HYDRONIC HEATING AUTHORITY
I SET UP MY own unscientific research to compare different insulation products used in snow-melt systems, as you'll recall from last month's column ("Answers to questions on snow-melt insulation," pg. 40). The piping to the 4-ft.-by4-ft. patches of different insulation products included a four-way reversible flow valve.
The application of the four-way valve to control the snow-melt system was so new that I had to develop my own timer from an off-the-shelf reversing relay. It works like a champ.
You can hear when the valve reverses and starts pulling warm water back in past the modulating boilers. They slow way down waiting for the cold water temperatures to come back and, when that happens, they take off like a herd of horses again. You can feel it real well on the near-boiler piping using the ooh-ah test (carefully using your hands to sense the pipe's temperature).
The only problem with the ooh-ah test is that it isn't very accurate. But it sure paints a clear picture in your mind's eye as to what is coursing through the veins of your hydronic beast at that particular point in time.
The instrumentation
The device of choice for this test was the HOBO four-channel temperature recorder that I've written about in previous columns. I had my fellow employees on this job install some capped copper pipes on the ends of spare 5/8-in. PEX tubing, placing one sensor in the middle of each insulation sample, buried in the dirt a couple inches below the insulation. The other ends of these sensor conduit tubes were marked with permanent marker and placed inside a manifold box near the end of the run for that section. They were matched with a sensor that was placed in the middle of each insulation coupon on the cement side of the insulation.
A total of 16 sensors were available during the comparison monitoring the sample coupons, top-of-slab surface temperatures, outside air temperature, supply/reverse flow/return manifold temperature and ambient temperature within the manifold box.
The reason for using the PEX with the bare copper sensing well on the end was that in case of sensor error, the sensor could be easily replaced, and removed at the end of the monitoring period. I secured the sensors to electrician's fish tape and used that to push the sensor to the end of its respective copper well.
I had documented the sensor well's leader length so that I could be sure that I was completely in the sensor well. It's a good thing I had done it that way, because a couple sensors had some anomalies initially and I had to replace them. So much for unpaid research. It pays to think ahead.
The HOBO data loggers were programmed to sample the sensor temperatures once every 60 minutes. This allows better graphics clarity, but is not as data-pointy as sampling every minute or 15 minutes. That is one thing that you become extremely aware of early in the data collection/analysis process — "Man, what am I going to do with all these numbers!?"
In this brief period, I collected 9,000 temperatures and time stamps. Staring at this much data for long periods of time can be rather humbling. Once you've converted all the data to a graph, it makes much more sense.
Remember, this was not a controlled scientific test, but rather a simple sampling and comparison under real time operating conditions.
The results
The "winner" was the thickest material tested. The thicker the material, the more R-value you have, the greater the resistance to the flow of heat.
When comparing materials like this, I think it is a good thing to look at more than just the raw data numbers that get generated during the sample period. There are people who swear by all the different types and brands of below-slab insulation, and there are people who swear at them. The loudest one to cuss is the poor soul who gets stuck with the system with no insulation below the tubing, and he has to break the bad news to the owner about heat not just rising, but traveling omni-directionally through the paths of least resistance, including downward. All the insulations compared proved some beneficial resistance to the downward flow of heat.
There is also the "unknown" factor, that being what would be the effect of no insulation? I witnessed an interesting phenomenon at the end of the test.
Right after I had pulled the sensors out and was cleaning up, I happened to have an IR non-contact thermometer in my hand. The slab surface was 25°F, and the ground was 34°F. (The system had not been running for a week.) I double checked my findings and confirmed them. The slab surface was significantly colder than the surrounding ground. I'm certain that it was the influence-of night sky re-radiation, and it occurs anywhere that there is a break in the earth connection. My gut feeling is that the insulation is worth it in the long run. If it weren't beneficial, then no one would use it, and that is obviously not true.
It makes you wonder about the potential benefits of a large, deep pit with an earth-connected heat exchanger connected to the snow-melt grid and what it could do for maintaining an "earth idled" system. The Japanese have successfully done just that on numerous projects with varying results. More on that in a future column.
Tune in next month for the final installment on below-slab insulation. Until then, happy flexible insulation hydronicing!
Mark Eatherton is a Denver-based hydronics contractor. He can be reached via e-mail at [email protected] or by phone at 303/778-7772.