Low pressure burner testing

The first question I had when designing my brewing system was what type of energy to use. My first choice was steam, but that was a pipe dream (pun intended). Next was electric for a variety of reasons, but that option required me to upgrade my electric panel, which was cost prohibitive. That left me with either propane or natural gas as options. I decided on natural gas because it is more economical, the idea of refilling tanks regularly was not appealing, and I could easily see myself running out of gas in the middle of a brew session.
An initial test run on the system produced poor results. After nearly 2 hours of heating, it hadn’t reached a boil. Some testing was needed to increase the performance of the burners (a link to the spreadsheet data is at the bottom of the page).
The areas to be tested were:

1. burner height – the distance from the burner to the kettle
2. heat transfer – using a more conductive metal to transfer thermal energy from the flame to the kettle
3. ventilation – insufficient ventilation chokes the flame
4. gas pressure – the burners were designed to work optimally at 11” WC of pressure, while most utility companies deliver gas to residential buildings at 7-9” WC.

For all tests, 10 gallons of 75°F water was heated on a single 10” low pressure burner, and temperature readings were taken every 5 minutes.
burner height

The first tests (test #1 in the chart below) involved adjusting the burner mount height. The heat shield/burner mount has a split open back to allow for rear ventilation above the burner, and comes with 3 sets of holes for adjusting the distance between the burner and the bottom of the kettle. The distance from the bottom hole to the top hole is approximately 1”, and while there was a slight difference in performance between the top mounting hole and the bottom, the results were not significant (which is why only one of the tests is charted). After 50 minutes, the best configuration was at 170.7°F.
heat transfer
I consulted my friend/assistant brewer, Eric, who is a mechanical engineer by trade, and for test #3, a homemade heat sink was built out of a thick plate of aluminum and aluminum angle iron to capture more heat from the exhaust and transfer it to the kettles. This increased heat transfer, but after 50 minutes the water had only reached 187.3°F.
ventilation
The next test involved creating a chimney (test #4). Just like with a house chimney, less dense hot air rises and is replaced by more dense, cooler air. This is referred to as the “stack effect”. The greater the thermal difference and the height of the chimney, the greater the stack effect.. We built a semi-enclosed chimney out of a 90 degree register box and sheet metal to force air to flow up from the bottom of the burner and out of the back. We tested it with and without a 6’ chimney (that’s not a typo, it was 6 feet tall), but both tests resulted in lower temperatures than with the aluminum heat sink.
gas pressure
I had the gas company come out and test/adjust the pressure to the house. I was at 7.3” WC at the main line to the house, and 6.7” WC at the line to my brew rig (acceptable range is 7-9” WC, according to the technician). He adjusted it up to 7.5” WC at the brew rig, which is the highest he was allowed to adjust it, but then volunteered to show me how to adjust the pressure myself “just in case I wanted to know how it works.”  I increased the pressure to approximately 8.6” WC, switched out the flex lines to the burner for larger diameter flex lines that were previously on order, and then ran two more tests.
Test #6 was with both the aluminum heat sink and the chimney at the new increased gas pressure. After 50 minutes, the temperature was 182.6°F, just shy of test #3, which took place before the gas pressure to the house was increased. This suggested that the chimney setup wasn’t helping.
In test #7, the chimney was removed and only the aluminum heat sink was used. At 50 minutes I was at 206.7°F, and at 55 minutes I was boiling.
Test #7 was by far the best results out of all of the tests, and is what I chose to stick with. A spreadsheet with the full test results can be found here. Below is a chart showing the performance over time for the tests mentioned above.