Error sources and analysis
As mentioned in other sections, the mixing methods accessible to technical divers cannot have the same accuracy as industrial methods; a first source of possible errors is already in this assumption.
The key is therefore to find a system that is an acceptable compromise between the need for convenience and accuracy that is - if not at laboratory level - not detectable with the measuring instruments and analysis at our disposal.
Gas mixing is an expensive and time consuming process; having doubts about the content is even worse, it can get frustrating and expensive.
Then, one of the worst things that can happen is having doubts during the dive, such as thinking back to the analysis and wondering whether it was reliable or not and questioning it etc.. etc..
For this reason, it is worth finding the right equipment and dedicating the right time and attention to the blending process with discipline, without taking anything for granted.
The final accuracy depends on several factors. It is important to have a good understanding of the various sources of errors and their possible impact on the final composition, in order to avoid or mitigate its effects.
Let’s start by asking ourselves some questions:
What might be the sources of errors? How much do they weigh ?
Are they as predictable as the sign and dimension ? Can we isolate and avoid them ?
Some errors are due to a lack of information, others to needs in terms of productivity, others to negligence, others to inadequate instrumentation. Some are very simple, others are more difficult to eliminate.
Lack of attention in the process of blending
Overconfidence is often a problem, and blending is no exception.
For example, it can happen that the lines are not flushed properly: hips, piping, booster, compressor etc.
A typical critical situation is when you add in sequence Helium then Air.
The practice of switching from one gas to another without stopping the compressor/booster and flushing the lines with the new gas, i.e. by changing the gas supply on the fly, is a frequent habit, certainly very practical but also quite uncertain.
The problem is that it is impossible to give a dimension to this problem. It is almost impossible to know how much of the old gas remains before being replaced by the new one.
If you are looking for the highest accuracy, this point must be respected: flush the lines with the new gas.
Another example: you can have a great Oxygen/Helium analyzer but with an exhausted or not properly calibrated sensor. Wrong analysis of the residual gas is a critical error, and the higher the residual pressure, the more significant the final error is generated.
• Lack of attention in the process of blending
• Inaccuracy of measurements (analysis and pressure)
• Non-constant temperature
• Compressibility of gases
• Molecular interaction between different gases
Classes of errors
Inaccuracy of measurements (analysis and pressure)
Bearing in mind that the methods we use cannot by definition have the same accuracy as the industrial ones, it is crucial to use the best measuring instruments and use them correctly.
The calculation requires the final pressure to be taken at the same temperature as initial pressure.
Example: the temperature is not stable for productivity needs: it is clear that, from a practical point of view, it would be better to go from start to final pressure in a continuous progression: the pros are clear, but what is not clear are the cons.
Here the problem is not setting the exact temperature of the gas: if you follow a procedure correctly with an isotherm at 25 °C while the gas is actually at 15 °C, the final difference is irrelevant.
Instead, what must absolutely be kept in mind is that, as the temperature strongly affects the pressure, if you want maximum accuracy (unless you are using the Continuous Flow method mix with the same residual/target mixture) you must keep a constant temperature when reading the pressure during the various steps of the filling process.
Since it is impossible to keep the temperature stable, because the heat of compression increases the temperature of the mixed-gas cylinders giving a false indication of the pressure in the cylinder, the best solution is to allow the mix to cool, then test the dropped pressure and re-adjust for a small, “no-temperature effect” pressure adjustment.
When productivity is an issue, the usual water cooling method, and possibly reducing the flow rate by filling more cylinders with the same mixture at the same time, are the only possibilities to mitigate this effect.
In this case you can adopt empirical adjustments to compensate this issue, well aware that this is not the best solution.
For more details see Blendig - the key role of temperature
To observe the pressure/temperature relationship see "Pressure and Temperature" in the Tools menu; in order to get an idea of what the impact on the final composition could be, you can use "What if - Ideal Gas and Real Gas" in Tools.
The last two sources of errors are related to the algorithm model, discussed in section Blending - "Real Gas", "Pure gases compressibility" and "What is different in mixtures":
A high-precision digital pressure gauge and a good Oxygen/Helium analyzer that are correctly and always calibrated is crucial.
Assuming that the correct information is found, one thing is to find it, and another is to integrate it into a software and provide the proper procedure to follow.
This is one of the strengths of Tech Gas Blender. The resulting accuracy of the calculation model makes the final deviations virtually undetectable with the measuring instruments available to technical divers.