Qualitative vs. Quantitative Reporting and the Standard Being Applied
A client's specification cites 'infrared inspection of electrical distribution equipment per the applicable ASTM standard' and requires 'temperature rise limits per the standard.' You perform a survey of 480 V motor control centers. Loads vary from 20% to 90% of rated. You find three connections: (1) 12 K rise over ambient at 90% load, (2) 22 K rise at 40% load, (3) 8 K rise at 20% load. Ambient is 30 °C. The client's engineer asks you to classify each per the cited standard and to state the pass/fail temperature limit.
1. What is the problem with the client's citation, and how do you handle the classification?
Read the worked answer
The citation is ambiguous and not self-executing. 'The applicable ASTM standard' for IR of electrical equipment is a qualitative/quantitative practice standard (e.g., ASTM E1934 for electrical and mechanical equipment), which describes how to perform and report an IR survey - it does not itself prescribe pass/fail temperature-rise limits. Temperature-rise limits for electrical connections come from other documents (e.g., NFPA 70B maintenance guidance, NETA ATS/MTS tables, or the client's own engineering spec). Citing 'the applicable ASTM standard' for limits is a category error.
Handling:
- Ask the client to name the standard and edition that governs acceptance (ASTM E1934 for the survey method; NFPA 70B or NETA for severity; or their internal spec).
- If they cannot, state in the report which standard you applied and why, and clearly separate the method standard from the acceptance standard.
- Do not invent a pass/fail limit; if none is contractually specified, report relative severity (delta-T versus similar connections under similar load) and flag the missing acceptance criterion as a limitation.
2. How do load and the qualitative/quantitative distinction affect how you rank the three findings?
Read the worked answer
Temperature rise scales roughly with load squared (I²R). A 22 K rise at 40% load implies a much larger rise at full load: scaling by (100/40)² ≈ 6.25 gives an extrapolated ~137 K - far more severe than the 12 K at 90% load (extrapolated ~15 K). The 8 K at 20% load extrapolates to ~200 K by the same crude scaling, which is physically implausible and shows the limits of the scaling - but it still signals a connection that should be re-examined at higher load.
Ranking and reporting:
- Qualitative: compare each connection to sibling connections in the same equipment under the same load - a connection 20 K hotter than its neighbors is suspect regardless of absolute temperature. This is the most defensible field call.
- Quantitative: report the measured rise with the load at measurement, and state that the rise is not directly comparable across different loads. Recommend re-survey at a common load (or at rated load) before applying any absolute limit.
- Provisional ranking: connection (2) is the highest concern (large rise at low load, strong extrapolation), then (1) (moderate rise at high load), then (3) (low rise, low load - verify at load). Flag all three for re-check at matched load and note that no acceptance limit can be applied until the governing standard is named.