A low-cost pipe freezer with limited applications

klugesmith, Sun Jul 02 2017, 02:34PM

Water pipe freezing is done by many plumbers on purpose. Forms a local plug of ice, to facilitate work downstream. I want to crawl under my house and do some pipe freezing to isolate a persistent leak of a few liters per day.

Commercial pipe freezing kits all include insulated boots that wrap around a pipe and admit coolant through a hose. In order of increasing cost, and decreasing cost of consumables, they use:
* canned refrigerant (freeze spray),
* liquid CO2 from a high pressure cylinder & hose,
* or glycol antifreeze from an electric chiller.

At box store, one can order an Arctic Freeze kit for $50. At a shop in the next county, one can rent a Rothenberger kit for $75 for 3 hours, $150 for 24h. Could use ordinary freeze spray in a home-made boot; one demonstration is posted by Hackaday in their usual "aren't we cool" style. Could use a loosely filled bag of crushed dry ice & a blanket of bubble wrap.

But no, Richard had to do it the fun way.
Let's see if these parts fit together to make something useful.
1499005296 2099 FT0 Dscn9958

1499005296 2099 FT0 Dscn9959

1499005296 2099 FT0 Dscn9960

1499005296 2099 FT0 Dscn9961


I've been obsessed with this for a week or so, even after recognizing more practical ways to get the job done & move on with life. Finished box #2 last night. Froze a test pipe with no surprises or disappointments. Sorry, no pictures yet.
Re: A low-cost pipe freezer with limited applications
Dr. Slack, Mon Jul 03 2017, 06:48AM

I guess you have to keep the origin of your name current, Smith of Kluges!

You've been a naughty documentor and not put a size reference (USB stick, CD, ruler even, not a coin - keep it international) in any of your pictures. I don't recognise the strapping in the G-clamp picture, is it Dexion pitch?

I am with you on the 'never buy if you can build' principle. However, I've come to realise over the last half century that it's not stopping one getting on with life, it is life.
Re: A low-cost pipe freezer with limited applications
klugesmith, Mon Jul 03 2017, 08:49AM

Oops, about the scale-less pictures. Is the railroad spike in one picture generally recognizable and of familiar size?

The concave cylindrical surface is supposed to fit 3/4" steel pipe, which has an OD of 1.050" (26.7 mm). Isn't the corresponding size in the UK called 20 mm? Does it still usually get BSP threading?

Over here, nominal 1" copper pipe has OD of 1.125" and comes in three weights. Type M, the lightest, has ID of 1.055" (26.8 mm). That serendipity begs to be exploited, eh?

1499069366 2099 FT180082 Dscn9973

1499069366 2099 FT180082 Dscn9971
The first experiment today was a learning experience. In order: placed boxes, dry ice, bubble wrap, alcohol, bubble wrap & rag on top. Crawl back to access hatch & blue sky overhead. After 11 minutes of chilling, verify ice plug by opening a valve downstream. Check water meter for creepage -- yay, it's stopped! Didn't move at all in the next 40 minutes, then started quite suddenly. On the return trip, take pictures during tear-down; later show them out of order.
Re: A low-cost pipe freezer with limited applications
klugesmith, Sat Aug 12 2017, 08:00AM

Indoor water leak is still un-pinpointed, but I got back to work on it after a few weeks off. Have continued to experiment and to learn.

Freezer model 1 (presented in OP) worked flawlessly for 4 experiments in the crawl space. With a single load of dry ice and alcohol "coolant", ice plug formed within a few minutes and generally persisted for an hour or so. The Boolean results (did water meter creepage stop or not) narrowed the offending flow to a single Tee off the main 3/4" line.

Now we were looking at a nominal 1/2" pipe, leading to two different fixtures, and still no evident wetness from the slow but persistent water leak. Model 1 was designed to fit only 3/4" pipe. I had originally considered making each box with two semi-cylindrical indentations, for the two pipe sizes, but wisely started with a simpler design. Another solution would be a thermally conductive adapter bushing in two sections.

