Bad experience with evacuated tubes.

[mkel]

10 month followup on evac tube repairs -- some lessons learned ---

OK ... it was just about a year ago I started repairing my broken heat tubes via acetone solder method. I left my original post below if you are unaware of what I had done.
I noticed it worked great all summer but my temps started dropping off about 2 months ago.... so I pulled a few random tubes to see what was up.
It appears I have had a total failure of all of the repaired tubes .... but it isn't all bad news ... in fact some good came out of it.

There was a lot of speculation that the solder would melt at the high temps inside the glass tubes... I found no failed solders.
all the tubes I pulled had lost the acetone and the copper was black from heat but all of the solders were fine and held on a vacuum test on the workbench.
Here is what did fail ... all of the refrigerant schrader valves.
For those who didn't know what I had done to repair the heat tubes.....
I had soldered schrader valves to the bottom of the tubes and added the charge through them and then heated the tube unit I got some vapor then closed the valve and put the valve cap on.... and at this point (one year later) it is totally obvious the schrader valves can't take the heat. I pulled the internal part of the schrader valve and the bad gasket out of the cap and recharged the tube putting the cap on tight with teflon tape. I have 2 tubes in the yard cooking for about a week this way and they are bloody hot.
But this isn't a solution I am comfortable with... I will try to find a solder cap that will fit my heat tube copper ( looked before and wasn't successful ) and if I fail to find any, I will try to charge a tube and use a refrigerant crimper at the end of the heat tube and then solder the end shut.

You will notice in the pic in my original post of the schrader valve I had put that if it didn't work I could solder the valve cap on ...
well I tried that and wasn't able to get it to solder very well.
possibly because I wasn't able to get the threads clean enough after a year in the tube.


Lessons learned ...
#1. the solder will hold even if the charge is lost and the tube cooks with no way to dump heat.
#2. Refrigerant valves can not take the heat inside the glass tubes and will fail.
#3. in theory and testing the acetone works great and produces excellent heat at the bulb.
#4. If you are doing one of these systems DO NOT put the tubes on your roof ... they are a bitch to take down and put back ... my biggest mistake in this whole experiment.

I will provide another update when I get a permanent solution to the problems I've had with the tube repairs.

Mike

I've been experimenting with evac tubes for a few years now. Have 80 tubes on the roof and last year I started to have heat tube failures. I want to add some comments about my experiences and maybe they will help someone else.... some of these comments are my opinions with no scientific tests.
Please don't rip me if I miss-spell something --

1. I read where a infra-red heat detecter could be used on the glass tubes to determine if a heat tube had failed inside. I did a reading on all 80 tubes and wrote down each reading. there were a number of tubes that read 4+ degrees higher than others (supposedly indicating the heat tube is bad). When I pulled the tubes I found no correlation based on this method.

2. My ratio of failed heat tubes was about 40%. They had split and I do agree with statements in this forum that this was a product of freezing.

3. Found a number of tubes that had NOT split yet were expanded at the bottom.

4. My supplier offered to replace all my heat tubes however I declined because I have no faith that the new tubes will have anything different in them and would just become a future failure. I have repaired all 80 tubes by soldering a refrigerant access valve on the bottom of the tube. Pulling the schrader valve out I can charge the tube with fluid and then replace the valve. Heating the tube I depress the valve until I feel a bit of moisture from the fluid. At that point I close the valve and replace the valve cap.
These tubes work great and because of the valve I was able to experiment with different mixtures and quantities to find the best ratio for my tubes.

5. In this process I cut all the tubes that had not burst. I found that every tube contained a different amount of fluid with a variance of as much as 8cc. Some tubes had almost nothing.. maybe 2cc. The fluid appeared to be dirty water and I suspect the copper dust is not a product of engineering but residue from poor manufacturing of deburing or pipe prep. The amounts of residue went from a trace to quite a bit and some had none.

6. SUMMARY -- I believe the tubes that split had much more fluid (water) in them than the ones that didn't split. the pipes that were expanded had more fluid in them than the ones that looked normal.
It appears the manufacturing process is sloppy at best. The random liquid levels in the heat tubes may explain why the infra-red heat reading didn't work. In experimenting with different amounts of fluid in the tubes I found a wide swing of temperatures depending on the amount. I lined up 8 tubes with different amounts of fluid from 2cc to 14cc. This provided interesting results and I found 7cc to be the perfect amount for my pipes. Less than 7cc worked well in direct sun but at 7cc I got some heat even when it was cloudy... only dropping off when clouds were very thick. Solution of choice turned out to be acetone. Nothing else I tried provided as good of results... and it won't freeze. There is ZERO chance of a failed acetone heat tube contaminating the water circulating in a evacuated tube system. Also for those who are paranoid of explosion... I put one of these acetone heat tubes in my burn barrel and cooked the crap out of it with no explosion.. solder did melt and acetone escaped with no event noted.

