Why You Tie Off to the Basket and Never to the Structure

Why this matters

If someone is currently working from an aerial lift platform with a lanyard clipped to a beam, a pipe rack or a pole, the action is to stop all boom function immediately, tell the occupant not to move, and get them re-connected to the platform's own anchor point before anything else happens - including before lowering the boom, because lowering is boom movement and boom movement is the mechanism.

The tie-to-the-structure decision is almost always made for a good reason. The beam is right there, it is enormous, it is not going anywhere, and every instinct a tradesperson has says a fixed heavy structure is a better anchor than a machine on tyres. That instinct is correct about strength and wrong about the event.

The claim of this card is that a platform anchorage has to move with the platform, because the moment it does not, the machine becomes the thing applying the load. A boom lift's hydraulics and drive develop forces vastly beyond a person's body weight, and none of the arithmetic that governs personal fall arrest describes what happens when a machine pulls on a lanyard.

The rule and where it lives

29 CFR 1926.453(b)(2)(iii) states plainly that belting off to an adjacent pole, structure or equipment while working from an aerial lift is not permitted, and (b)(2)(v) of the same section requires the lanyard to be attached to the boom or basket. The general industry counterpart for vehicle-mounted elevating and rotating work platforms is 29 CFR 1910.67. Name the Part that governs your work; a service call on an operating building is usually general industry and the same tech on an active construction project is usually construction.

That is one of the rare cases where the standard's text is the whole rule and the interesting question is why. The rest of this card is the why, because a rule a crew understands survives the beam being right there and a rule they have only memorised does not.

What fall arrest arithmetic describes, and what this event is not

Personal fall arrest is designed around a body falling under gravity. Every number in 29 CFR 1926.502(d) is derived from that: the 6-foot limit on free fall, the 3.5-foot limit on deceleration distance, and the 1,800-pound cap on maximum arresting force with a body harness. Those figures bound what a falling human's momentum is allowed to do to a body through a functioning energy absorber.

A boom pulling a tethered person is a different event with a different energy source, and it is worth being precise about which terms survive.

   structure beam
  ==================
          \
           \  a lanyard to the beam stays
            \ where the beam is
             \
        +-----+-----+
        |  basket   |  a lanyard to the basket
        +-----+-----+  anchor travels with it
              |
             boom
             /
            /
        chassis

Worked example: three feet of slack and eight feet of boom

A tech is in a boom platform working alongside structural steel. He clips a 6-foot shock-absorbing lanyard to a beam about 3 feet away. The ground operator, or a bump on a control, moves the boom.

Line 1, the slack. 3 feet of separation at the beam means about 3 feet of movement before the lanyard is taut. Nothing is felt during this.

Line 2, the absorber. Past taut, the energy absorber begins to pay out. 29 CFR 1926.502(d)(16) limits maximum deceleration distance to 3.5 feet, so 3.5 feet is the absorber's designed travel, taken as a bound rather than a typical value.

Line 3, total movement the system can accommodate. 3 plus 3.5 is 6.5 feet.

Line 4, the movement available. A boom function at height traverses far more than that. Take 8 feet, which a modest slew at a long radius produces without the operator noticing anything unusual.

Line 5, the excess. 8 minus 6.5 is 1.5 feet of boom travel applied through a connection with no remaining extension in it. Once the absorber has bottomed out, the load path is machine, to lanyard, to harness, to body, and it is essentially inextensible.

Line 6, why the 1,800-pound figure does not help here. That cap is a limit on what an arresting system is permitted to impose on a falling body. It is not a limit on what a hydraulic boom can apply through a bottomed-out absorber, and no provision of 1926.502(d) governs the machine's output. Quoting it in this situation is applying a number outside the condition it was derived under.

Line 7, the two ways it resolves. Either the tie holds and the occupant is pulled against the platform rail. A rail at roughly waist to chest height sits just above the centre of gravity of an adult standing on the platform floor, which is exactly why a rail works as a barrier and exactly why a sustained lateral pull rotates the body over it. Or the tie holds, the occupant holds, and something structural gives - the platform anchor point, the rail, or the connector.

Line 8, the reverse case, which is the one people actually picture. Suppose it is a genuine fall rather than boom movement, and the occupant goes over the rail while clipped to the beam. Now he is suspended outside a platform that may have moved, and on most boom lifts the platform controls are in the platform he is no longer in. 29 CFR 1926.502(d) requires the employer to provide for prompt rescue or to ensure employees can rescue themselves, and a tie to fixed structure has just removed the machine that was the rescue plan. The clearance arithmetic for that fall is owned by this library's card on the transition fall and is not re-derived here; the point specific to this card is that the retrieval got worse at the same instant.

