Homespun multi-pass convector

Gordan

Nearly a year ago I was soliciting feedback here on an idea I had of stacking several old convector elements I have on hand, and using a few 5" computer case fans across the top to create a low-water-temp convector. This weekend, I finally got around to rigging this up and testing it, and the results were encouraging.



Basically, the beast is made from three 3' x 6" deep three-tube convector elements, piped in series and with supply at the top and return at the bottom. I made a box and baffling from polyiso board to enclose the elements, and placed four fans into tight-fitting cut-outs on top, with about 2'5" space between the top of the top element and the fans. The effective length of each element is probably about 30". I piped this in series with the heat source (TT PS-60) with an Alpha as the circulator. I then throttled the flow until the Alpha just began indicating 2 gpm. The inlet air temperature was 76 F, and the difference between the peak supply and the peak return was 11 F (132 F - 121 F), which would give close to 10kBTU/hr. This was at high fan speed - these fans are rated at about 80 cfm each free air, for what that's worth, which is probably not much. 



My goal was to meet a 8kBTU/hr worst case load. Now, to do it with this single element, I had to run the fans at a higher speed than I would ideally like to, so I think I'll follow my original plan and build another one using the three 6' elements I have, which should give me ample headroom to crank the fan speed down and still easily meet my worst case load.



One interesting thing I noticed is that even though the PS-60 has a high volume heat exchanger, it doesn't seem to much like low flows, even at lowest modulation. The boiler's indicated supply temperature would climb rapidly to 150 F even though the supply temp I measured at the convector inlet peaked at 132 F (measured using a contact probe directly on the middle tube at supply end.) This was at 2 gpm.