Hope some of this info helps others. It is a shame the manufacturing of these tubes has left the customers with this problem. Worst part for me was the huge number of trips up and down from the roof removing and replacing the tubes.
[ATTACH=CONFIG]1546[/ATTACH] [ATTACH=CONFIG]1547[/ATTACH] [ATTACH=CONFIG]1548[/ATTACH]
mkel

[inetdog]

Temperature differences from one end to another in a heat pipe.

I am a fizzycist by training, but not an engineer (and I can spel) , and here is what I have learned about heat pipes:

If a heat pipe is working properly, the only gas inside the pipe will be the vapor phase of whatever liquid is the "working fluid" of the pipe. The vapor pressure under these circumstances will be dependent only on the temperature of the inside wall of the pipe. And since the pressure is constant from top to bottom of the pipe (there is not a strong wind blowing from one end to the other inside the pipe!), the temperature must be the same at both ends *as long as the system is in equilibrium*.

It is somewhat like a pot of boiling water: The water temperature is almost identical from the heated inner surface of the bottom of pot to the top surface of the water. When you apply more heat on the stove, the water boils faster but the bottom water is still not any hotter than the water at the top. (OK, there will be a small difference in pressure from the bottom to the top, since water is fairly dense. But we can ignore that effect in the heat pipe since it is mostly full of vapor.) And the transfer of heat from bottom to top in this case will mostly happen by convection which is stirred by the boiling effect. In a heat pipe the fluid is boiling at the bottom, the vapor moves freely up the pipe, condenses at the top, and runs back down to the bottom as liquid. The is referred to as "mass transport", and is more efficient than either convection or conduction.

You may get some temperature difference between the outer wall of the bottom of the pipe and the inner wall if you are adding heat to it fast enough. And the same a the top end where it connects thermally to the manifold and you are removing heat from it. But that should be only fractions of a degree.

If there is a significant difference in temperature along the heat pipe from one end to the other, or from the top of the heat pipe to the manifold, then is it a sign that there is more than just the liquid and its vapor inside the pipe, and that is interfering with the movement of vapor from the bottom of the tube to the top. If there is no liquid left inside the tube, the temperature difference from one end to the other will be even greater.
So a heat pipe can fail either by air getting in (partial fallure, the more air the poorer the performance) or by the water all leaking out as vapor (at that point you just have a copper tube, which will still conduct heat from one end to the other, just not very well.)

Two of the many factors in the choice of working fluid are the latent heat of vaporization, which is very high for water compared to most other useful fluids, and the useful range of the vapor pressure versus temperature curve. A liquid with too low a boiling point at atmospheric pressure (lower than the manifold temperature, for example) will have to operate at a potentially dangerously high pressure inside the heat pipe

End of lecture, sorry about that....

[MikeSolar]

The liquid inside the tube refluxes (evaporates/condenses) in the tube - if there is a hole in the tube all will escape and no reflux action - end of game

Air diffuse through copper? Not gonna happen under these conditions - that 'theory' hasn't checked in with science or practical knowledge - if air does diffuse through copper does then cancel all refrigeration units -heat pumps etc.

I know this is an old post, just working my way through

No refrigeration system would work for long if ANY vapour went through a copper wall. Period. Also, every hot water heating system would then have O2 dissolving through it and rotting out the pumps etc. Ain't gonna happen.

[MikeSolar]

[QUOTE=Art VanDelay;37847]This thread contains a juicy irony:

I can't find a US engineer selling heat pipe evacuated tubes who knows for sure how they are freeze protected. I've gotten a few folks to speculate, and I don't mind speculating either. Pointed questions to the Chinese manufacturers result in the "trade secret" answer.

A few things are pretty clear, however:

1. Quality problems or design errors have resulted in a high historic heat pipe failure rate.

2. Since some heat pipes work and don't freeze break, it is possible to make the technology work for North America.

3. All the large Chinese manufacturers use water in the heat pipe as the working fluid.

4. Overfilling of the heat pipes with water can cause freeze breaks.

5. Not all of the heat pipe failures are from freezing. A high percentage of pre-2006 Apricus tubes stop functioning without any external evidence.

So, be sure you get an enforceable warranty from your supplier. Better yet, get the supplier to explain how their freeze protection works and how they maintain the quality to ensure they are manufactured correctly.