Three separate failures, not one

It is worth separating them, because a crew that hears only the first will accept a tie to structure whenever the boom is parked.

The machine applies the load. Lines 1 through 7. It requires only that the boom move, and boom movement includes drift, a bumped control, a second person operating from the ground, and lowering.

The retrieval is removed. Line 8. This one is true even if the boom never moves at all, which is why "I'll be careful with the controls" does not answer it.

The anchorage is unevaluated. A beam is strong. Strong is not the test. 29 CFR 1926.502(d)(15) requires an anchorage used for personal fall arrest to be independent of any anchorage used to support or suspend platforms, and capable of supporting at least 5,000 pounds per employee attached, or to be designed and used as part of a complete personal fall arrest system with a safety factor of at least two under the supervision of a qualified person. The general industry counterpart is 29 CFR 1910.140(c)(13). Whether the flange you clipped to satisfies that, in that direction of pull, with that connector, is a question for a qualified person and not for a glance.

The word anchorage sets a trap, and it is worth naming

This library carries several articles about anchors: how an anchor transfers load into concrete, which anchor type suits which base material, why anchors fail. Those are concrete and masonry fasteners for hanging and mounting equipment, and they are a different subject with a different rating basis.

A fastener anchor is rated by a published allowable load for a stated embedment, edge distance, spacing, base material and installation condition, and it is designed for a static, largely predictable service load. A fall arrest anchorage is rated under 1926.502(d)(15) against a 5,000-pound-per-person criterion or an engineered safety factor of at least two, for a dynamic event whose direction of pull is not known in advance.

The two are not convertible. A threaded rod anchor holding a piece of equipment, sized correctly and installed correctly for that equipment, is not an anchorage, and a lifeline clipped to it is attached to something whose rating never contemplated this load. If a reader takes one sentence from this card into the rest of the library, take that one.

When the structure is the only thing in reach

Two situations come up genuinely and neither is answered by tying to a beam.

The work is beyond the platform's reach. That is the machine being wrong for the job, not the anchorage being wrong. The answer is a different machine, a different setup position, or a different access method entirely. Reaching out of a platform and tethering to what you are reaching toward is the specific arrangement this card describes.

A transfer out of the platform onto the structure. Whether this is permitted at all is governed by the machine manufacturer's instructions and your site's written programme, and where it is permitted it requires continuous attachment through a second connector, so there is no instant with nothing connected, and the structural point used must be an anchorage evaluated under 1926.502(d)(15) rather than the nearest steel. Do not improvise a transfer. If your shop does transfers, it has a written procedure for them or it does not do them.

One more thing that is not an option: adding your own attachment point to the platform. Drilling, welding or clamping a new anchor to a basket modifies the machine, and the platform anchor point you are trying to supplement is rated by that machine's manufacturer for that platform. That rating is not transferable to a point you installed.

How to verify you got this right

Before the platform leaves the ground:

• Harness on and adjusted, lanyard connected to the manufacturer's designated anchor point on the boom or basket, checked visually by a second person.
• Confirm the anchor point on the platform is the manufacturer's, unmodified, and not something added by a previous crew.
• Agree out loud who may operate the boom, and confirm nobody at the ground station touches a control without a verbal exchange with the occupant.
• Confirm the reach required is inside the machine's envelope for the load in the platform, so nobody is tempted to close the last few feet with their body.

At height, every time you reposition:

• Before any boom function, look at where your lanyard is clipped. This is the whole discipline. The tie to structure is almost never made at the start of a job; it is made mid-task, by someone who unclipped to reach around an obstruction and clipped to the nearest solid thing without deciding to.

The failure mode is not a person who does not know the rule. It is a person who knows it perfectly, is holding a connector in one hand and a fitting in the other, and clips to the beam for thirty seconds. Then the boom moves.

References

• 29 CFR 1926.453, Aerial lifts (construction), including (b)(2)(iii) prohibiting belting off to an adjacent pole, structure or equipment, and (b)(2)(v) requiring the lanyard to be attached to the boom or basket.
• 29 CFR 1910.67, vehicle-mounted elevating and rotating work platforms, the general industry counterpart.
• 29 CFR 1926.502(d), personal fall arrest system criteria, including (d)(15) on anchorages and the 5,000-pound-per-employee criterion, (d)(16) on free fall, deceleration distance and maximum arresting force, and the requirement for prompt rescue.
• 29 CFR 1910.140, personal fall protection systems for general industry, including (c)(13) on anchorages.
• Machine manufacturer's operating manual, which owns the designated platform anchor points and any permitted transfer procedure.
• See related: universal-the-fall-that-happened-on-the-way-down-not-the-way-up for the fall clearance stack, and universal-what-an-aerial-lift-requires-that-a-scissor-lift-does-not for which standard governs which machine.