Twinsprings Research Institute

I could not get the new Penguin chiller to cool the 50 gallons of water in the old hot water tank. The company was very helpful but also could not figure out why the chiller was not chilling. It actually did chill at night a few degrees but during the day the temperature kept going up. I sent the chiller back with courtesy shipping and when they plugged it in it worked with a 5 gallon bucket of water, chilling it 8 degrees in 5 minutes.

The support person was Eric and he tried to figure out what was wrong with the circulation system by long distance and photos. He could not figure it out either but here are our theories:

1. The pump was not delivering the water to the chiller quickly enough.

2. The chiller was “frozen” inside so could not chill the water–this would be from inadequate circulation.

3. The piping is not allowing for proper flow–possibly entering and exiting the chiller opposite of what it should.

4. The ambient temperature in the room was too high to chill the water.

5. The electrical outlet delivered less than the required amount of power.

6. The coolant was low from shipping damage.

7. Something, the pump or other heat source was adding heat to the system so that it could not properly chill.

8. Something else?

Evidence

1. There is no evidence that the pump was not operating. It was humming away and when I disconnected the pipes they were full of water.

2. Although the system could have been “frozen” it would have stopped the water flow which would have made the pump react and I didn’t see any such reaction.

3. The input and output were clearly labeled, and warm water goes in while chilled water goes out. What I saw was warm water going in and that coming out not even a degree cooler while the next round of water going in was even warmer.

4. The water could have been getting warmer just by being in a warm room–the ambient temperature was about 80.

5. We just discounted this theory, the outlet seems normal–although I didn’t check the power, it is likely the chiller works at somewhat variable amperage.

6. We thought the coolant was low until it was returned and seemed to work just fine with 5 gallons of water circulating through it.

7. The pump may have been adding enough heat to the water alone with the ambient temperature that the 1/2 HP chiller could not overcome that heat. Evidence that at night the temperature did start to go down does support this theory.

8. Something else entirely may be going on. After all the first chiller was also sent back because it was not chilling! Maybe it was just fine too and the problem is somewhere in my system.

Despite the issue that the chiller was not working in my system for only a 50 gallon resevoir, it may be that 5000 btus would do little to chill the mass in the house. Based on the Manual J cooling load, it would take at least 24000 btu’s and then some to make the floor cooler than the air in the room. But that logic might be based on traditional air conditioning, not radiant cooling.

Eric concluded that no matter what the issue with my system his chiller would not work for this application.

However there are other concerns with what you are trying to accomplish. The first concern is chiller sizing, I’m no expert on radiant cooling, or if it’s even feasible. With air conditioning the unit is sized to remove heat and humidity from the air, with a chiller you are trying to remove heat from a large mass, in addition to the air. This would lead me to believe that you will likely need more BTU in a chiller setup than you would with a traditional air conditioner. The next concern would be that you would need to cool your floors to about 10F lower than you want the air temp to have effective heat transfer. With air conditioning the removal of humidity is half the function of the unit and amplifies the cooling effect, radiant cooling won’t remove humidity. When the floor gets cool it will start forming condensation, and prolonged condensation will lead to mold and a host of other problems.

So this argument won the day.  I was the one who could not get the chiller to work in my system and could not find an answer to the problem. I got my money back so I should not complain that my design is not acceptable to a company whose products are not used for radiant chilling. But I now I have to look at what else I can do. I’m tempted to spend the money on a larger chiller, but would have to buy it from Amazon where a return is as easy as from Penguin. I found a 1 1/2 HP chiller for almost $1500! But I would have to buy a larger pump–it requires circulation of at least 10 gallons per minute. Or I could try a small exterior heat pump with a water coil instead of fan unit. There are less expensive split units but would have to figure out how to build the heat exchanger to use water instead of air as mini splits send coolant to an inside unit that cools air instead of water.

But as far as the science of such a system, cooling the thermal mass instead of the air, there are papers that support the idea paired with an enthalpy ventilation system, a dedicated outdoor air system or DOAS. Energy Recovery Ventilators use enthalpy to “condition” the incoming air by exchanging heat and humidity with outgoing air.

Colorado generally has low humidity so condensation on the floor is much less likely. We probably don’t need the mass to be a full 10 degrees cooler than a typical air conditioned inside temperature as Eric believes.  Radiant cooling should make the house more cave-like, a constant cool temp of the whole structure cooling the air inside. The theory of mass is that it has a long term effect so that by the time the heat of the day overcomes the cool of the mass, the temperatures are dropping outside again. This would work because Colorado tends to have cool nights. Eric lives in Florida and such a system would not be feasible there.

From the ASHRAE Handbook for HVAC Applications (which I apparently found posted illegally as the original link was down so this link is to the subscription site) there is this argument for the efficiency of radiant cooling.

One way to reduce cooling energy further along the path to net- zero-energy buildings is thermally active building systems (TABS) or active-core cooling: direct cooling of building mass by chilled water in conjunction with chillers designed for very high part-load efficiency in low-lift operation and enthalpy-recovery dedicated outdoor air systems (DOAS).
TABS differs from passive storage by cooling the mass directly from the inside, thus eliminating charging-mode convective or radiative coupling resistances. By directly precooling the mass, instead of the occupied space, substantially better storage efficiency and larger effective diurnal storage capacity are achieved. Precooling energy percentage savings may be 5 to 35% higher because the condenser-evaporator temperature difference is lower to begin with and because, with the elimination of supply fans and mechanical cooling, energy use is dominated by compressor operation at very low pressure ratios (Armstrong et al. 2009; Katipamula et al. 2010a). Economizer-mode energy, which involves pumps only rather than fan transport energy, is extremely low as well.

I looked up the research that the ASHRAE Handbook cites for radiant cooling and much of it is centered on control strategies to integrate low lift chillers (small differences between incoming and outgoing water temperature) with ventilation controls to increase cooling efficiency. This explanation is from the introduction to a report for the DOE by Katipamula et. al. entitled, “Cost-Effective Integration of Efficient Low-Lift Baseload Cooling Equipment: FY08 Final Report”