often we see the assessment for sampling bias;
but is there any acceptance limits for sampling error.
The short answer is no: sampling bias is the result of a poor sampling plan.
To avoid sampling bias the sample should be taken randomly. Sampling bias is, for example inspecting only the green widgets or examining services performed by a particular group of personnel. This is not representative of the population.
However, sampling might not be biased if the experiment might be repeated exactly for other widgets or people and compared to see if a particular aspect of performance varies between these two groups if all else is exactly equal (which is seldom the case).