when you say you figure by volume though do you mean the total number of ml of total solution to that of total ml of water.
I mean you take X amount of water. Lets say you use 1 gallon. That's ~3780ml. Then to it you add 22ml of nutrients (total). The final solution is 3802ml.
Just so I've covered my own ass here... The following example is an approximation of nutrient levels available. They represent what can be at best considered "ballpark" and should be considered for comparison rather than for accuracy. Keep track of what you are adding, stay consistent in your record keeping and your formula, and you can then use the data to improve future grows of the same strain by knowing where it likes the relational (inaccurate consistently) nutrient level to be.
At
20ml of CNS 17 Bloom 2-2-3 Ca2.5 Mg0.5
and 2ml of Hydroplex 0-10-6 Ca0.01 Mg0.5 added to exactly 3780ml for a total solution of 3802... The formula's look like this:
N=(2x10,000)/(3802/20)=20,000/190.1=105.2
P2O4=((2x10,000)/(3802/20))+((10x10,000)/(3802/2))=(20,000/190.1)+(100,000/1901)=105.2+52.6=157.8
P=157.8x0.44=69.4
K2O=((3x10,000)/(3802/20))+((6x10,000)/(3802/2))=(30,000/190.1)+(60,000/1901)=157.8+31.6=189.4
K=189.4x0.83=157.2
Ca=((2.5x10,000)/(3802/20))+((0.01x10,000)/(3802/2))=(25,000/190.1)+(100/1901)= 131.5+0.05=131.6
Mg=((0.5x10,000)/(3802/20))+((0.5x10,000)/(3802/2))=(5,000/190.1)+(5,000/1901)=26.3+2.6=28.9
Total guaranteed nutrient ppm: 612.9
105.2-157.8-189.4 Ca131.6 Mg28.9
Add in the tap water starting point and unaccounted for elements like Sulfur, Iron, Manganese, etc, and compensate for additional NPK unlisted on the bottle... That could be anywhere from 160ppm more to 360ppm more. So, actual measuring should be something like 770ppm to 970ppm. Translate that back into EC and it could be anywhere from 1.2 to 1.5... which is a pretty huge deviance.
So truth be told... you can't really figure out anything accurately just by crunching the numbers because you can't be sure what you're actually adding to the water, or what is in the water to begin with (short of RO filtered I suppose). Just use your brain, try to come to your own conclusion using the data at hand, and don't underestimate the necessity of actual measurements.