say every two feet, on one or more loops.

Since you're running your loops in series you wouldn't need more than a send and a return temperature which you could read at the inside unit. If your loops were in parallel having a send and return on each parallel leg would allow you to verify each leg was providing equally.

Having more sensors than that (ie send and return on each well in series and/or every 2 feet) would be cool to have to see HOW it is working, but would be unneeded for anything except additional info. The main send/return temp change (and actual water temperature) is the big thing.

Looking forward to seeing it in action!