Handbook 44 scale test procedures
Updated · 4 min read
On this page
Handbook 44 Section 2.20 sets out which tests a scale gets and how the results are judged (the "N" notes for procedures, the "T" sections for tolerances). This guide walks through the ones a field technician runs on most commercial scales, using the 2026 edition. Vehicle, livestock, railway track, crane, monorail and in-motion systems have their own test patterns in the handbook; they're mentioned, not covered.
The tolerance each result is compared to comes from Table 6. See Handbook 44 scale tolerances explained.
Before you start
- Record the scale's identification: owner, location, make, model, serial number, accuracy class, capacity, e and, where marked, the NTEP Certificate of Conformance number.
- Check that the scale is level, clean and suitable for its use, and that the zero is set.
- Make sure your test weights are within their calibration dates; see test weights for scale service.
- If you adjust anything, record the readings as found first, then as left.
Increasing-load test (N.1.1)
Run on every scale. Apply test loads in steps up to capacity (or as much known test load as you have), approximately centred on the load-receiving element. If you have less test load than the capacity, concentrate it over the main load supports to get the most out of it.
Each indication must be within tolerance for that test load.
Decreasing-load test (N.1.2)
Run on automatic-indicating scales, with the load centred, as you take the weights back off:
- Scales marked I, II, III or IIII with n of 1,000 or more: test at the largest test load of each tolerance step. On a Class III scale, that's 4,000 e, 2,000 e and 500 e.
- Scales with n under 1,000, and all other scales (except portable wheel-load weighers): one test at one-half of the largest load used in the increasing-load test.
Shift test (N.1.3)
The shift test shows whether the scale reads the same wherever the load sits. For most platform and bench scales (N.1.3.7):
- Capacity 500 kg (1,000 lb) or less: one-third capacity (at least 30 %, at most 35 %), centred in each quadrant of the platform in turn.
- Capacity over 500 kg (1,000 lb): either the one-third-capacity quadrant test, or one-quarter capacity centred over each corner in turn.
A single weight goes in the centre of each position. With several weights, use the same pattern at every position, and don't stack them into a smaller footprint than a single weight would occupy.
Judging it (T.N.4.4): each result must be within tolerance, and the range (highest reading minus lowest) must not exceed the absolute value of the maintenance tolerance for that load.
Vehicle, axle-load and livestock scales use the section and axle patterns in N.1.3.3; equal-arm scales, dairy-product test scales, railway track and monorail scales have their own subsections.
Repeatability (T.N.5)
Weigh the same load several times under reasonably steady conditions. The results must agree within the absolute value of the maintenance tolerance for that load, and each must be within tolerance.
Discrimination test (N.1.5)
Run on automatic-indicating scales near zero and near the maximum test load (except Class I and II digital scales where e = d and is under 5 mg). On a digital scale, start just below the lower edge of the zone of uncertainty for increasing loads, or just above the upper edge for decreasing loads, and check that a small added load changes the indication as the handbook specifies.
Sensitivity test (N.1.4)
Only on non-automatic-indicating (weighbeam) scales: at zero and at the maximum test load, add or remove the load in the sensitivity requirement and check the beam moves as required.
Zero-load balance (N.1.9)
After you remove each test load, the zero-load balance should not change by more than the minimum tolerance applicable.
Other procedures
- Substitution test (N.1.11): replace known test weights with material, re-establish the same indication, and keep testing above the weight you have on hand. Useful on large scales.
- Strain-load test (N.1.12): apply known test weights on top of an unknown amount of material already on the scale.
- Ratio test (N.1.7): for scales with counterpoise weights and equal-arm scales.
- RFI test (N.1.6): at an installation where radio-frequency interference has been verified and is usual and customary.
What goes in the record
For each test: the test load, the indication, the error and the tolerance it was judged against (maintenance or acceptance), as found and as left. Add the weights used, with their calibration certificate numbers and dates, and the technician's registration number. If you place the scale in service, the state usually wants a placed-in-service report too.
The scale tolerance calculator gives the decreasing-load and shift test loads for a scale and checks each reading against Table 6.
Certorum for scale service agencies
Handbook 44 tolerances checked as you enter each test load, as-found and as-left results on one report, placed-in-service reports, test weights on record and the invoice from the same visit. Free for up to 40 scales.
More guides
- NIST Handbook 44 scale tolerances explained
- Placed-in-service reports for scales
- Test weights for scale service
- ASHRAE 110 fume hood testing: procedure, ratings and acceptance criteria
- Biosafety cabinet classes and types explained
- What a biosafety cabinet certification report should include
- How often do biosafety cabinets, fume hoods and cleanrooms need certification?
- Cleanroom air changes per hour: how to calculate and verify them
- ISO 14644-1:2015 cleanroom classification: class limits, sample locations and sample volume
- Biosafety cabinet certification: NSF/ANSI 49 field tests and acceptance criteria
- USP <797> cleanroom certification: what's tested every six months