The Bubbling Barometer

I like brewing beer.  Anyone who has watched home-brew beer fermenting will be familiar with the 'bloop-bloop-bloop' sound effects of carbon dioxide escaping from the brew keg through a brewer's airlock.  One day whilst watching a merry brew a'brewing, I pondered if one could make a highly sensitive barometer by using a really large empty keg (say of a few hundred litres capacity) and by allowing just changes of ambient air pressure to cause air flow through an airlock.  After several months of 'thought experiment' on how one might construct such a barometer, fate stepped in one morning when I walked into our laundry to find our 10-year-old 400-litre mains-pressure electric hot water cylinder leaking water all over the floor (400 litres ≈ 100 Gallons).  I'm pleased to report that our highly effective local plumber had the worn cylinder replaced within a couple of hours, leaving me with with a large thermally insulated steel pressure cylinder, precisely what was required for my much-pondered barometer project.

The first problem was to find the leak.  After partially repressurising the empty cylinder with water via a garden hose, I could hear air escaping through the pin-prick hole.  By cutting away the cylinder's outer steel jacket with an angle grinder, I located the offending hole and repaired it (using solder and my large 250-Watt soldering iron).  After repairing the hole I patched the outer cladding with tape.  Then after removing the four port holes on the cylinder, I left it in the sun for a week, to drain and to dry completely on the inside.

As any student of chemistry knows, PV=nRT, which is a formal way of stating that for a closed system, pressure is proportional to absolute temperature, and inversely proportional to volume.  To have the cylinder operating as a barometer, where gas volume is proportional to pressure only, it is essential to have the system held at constant temperature, say better than ±0.1°C per day.  But where does one find such extraordinary temperature stability?... the answer lies right under one's feet, so the next task was to bury the cylinder into the ground, deep!

A 400-litre electric water cylinder is a cumbersome and heavy (≈110kg or 240lbs) piece of hardware, and moving it around and burying it deep-down presented a few minor problems.  Fortunately it just squeezed into our Ford Laser station waggon.  Even more fortunately I already had a made-to-measure hole in which to bury it.  Our neighbour had kindly dug us a large hole for future use as a rubbish pit, but these plans now changed, and this large hole became the final resting place for our dearly departed hot water cylinder.  After a brief burial ceremony (pictured right) attended by baby Max and grieving relatives, and including a brief reading from the Good Book*, our cylinder was finally laid to rest some 1.5m down (* The CRC Handbook of Physics and Chemistry, 63rd Edition).

With all cylinder ports securely capped, and an air-line attached, I commenced a partial burial so that I could assess the behaviour of the air flow in/out of the cylinder as atmospheric pressure changed.  As the the cylinder cooled and assumed the temperature of the surrounding soil, air was predictably sucked into the cylinder via the air hose, and the airlock bubbled (the airlock is actually filled with olive oil, as I do not wish to humidify the air within the cylinder).  Upon reaching a moderately temperature stable state, the airlock ceased bubbling and the liquid level came to be affected by changes in ambient air pressure only.  By keeping an eye on the ambient air pressure which is logged hourly by a local Bureau of Met Automatic Weather Station, it became clear that there was a very close correlation between airlock fluid levels, and the BOM readings.

The first test of the Bubbling Barometer came a couple  of days after I buried the cylinder, when I noticed on a satellite image (left) that a weather front appeared to be moving toward us.  The corresponding Mean Sea Level (MSL) pressure chart forcasted falling local air pressure throughout the day until the trough was encountered, and indeed between 9am and ≈3pm we observed a steady fall in air pressure of around 6hPa (ie. hectoPascals, which are equivalent to milliBars)  Needless to say this fall in pressure caused the barometer's airlock to start bubbling as air was steadily sucked out of the cylinder.  And it was no subtle effect, with a bubble passing through the airlock every few seconds.  So if one did not have access to real-time satellite images and MSL pressure charts, and had to rely on the barometer alone to forecast approaching weather, the Bubbling Barometer gave an excellent indication that a significant weather change was approaching.

