Continuous blending of trimix

n2diving

We have three different compressor systems that are configured to make Trimix. The process is this....

Our best practice is that we start the system and begin making 36% with a membrane until everything is warmed up and stable this step takes a few minutes... then we turn down to 32% until it's also stable - this step take just a minute or two... then we reduce the volume of 32% and turn on the helium flow... We use a Dywer flow meter supplied by 15 psi He, which is delivered using an extremely high quality gas regulator specificly designed for He on a 2650 supply cylinder... the regulator quality is critical and was the source of several problems until we used a proper He specfic high quality regulator.

By trial and error we have found exactly the setting for 32% (we adjust this using an LP compressor psi setting that feeds the membrane) and flow of helium... the 32% and He flow rates will be a function of the HP compressor output volume and is very consistently repeatable.... for our 8 cfm compressor the He flow is about 60 scfh and for our 18 cfm compressor its about 150 cfm (from memory)... note we are using Dwyer flow meters calibrated for air, we have no idea what the true flow rate of He really is... the flow meter just gives us a repeatable way to monitor and control the rate of flow.

As you adjust He flow rate, you enter a feedback loop where the 32% flow rate must also be adjusted so that the total flow rate of gas entering the compressor intake is exactly equal to the output, and we have a small pressure gauge at the input manifold on the high pressure compressor to verify no more than 1 psi of back pressure.

On the HP output side we have a regulator with a flow restricted orifice that feeds a trimix analyzer.... here we can tell what the exact analysis of the output gas is moment by moment... as the environmental conditions change i.e. compressors warm up, weather changes, etc.... you will have to tweek the 32%/He flow from time to time through the day...

Initially it can take some time to learn what variables affect the mix but once your system's specific behaviors are well understood, they are extremely repeatable... We've learned that we just turn everything on at once, set the pressures and flows up cold and within a very short order the mix will settle down to exactly what we would expect... one or two tweeks a few minutes later and it will be remarkably stable.

We have found that keeping the nitrox supply mix stable at 32% is the best approach... this allows us to make 21/35, 18/45, 15/55 just by varying the amount of He.... However, we most routinely make 21/35. There is no detectable He loss through the compressors and we can empty the He supply cylinders to 25 psi with ease.