Thierry, cool that you are measuring BC values!! Keep it up, and share your results!
When calculating BC there are a few areas where errors can creep in: 
(1) What ballistic calculator do you use to measure the BC? Many calculators, e.g., the one at the PyramydAir.com website, do not allow to enter atmospheric data (air pressure, temperature, and humidity), which will cause miscalculations.
However, the ChairGun* ballistic calculator (Windows, and Apple, not the smartphone app) gives you that option.
*
https://uk.hawkeoptics.com/chairgun.html (2) What drag model does the ballistic calculator use? Many calculators use only a constant drag model* or a drag model that is appropriate to bullets typical for fire arms (the G1 drag model)**.
However, the ChairGun ballistic calculator gives you the option of using the GA drag model, which is more appropriate for airgun pellet.
*
https://www.bfxyz.nl/ **
http://www.jbmballistics.com/ballistics/software/software.shtml http://www.shooterscalculator.com/ballistic-trajectory-chart.php?t=f13e8e2b http://www.cleverinsite.com/ballistics/calcs.asp (3) Amount of velocity data points you need to collect It’s normal that at the muzzle, most pellets will not have exactly at the same velocity. And if its velocity coincides with the velocity of another pellet, both pellets will most likely not have the same velocity once they gets to your next measuring point downrange. These coincidental differences in velocity can make large enough differences in your BC calculations that you can get incorrect BC values – by coincidence.
To overcome these differences and the bias this introduces into your BC calculations, you just need to shoot and measure enough times that these coincidental differences will be averaged out. The larger the differences, the more shots (velocity data point) you should take.
For example, if your spread between muzzle velocities is 20fps, taking the average of only 5 measurements at the muzzle is probably not enough. You’d need more like 10, 15, or 20 data points to establish a reliable average muzzle velocity. And if you have a larger velocity spread of say 40fps at your 55m measuring point, you will need to collect many more velocity data points. Otherwise, your velocity average could be – by coincidence – far off from what it really is, because your data points by coincidence are not representative of the true BC.
The larger the difference in fps between the shots at one measuring point, the more measurements you’ll have to make in order to get a more reliable BC value.
(4) You could reduce the amount of velocity data points needed if you had two chronographs, and calculate the BC for each shot individually instead of using an average. If your list of say 15 BC values is very consistent (small spread between the values), you probably have a reliable average BC value. However, if the BC values are spread far apart, e.g., from 0.040 to 0.065, you’ll need more BC values to get a more reliable average.
(5) There are other errors that can creep in when using a chronograph that can mess up our readings. For instance if we have the chronograph set up at an angle, or the sensors are tilted (Chrony model), or if we measure indoors and use only one artificial light source. Others have already explained these possible errors quite well, see the following articles:
http://precisionrifleblog.com/2012/07/20/chronograph-accuracy-tips-15-practical-tips-to-increase-accuracy-reliability/ http://www.frfrogspad.com/chrono.htm https://www.gatewaytoairguns.org/GTA/index.php?topic=158296 Hope this makes sense. Happy shooting and measuring. It’s an exciting hobby!
Matthias For a comprehensive list of published BC values (.22cal) from a wide variety of sources, cf.:
https://www.airgunnation.com/topic/bc-table-22cal-comprehensive-internet-wide-collection-of-ballistic-coeff-data/ 
Check out this thread by Bob Sterne — very interesting and helpful!