In my eagerness to make use of un-evaporated dry ice, I made a "dry box" from corrugated cardboard and bubble wrap. Top flaps fold and lock like a cardboard cat-carrier. Took it directly (as straight as I could crawl) to the work zone and loaded it up with remaining dry ice chunks. Familiar ice plug failed to form. frown
1502522268 2099 FT180082 Model2

Before fetching more dry ice, I made a fancier dry box and lid (model 3) from a laminated sheet of resilient, closed-cell plastic foam. Polyethylene?
1502522775 2099 FT180082 Dscn0134
Set that up near to Model 2 and filled both with fresh dry ice & no liquid. Wait, wait, wait -- no ice plug. Crawled back into the hole & found plenty of dry ice still in each box. Consolidated it into the newer, fancier box. Working in a supine position, in cramped quarters, I felt one or two small bits of dry ice fall into the long sleeve of my coveralls. Had not expected this project's first injury to be a burn instead of a cut or a head-bump.
1502523345 2099 FT180082 Dscn0142

With no evidence of an ice plug, I reluctantly started cutting metal for a smaller version of Model 1. Made some design changes that were supposed to save time in fabrication and assembly, but didn't. It was no big deal to correct the curvature of these half-pipes.
1502523957 2099 FT180082 Dscn0136
At bedtime, with the new "wet boxes" far from finished, I was surprised to find lavatory water not flowing. The plastic foam dry box had done its job, but it took more than an hour. Hadn't expected thermal contact to be that bad, even knowing there would be a film of vapor between the solid steel and CO2 surfaces. After all this trouble, I had one more data point (water meter was not creeping).
Re: A low-cost pipe freezer with limited applications
klugesmith, Tue Aug 15 2017, 06:38PM

Got to chip away at the storytelling.

For Model 1S (small) I finished one box using thin steel instead of copper. Took a while to scrounge an empty tin can without corrugations. Found the metal to be 0.25 mm (10 mils) thick, about the same as a scrap of galvanized roof flashing that I'd previously dismissed as too thick. Lacking experience with bending allowances, it took me a long time to fit the floor-and-walls part. The fancy semicircular joints were eager to leak; wish I'd simply butted flat sheets as in Model 1. Small holes in the copper walls are for threaded standoffs (inside the box) as an alignment fixture.
1502820041 2099 FT180082 Dscn0446
Instead of finishing Model 1S box 2, I made Model 4 box 1, using no copper and no solder. Aluminum half-pipe and an old tin spice box, assembled with 5-minute epoxy. The cured adhesive was more brittle at room temperature than I expected, perhaps because it was old stock or the mix ratio was off.

Dry boxes (Model 2 and Model 3) are in the background of the picture above.

A hybrid of Model 1S and Model 4 was tested outdoors, on a slightly inclined pipe, the very night the parts were made. Note frosty pipe due to huge excess of cooling capacity.
1502820996 2099 FT180082 Dscn0388


That was my first trial using a storable refrigerant.
1502822034 2099 FT180082 Dscn0384
An assortment of duster spray and freeze spray cans were on hand, both R-134a and R-152a (1,1,1,2 tetrafluoroethane and 1,1 difluoroethane).

But first let's deplete a much older container of much lower-cost material: propane, aka LP gas, from a BBQ-size tank so old it has a POL valve. Propane is also less toxic and vastly more environmentally friendly. With a kluged adapter on the upside-down cylinder, I collected about 8 oz (250 ml) of liquid in a glass vacuum flask. Used that to fill the freezer boxes. The level of unused fluid in the thermos, after pipe freezing experiments, dropped by about 1 cm overnight. In following days, it wasn't easy to get rid of lingering gas odor in the flask.
Re: A low-cost pipe freezer with limited applications
klugesmith, Thu Aug 17 2017, 04:46AM

This'll be consolidated into top post when things have settled down.

Dry ice has a couple of problems as a pipe freezing agent. Aside from its limited shelf life, it makes the ice plug and things around it much colder than they need to be.
Here's a chart that showed me how much the heat-removal work goes up as the temperature goes down.
1502942363 2099 FT180082 Pipe Freezing Work
The model is very simple: a section of pipe, the water inside it, and the coolant container. All brought to the temperature where the boiling coolant settles down. Here 81 grams of steel, 12 grams of H2O, and 80 grams of copper.
Dots along the Total curve, to the right of the water freezing step, represent the normal boiling points of two "freeze sprays" (R-152a and R-134a), propane, and CO2. In the last case, we evaporate as much coolant super-cooling the ice plug as we did to form the ice plug. (Of course we could used less coolant & had it boil dry at some intermediate amount of cooling.)

The order changes if we look at the mass of coolant that would be evaporated: 28, 42, 25, and 25 grams. Based on latent heats of 320, 217, 428, and 574 J/g; I'm surprised at the large difference between the HFC compounds.

The chart also shows the significant cost of chilling the metal coolant container.

On propane night I finally made a box of plastic foam, that can hold liquid on all sides of a horizontal pipe. The box is divided in a horizontal plane, and sealed with some buttery hand lotion.

1502944563 2099 FT180082 Cr0409

1502944718 2099 FT180082 Dscn0399

I think Model 5 is the best of the series, by a large margin. Inexpensive, easy to make, easy to use, and more efficient.