Induced-charge errors in charge-transfer measurement

Induced-charge errors cause a degree of under-recording when charge-transfer measurements made with high-speed, unshielded test-probes are used for assessing electrostatic hazards. Published information does not quantify the errors well enough to allow appropriate test tolerances to be set. This work tries to rectify that situation by calculating the collection-efficiency (ratio of measured to actual charge-transfer) for sparks and brush discharges to earthed spherical probes from representative flat or convex test bodies. Our results include data on the worst-case induced-charge errors for charge-transfers of 60 nC, 30 nC and 10 nC, which have been suggested as hazard thresholds for Gas Group IIa, IIb and IIc materials respectively. For 60 nC charge-transfers from non-concave surfaces, our data indicate that the possible under-recording of charge-transfers measured with fast, unshielded probes can be dealt with by applying a correction factor of ×2 to ×2.5. For 30 nC and 10 nC charge-transfers, or for discharges from concave surfaces, the necessary correction is both larger and more sensitive to gap geometry. Consequently, we do not recommend the use of fast, unshielded probes for assessing the electrostatic hazards arising from charges on concave surfaces or with materials that are more sensitive to ignition than those in Gas Group IIa.