__________________
"Did you get my e-mail?" - "The biggest problem in communication is the illusion that it has taken place" - George Bernard Shaw, 1856
How would your client know what is too high? Where did he get his information?
I would read the instructions very carefully. After all, when you attempt to measure this, you are actually measuring the contact resistance of three contacts in series. So, how do you eliminate the two points of resistance caused by your meter leads?
Lastly, the contact resistance of any cold connection will not be the same as what you get when you apply power to the contact. There are many factors that may or may not apply depending on the construction. The branch of EE that deals with this is called rheology.
Low resistance measurements are much more difficult to read correctly than you might think. It is about 60% technique and 40% instrument and sometimes it is difficult to figure out which one is right or wrong. Good luck!
__________________
A great troubleshooting tip...."When you eliminate the impossible, whatever remains, however improbable, must be the truth." Sir Arthur Conan Doyle
You also have not given enough information about your circuit breaker: What is its voltage & current rating, manufacturer, etc. Depending on the size and type, the magnitude of the contact resistance may be acceptable, depending on the test specifications. If this breaker is rated 3000 or 4000 amps, these readings are high, but for a breaker of 60 to 600 amps, the values could be acceptable.
The bigger concern is the variation in readings between the phases: L2 is 3x the L1 value, and L3 is 2x. Was this breaker operated and maintained before the readings were taken? Often the mechanism has been sitting for ages without operating, so lubricants become thickened and the contacts oxidized. The breaker should be manually exercised a number of times, and the readings taken again to see if there is a change. I wouldn't be surprised if the values come down.
__________________
To get the right answers, first you need to ask the right questions.
<...Any guideline to follow what is the requirement....>
A <...test...> has to have a pass/fail boundary established BEFORE carrying out the measurement. Otherwise, all one is doing is obtaining a measurement, and not carrying out a true <...test...>.
So, go back to the test protocol document and read what it says. If one doesn't exist then get someone to write one, and get it approved for use before carrying out the <...test...>. Otherwise all the measurements are either academic, or a complete waste of time.
Sheesh.
__________________
"Did you get my e-mail?" - "The biggest problem in communication is the illusion that it has taken place" - George Bernard Shaw, 1856