For you "Evacuated Tube Geeks", here's a quote from a forthcoming NREL report on the topic of Chinese collectors:

"Production and use of evacuated tube collectors is growing worldwide; the nature of glass tube production favors high degrees of automation. In the late 1970

[daves]

Curious Daves, have you adjusted your output calculations for area?

That is not my site. I agree the flat panel has much bigger aperture size. But the owner compares output per area taken on roof.

Once vacuum tubes will be 100% working for many years and you have enough space, then vacuum tubes are better choice.

Since vacuum tubes are still unstable (blow HP, loss of vacuum, high price) its still better plat panel imho. Btw, my choice is UP, but I think flat panel is better anyway.

[tandrews]

Curious Daves, have you adjusted your output calculations for area?
That is, for a given flow rate through both types are you calculating your output temps (edit: should have been BTU) for the collector area?

You note your choice was based on marketing so without this adjustment it would not be a fair comparison right?
It looks like the evac system is not even half of the flat panel considering the exposed area of the internal evac glass.

[daves]

I'm interested in the heat pipe collectors because I think they would be more reliable in a cold-climate recirculation or thermosiphon freeze protection strategy.

Heat pipes has too many risks. Freezing water in heatpipe, lost vacuum in the glass tubes, heat transfer loss between the head of HP and the manifold. The best choice is flat collector, even if it has poor performance in winter. If you have enough space, you can make many flat collectors, make much bigger aperture size to cover the heat loss and get some heat from it.

See http://cwc-das.com/logger/meters/sol...?genimage=true (the picture is rendered every time with actual temperatures) - In this case I do think flat collector has bigger aperture size and vacuum tubes could be gone (like experiences from this thread). Anyway, good to watch.

U pipe is imho kind of compromise. Vacuum is in the glass like in thermos bottle, not in tube like heatpipe has. So less risk with vacuum loss. Also U pipe is filled with antifreeze fluid used in system, so it can not freeze. Also it is directly connected to the manifold, so the are no loss due to oxidation or some plays like on heatpipe. Finally it has round shaped aperture, so it covers much bigger size than common heatpipe.

I wanted to buy heatpipe vacuum tubes from china, but i give it away. To many risks for too much money. U pipe has much bigger chance to work in long term.

[Art VanDelay]

U pipes are fine with antifreeze

I'm interested in the heat pipe collectors because I think they would be more reliable in a cold-climate recirculation or thermosiphon freeze protection strategy.

For more information on that point, see http://www.heatinghelp.com/forum-thr.../direct-system

[daves]

water in heatpipe

I have email from china manufacturer, where he wrote they use clear water in heatpipes since 2009. For extra 2$ per tube they can add copper powder to make "antifreeze". He agreed they have still problem with freezing.

Then I found this forum, while looking if its true. So it is.

Maybe instead of spending lots of money for diy heapipes, isnt the U pipe better way with less risk ?

[tandrews]

I'm not yet convinced that you had dieseling or autoignition. Did you see flame marks or other evidence? All the other phenomena can be explained by the sudden release of a compressed gas out of the bottom tip. But if you had a flame, this is the answer to a question I posed above, which was why don't the manufacturers use acetone?

I'm not convinced either, nor am I trying to convince anyone. Simple observation and hypothesis.
I don't agree though that simple gas release was the cause. I heard the tube when it failed, the noise travelled through the wall/siding of the house, two layers of R5 silverboard and foil bubble insulation and there were still two distinct bangs. One when it popped and one when it hit the roof I'm guessing. 63" of 3/8" copper hvac, 8' of copper wool and an end cap left the evac tube all at once. I know the tube would act as a rocket launcher, but it wasn't a "foomp", rather a bang and that's a fair bit of mass to move. That, and the tube is not connected to the wrap, rather sits on a bed of it at the bottom of the evac tube. (lurkers please don't drag air testing into this)
I think I suggested it was likely liability that prevents acetone use from manufacturers. Also theory.

I would not expect flame marks from a stoic ignition as there are no flames. Just kablooey and it's out.
I didn't see any, but they would be at the bottom of the tube where it's dark anyway.

A few photos would be awesome.

I understand, pix or STFU...
Not high on the todo list, but may happen.
Sit by.

As for whether the top should be ever hotter than the bottom, the top is always being cooled by ambient air convection. There is a little air convection at the bottom, but that air is much hotter than ambient.

More likely zero air convection in these instances due to the wrap, but the top of the heat tubes sit only a couple inches from the end of the evac tube in the heated air that does escape.
I think I had leaky heat tubes. If I get failures on top when it's all insulated I'll change my theory. I hope to put all that energy in the tank though.

