1.1 The weight is not a guess
Ask a crew what a load weighs and you will often get a number delivered with total confidence and no source behind it. That is the first failure in most rigging incidents, and it is the one the rest of the lift is built on: the sling selection, the hitch, the angle, the crane's radius. If the weight is wrong, everything downstream is wrong by the same factor — and the standard does not accept a confident guess.
1 What the standard actually demands
There are two rules working together. The first is short: 1926.251(a)(2)(ii) forbids loading rigging equipment in excess of its recommended safe working load as marked by the manufacturer, and 1910.184(c)(4) says slings shall not be loaded in excess of their rated capacities. You cannot obey either one without knowing the weight.
The second is where the enforcement teeth are. 1926.1417(o)(3), in the cranes and derricks standard, tells the operator exactly how the weight may be established:
- 1From a source recognized by the industry — the load's manufacturer, the shipping papers, a nameplate — or by an industry-recognized calculation method, and the standard gives its own example: "calculating a steel beam from measured dimensions and a known per foot weight." Or by other equally reliable means. And: "when requested by the operator, this information must be provided to the operator prior to the lift."
- 2Or by test-hoisting it. The operator begins hoisting to see, using a load weighing device, load moment indicator, rated capacity indicator or rated capacity limiter, whether the load exceeds 75 percent of the maximum rated capacity at the longest radius that will be used. If it does, the lift stops until the weight is verified by method one.
Read what that second route implies. Even the version of the rule that lets you find out by lifting builds in a hard stop at three-quarters of capacity — because the whole point is to discover a heavy load while there is still margin, not while it is over the crew.
2 Four sources of a weight, ranked
In practice you will get the number from one of four places. They are not equal, and it is worth being honest about which one you are using.
| Source | How good | What can still go wrong |
|---|---|---|
| Scale or load cell | Best | Weighs what is on the hook right now — which is what you want, provided nothing gets added afterwards. |
| Nameplate or shipping documents | Very good | Often the dry or shipping weight. Fluids, skids, crating, attached piping and the contents of a tank are extra. |
| Calculation from measured dimensions | Good, and explicitly accepted | Only as good as the dimensions and the material. Named by 1926.1417(o)(3)(i) as an acceptable method. |
| Somebody's memory of a similar load | Not a source | Everything. This is the one that shows up in incident reports. |
3 Everything that is not the load
The number the crane sees is not the number on the nameplate. Between the hook and the object there is a stack of things that also have weight, and every one of them comes off the crane's capacity for the day.
- The rigging itself. Slings, shackles, spreader bars, below-the-hook devices, tag lines. A pair of large wire rope slings and a spreader beam is not a rounding error.
- The block and the ball. On a crane, the headache ball or the hook block is deducted from capacity before you start.
- Contents. A tank, a hopper, a machine with a full sump, a bin nobody emptied.
- Attached everything. Skids, dunnage, crating, bolted-on piping and conduit, ice, mud, and the water that has been sitting in the bottom since the last rain.
- Bond to the ground. Not weight, but it acts like it: a load frozen down, suction-bound in mud, or still bolted at one corner will read as far heavier than it is — and then let go all at once.
That last one deserves its own sentence, because it is not solved by better arithmetic. A load that will not break free is a load you stop and investigate, not a load you pull harder on.
4 How wrong is "a bit wrong"?
Weight errors do not stay the size they started at, because the sling angle multiplies them. Take a 4,000 lb load carried by two legs at a 60-degree sling angle. Each leg carries about 2,309 lb. Now suppose the real weight was 5,000 lb — a 25 percent underestimate, which is entirely ordinary when someone forgets the contents of a tank.
| Assumed | Actual | Tension per leg at 60° | Tension per leg at 30° |
|---|---|---|---|
| 4,000 lb | 4,000 lb | 2,309 lb | 4,000 lb |
| 4,000 lb | 5,000 lb | 2,887 lb | 5,000 lb |
At 60 degrees the error costs you 578 lb of unplanned tension in each leg. Flatten the same rigging to 30 degrees and the same 25 percent error costs 1,000 lb per leg — and the leg is now carrying the entire weight of the load on its own. The point is not the arithmetic, which Part 3 covers properly. The point is that a weight error and an angle problem compound, and crews usually discover both at the same time.
5 Writing it down
A weight that lives only in someone's head does not survive a shift change. It also cannot be handed to the operator, which 1926.1417(o)(3)(i) requires when the operator asks. So the weight goes on paper, with three things next to it:
- 1The number, and its unit. Pounds or short tons, written out. "12 tons" from a European supplier may be twelve metric tons — about 26,455 lb rather than 24,000 lb.
- 2Where it came from. Nameplate, packing list, calculation, load cell. One line.
- 3What is included. Load only, or load plus rigging, plus contents, plus skid. State it, because the next person will assume the opposite of whatever you meant.
On a routine repetitive lift this is a line on the pre-task plan. On anything unusual it belongs in the written lift plan covered in Part 5.
6 Repetitive lifts: build the list once
Most sites lift the same twenty things over and over — the same transformer, the same bundle of rebar, the same pump skid, the same precast panel. Weighing each of them once and posting the list turns the hardest step of every future lift into a lookup, and it removes the incentive to guess when the crew is in a hurry.
A useful standard load list has one row per item and five columns: what it is, the weight, the source of that weight, whether the figure includes rigging, and the date it was established. Anything whose configuration changes — a tank that is sometimes full, a basket that is sometimes loaded — gets two rows, not an average.
The list is a jobsite document, not a permanent one. When a supplier changes, when a unit is modified, or when the same nameplate turns up on a heavier revision, the row is re-established rather than inherited. A stale load list is more dangerous than none, because it carries the authority of paper.
- You cannot obey 1926.251(a)(2)(ii) or 1910.184(c)(4) without knowing the weight — the two rules depend on it.
- 1926.1417(o)(3) accepts an industry-recognized source or calculation, or a test hoist with a weighing device that stops at 75 percent of rated capacity at the longest radius.
- The operator can require the weight to be provided before the lift, so it has to exist in writing.
- Add the rigging, the block, the contents and everything bolted or frozen to the load; a bound load reads heavy and then releases all at once.
- Weight errors are multiplied by the sling angle: 25 percent low at 30 degrees is 1,000 extra pounds in each leg on a 4,000 lb load.
- Write the number, its source, and what it includes. A weight without a source is not a weight.
Free educational content — not OSHA-authorized training, no certificate or card issued. Follow your employer's program and the standards cited.