I was curious to know just how temperature-stable the buried cylinder really was, located at its mean soil depth of 1-metre.  Before burial of the cylinder I placed a 50mm dia PVC pipe (see right) down into the cylinder pit so that later I could lower down a thermometer to check the soil temperature.  Recently I purchased a small temperature logger from Jaycar.  This device, looking like and overgrown lipstick container, is perfect for the task, and the first results are shown on the plot to the left (click on plot for full-sized version).  These data, showing around four days of 1-hourly temperature measurements, indicate a very slow increase in soil temperature at 1-m depth (≈1/6 °C per day).  The data logger has a measurement resolution of ±0.5°C, and within this accuracy there is not the slightest indication of a diurnal (i.e. daily) variation.  Which is what we want to see, and got me thinking further about our local soil temperature profile.

Having established the that barometer worked as intended, the next task was to bury an air-line up to our house so that the airlock display could be mounted somewhere convenient inside the house.  The distance between the buried cylinder and the house is about 100m (≈330'), and for this link I used 4mm* polythene tubing, of the kind commonly used for garden watering systems (*ID=4mm, OD=6.2mm).  To protect this thin tube from damage I placed it inside 13mm polythene pipe.  These polythene pipes, and their various joiners and fittings, are surprisingly cheap (eg. 100m of 13mm tube= AUS\$27).  The only big issue with laying the air-line was the required 100m of trenching.  We are located in Scribbly Gum (Eucalyptus Rossii) open woodland within a bushfire prone region of NSW contiguous with the vast Pilliga Forest, so burial of the air-line to a depth of 100mm is advisable.

The volume of air contained within the 100m of air-line was a concern, given that it is buried near the surface where it is not in a temperature stable environment.  It's a concern until one considers that the total volume of this thin air-line is 1.25 litres, which is negligible compared with the buried cylinder's 400-litre volume.  On the subject of volume, if the ambient air pressure changes by 1-hPa, what would be the volume of air that would flow in/out of a 400-litre cylinder?  The elevation of our house is 541m (1774'), where the mean air pressure is about 950hPa (compared with standard sea-level pressure of 1013.25hPa).  Since volume is proportional to pressure at constant temperature, then a change of 1-hPa in sea-level pressure would displace around (400/1013)×(950/1013)≈0.37 litres of air in/out of our 400-litre cylinder.  So a modest air pressure change of say 10hPa would displace around four litres (≈1 Gallon), which explains why the airlock display generally starts bubbling rapidly when a significant change in the weather is occurring.

The buried cylinder is located in a very temperature stable environment, but if for example the temperature did change by 1°C, what volume of air that would move in/out of the cylinder?  The absolute temperature of the cylinder is currently around 290°Kelvin, so a 1°C change of temperature would displace around 400×(1/290)≈1.4 litres of air.  This is equivalent to the amount of air that would be displaced by a 3.7hPa ambient air pressure change, which reconfirms the need to hold the cylinder temperature very constant.

Eventually I made a nice job of mounting the brewer's airlock on an attractive wooden base made of Tasmanian Oak , and this simple instrument now takes pride of place on my wall of meteorological instuments.  Compared with all other pressure measuring instruments I own, it provides the quickest visual indicator of what the air pressure is doing; rising or falling, and the rate.

The liquid I chose for the airlock is glycerine, and a 1hPa change in air pressure causes the fluid level [between the two bulbs] to change by 8.1mm.  It takes a pressure change of around 5hPa to get things bubbling.  To smooth out the rate of bubbling, so that there are many small bubbles rather than a periodic monstrous "glurp", the rate of airflow through the instrument is controlled by a hyperdermic needle placed through the rubber bung at the top of the airlock.

'H' and 'L' labels indicate a Higher or Lower pressure tendency.

One curious feature I have noticed during times when the barometer is bubbling, is that the bubbling may periodically stop for a few minutes, and then resume.  What I think is causing this effect is the passage of microbaric pressure waves, which present a tiny pressure moduation on top of the general barometric trend.  Which has now got me thinking about the measurement of microbaric variations, and their causes... and which may well spawn a new instrument when I can think of a good design.

If you should have inclination (and space!) to build an instrument similar to this, please feel welcome to contact me with any queries.