Two standardless quantitative methods for evaluating EDS X-ray spectra were investigated in regards their basic metrics. Both methods have similar total errors, but the error contributions are from different sources. In the P/B-based method, error is more related to counting statistics and therefore can benefit from high count rates achievable with modern silicon drift detectors. To reduce systematic uncertainties in the net-count-based standardless approach, measured values need to be supported by data in a previously measured database. Using the P/B-based method, it is now possible to achieve standardless EDS quantification within ±10% relative deviation from true composition for 95% of results.
