Sunday, August 22, 2010

Variation in Riser Flow For Wide Solar Collectors

Solar space heating systems often use wide solar collectors in which a large diameter bottom manifold distributes a heat transfer fluid (often water) to a large number of vertical risers.  The risers are equiped with fins to absorb the sun and transfer the heat into the water flowing in the risers.  The heated fluid is collected by an upper manifold and returned to the storage tank.  The collectors are often arranged so the inlet is on one lower corner and outlet is on the opposite upper corner in order to improve the evenness of water distribution to the risers.

This test attempts to assess how even a distribution of fluid to the risers this design achieves.  If the distribution is uneven, then the efficiency of the collector will suffer.  Even though this design is very common, there is surprisingly little data that I could find on the evenness of the distribution.

The picture shows the my 11 ft wide collector used for the test.

Same collector with glazing removed to show manifolds and risers.

The test was done by removing the glazing from the collector, and with full sun on the collector, and normal fluid flow through the collector, the temperature of each riser was measured with an IR gun style thermometer.
The idea being that if a good fluid distribution is being achieved, the risers should all have a similar temperature profile with the lower part being near the tank temperature, and the temperature gradually increasing as you go up the riser.

Riser No 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
top 124 123 123 124 126 129 124 128 126 132 127 125 129 125 128 126 123 123 122
mid 123 122 121 123 124 124 120 125 123 125 125 125 125 124 123 123 123 123 120
bottom 122 117 119 117 119 117 115 119 118 119 120 121 119 117 117 115 116 120 115
Time 12:47 PM 12:56 PM
Temp rise 2 6 4 7 7 12 9 9 8 13 7 4 10 0 8 11 11 7 3 7

The truncated table above shows the results -- see the full report link below for the full table -- there are 20 risers in the full table.

Happily, the distribution is pretty even, with most risers showing similar temperature patterns.  There are variations, but probably not enough to seriously effect efficiency.

It does appear that the evenness of the distribution can be effected by relatively small changes in the details of how the plumbing is done.  Have a look at the full report for more details on this.

All the gory details on riser flow distribution test...

If you have additional data on riser flow pattern tests or analysis, please let me know.



  1. I think that your best approach would be to drill a hole in each tube and run a line to a Pitot tube. Here is a link to a google query for Pitot tube construction:

  2. Thanks Stuart.
    I was hoping to find a way that would be somewhat less work, as there are 20 risers on my collector and 46 on the Thom's collector that brought this issue up.


  3. Hi Gary

    The bernoulli effect of water flow causing pressure differentials across the array of risers can cause significant difference in flow rates in each riser tube. The rate of this imbalance is determined by the flow restriction in each riser and he flow rate through the header tubes. If the restriction in each riser is higher than the pressure differential then much of the effect is negated. However if your riser tubes are say 15mm dia and less than 2m tall then they dont have much restriction and thus small pressure differentials will cause the flow rates to vary or even reverse along a long length of header array.
    I have observed this first hand when building some of my early arrays. If I welded just the bottom header tubes leaving the top off and fed a hose in the bottom, then adjusted the tap so the water was streaming a few cms from the risers, then some further-est from the connection had significantly more water flowing out the top. This was in an array of 12 risers 1.5 m high. If I was to place 3 of these arrays in series then I would expect the risers in the center to get hotter as the flow rate would be less or non existent. A few taps with the hammer to restrict the offending tubes cured the problem. Making the height of the riser tubes 2.4m negates the effect somewhat. But in a large array I would be very careful to get the flows balanced or loose efficiency.

    In non pumped systems then the above really doesnt apply, the water flow rates are not high enough to cause any significant pressure imbalances.

    Mike (NZ)

  4. Thanks Mike -- good observations.



Note: Only a member of this blog may post a comment.

/* Start Analytics ---------------- */ /* End Analytics ---------------- */