But since we were talking about safety, once you install these babies in the manifold, you've sealed the top of the vacuum tube. If you stop all the hot air from convecting out, then everything has the opportunity to get even hotter. But now the failure mode isn't copper wool being blown out or a copper bottle rocket....... well, we might not be as amused by your next report, and "engineering failure" would be the correct description.

Nice seqway. Someone was amused?
My goal is to transfer ALL the heat through the tubes to the manifold and on to the milk tank. I don't plan to seal the evac tube ends however, rather place manifold and insulate around it to allow air heat to transfer to manifold as well (a loop of copper with manifold tubes soldered on). Hotter is better as long as I collect it.
I can't have an engineering failure if it wasn't engineered. DIY experimental would be more apt. (Easy mods... not a secret and I'm sharing)

Like I said, if this was a simple temp desolder tube fart scenario I would have more than two failures by now.
In order to to ensure tubes had zero O2 in them, one might have to heat inductively in a nitrogen atmosphere and hard solder.
Quality control requirements, liability, expense - all good reasons not to do acetone commercially and water is a better transfer medium.

[Art VanDelay]

I'm not yet convinced that you had dieseling or autoignition. Did you see flame marks or other evidence? All the other phenomena can be explained by the sudden release of a compressed gas out of the bottom tip. But if you had a flame, this is the answer to a question I posed above, which was why don't the manufacturers use acetone?

A few photos would be awesome.

As for whether the top should be ever hotter than the bottom, the top is always being cooled by ambient air convection. There is a little air convection at the bottom, but that air is much hotter than ambient.

But since we were talking about safety, once you install these babies in the manifold, you've sealed the top of the vacuum tube. If you stop all the hot air from convecting out, then everything has the opportunity to get even hotter. But now the failure mode isn't copper wool being blown out or a copper bottle rocket....... well, we might not be as amused by your next report, and "engineering failure" would be the correct description.

[tandrews]

Russ's suggestion prompts another possible cause for my failures.
If I had a poor solder seal, however small the leak, the cycling of pressures night to day would vent acetone and suck air until stoic conditions were met.
Since they didn't fail immediately upon installation this may be an equally likely scenario.

I'm torn on the hottest part of the heat tube.
On one hand, I thought the amount of copper wrap around the heat tube may have prevented cooling air from circulating, creating a more then others sealed oven at the bottom of the evac tube. Maybe.
On the other hand, the acetone should boil off from the bottom transferring heat to the top. (Where manifold would take it away if installed, air since it wasn't)
On the third hand, acetone will only boil until heat tube pressure tells it to stop, then the fluid will rise in temp. No air, no autoignition.
On the fourth hand, I may have had a poor solder seal - leak, suck, stoic, flashover igniting tube gas leakage.
On the tail, there was ignition, so regardless of failure mode, evac heat tubes can reach acetone autoignition temps.

One tube did not launch but blew out copper wrap. So - autoignition of the gas mix in the evac tube.
Since the other launched with relatively high energy, it likely dieseled once it reached stoic (inside heat tube) and sufficient pressure due to temp.

Any way you view it, 2 out of 35 is not a bad failure rate for a newby solderer/acetone air evacuator.

I did think about how to test these heat tubes after build, but couldn't come up with a practical test to ensure they were sealed.
8ccs of acetone might be difficult to differentiate in weight from extra solder, or variations in tube length regarding a weight test.

[russ]

All joints leak is a generalization but not far off. Over the years I worked with the direct reduction process (of iron ore) we tried to minimize hydrogen leaks to an acceptable point. Trying to stop all leaks was a fools game - a lot of work with no success.

Of course the plants were outside so if anything leaked it went up-up and away. In a confined space H2 can cause exciting moments. You have to consider that static electricity is more than enough to ignite hydrogen - not every time but often.

[Art VanDelay]

"It's a serious problem - but rather a shame - that Apricus was't able to supply quality heat-pipes. The heat-pipe replacement under warranty caused extra cost that we couldn't accept."

"Air diffuse through copper? Not gonna happen under these conditions - that 'theory' hasn't checked in with science or practical knowledge - if air does diffuse through copper does then cancel all refrigeration units -heat pumps etc."

Well, now we know Apricus has had an unacceptable number of dead heat pipes. If I were forced to reach a conclusion, I'd blame leakage, not diffusion. According to Refrigeraton Technologies, "all joints leak":

http://www.refrigtech.com/Product/Manuals/LeakDetectionManual.pdf

[Art VanDelay]

The top is cooler in your case

If you hadn't yet assembled the header, I would think that the top is the coolest spot on the heat pipe. It's losing heat to the air while the bottom tip has more insulation around it.