The Rolling Offset Cut Short Because the Roll Was Ignored

Why this matters

A rolling offset is an offset that moves in two directions at once: up and sideways, or down and over. It is one of the ordinary moves in drain and vent work, and the arithmetic is no harder than a simple offset once you have the right input. The trap is that one of its two directions is often the one you can see from where you are standing, and the other is not.

This case is a fitter who cut the same short piece twice. Both times he correctly worked out that the piece was short because the travel was wrong. Both times he went looking for the error in his arithmetic and his tape, and both were fine. The number that was wrong was one he never wrote down, and being right about the symptom did not help until he found it.

The job

A remodel was moving a lavatory. The new 1-1/2 in drain arm came through the wall and down into a basement ceiling, where it had to tie into an existing horizontal drain. The existing line ran parallel to the joists, and the new stub ended above it, offset from it. The fitter planned a rolling offset with two 45 degree elbows: out of the stub, across on the slant, and into a fitting on the existing line, which would carry on at its own slope.

Standing under the work, he measured from the centerline of the stub down to the centerline of the existing line's fitting outlet: 16 in. He wrote "offset 16" on a joist.

The first cut

For 45 degree fittings the travel, the center-to-center length of the slanted piece, is the offset times 1.414. See related: Pipe Offset Constants for Every Fitting Angle.

• Travel = 16 in x 1.414 = 22.63 in.
• Each 45 elbow's take-off, the distance from the fitting's center to the bottom of its socket, came off the fitting maker's dimension sheet. Say it read 1-1/8 in for this fitting (an illustrative figure; read your own fitting's sheet).
• Cut length = 22.63 in - 2 x 1.125 in = 22.63 - 2.25 = 20.38 in, cut at 20-3/8.

He primed and cemented a 45 on each end, offered the assembly up, and the far fitting stopped well short of the existing line's outlet, swung off to one side of it. Cemented joints do not come apart, so the piece and both elbows went in the scrap bucket.

The second cut, and why it came out the same

His read on the failure was correct: the slanted piece was too short, so the travel was wrong. He then did what most people do and checked the things he had calculated.

The constant. 1.414 is the right multiplier for 45 degree fittings. Checked against the fitting chart, fine.

The take-off. He wondered whether he had read the dimension sheet for the wrong size and looked it up again. Same figure. Fine.

The tape. He re-measured the drop from stub to outlet more carefully, hooking the tape on the stub's fitting and reading straight down to the height of the outlet's center mark, and got 16-1/16 in. That changed the travel to 16.0625 x 1.414 = 22.71 in and the cut to 22.71 - 2.25 = 20.46 in, about 20-7/16.

The second assembly was 1/16 in longer than the first and arrived in the same wrong place. Two pieces and four elbows scrapped, and the ceiling still open. Every number he had checked was right. The trouble was the number he had not written down at all.

What the joist note left out

The supervisor came down, looked up at the stub, then walked around to the other side of a joist and looked again. From there it was obvious: the new stub was not directly above the existing line. It sat over to one side, between two joists, and the existing line's outlet faced into the next bay.

He hung a plumb bob from the stub's centerline and measured across, level, to the outlet's centerline: 12 in. The run did not just drop 16 in. It dropped 16 in and rolled 12 in sideways.

 END VIEW, looking along the existing line

                     STUB (leave)
          +---------*
          |       / |
          |     /   |  RISE
          |   /     |  16 in
          | /       |
 OUTLET   *---------+
 (arrive)   ROLL 12 in

 diagonal = TRUE OFFSET
   sqrt(16 x 16 + 12 x 12) = sqrt(400) = 20 in
 SETBACK along the run, 45 deg fittings
   = true offset = 20 in
 TRAVEL = 20 in x 1.414 = 28.28 in
 FIGURED FROM RISE ONLY = 16 in x 1.414 = 22.63 in

The vertical leg is called the rise whichever way the pipe travels; here it is a drop. The offset a pair of fittings has to make is not the rise and not the roll. It is the diagonal of the rectangle they form, seen looking along the run, and that diagonal is the square root of the rise squared plus the roll squared: square root of (256 + 144) = square root of 400 = 20 in. With 12, 16 and 20 being 3, 4 and 5 times four, the numbers make a quick check of themselves.

The travel for 45s is then that true offset times 1.414: 20 x 1.414 = 28.28 in. The first piece had been figured from 22.63 in. The first piece's shortfall was 28.28 - 22.63 = 5.66 in, a hair over 5-5/8. The second, figured from 22.71, fell short by 28.28 - 22.71 = 5.57 in. Neither recut changed the input that was missing, so neither could close the gap.

Checking it a second way before cutting again

The supervisor had him check the travel with the whole box before anyone touched the saw. A rolling offset with 45s fits inside a box whose three sides are the rise, the roll and the setback, the distance the offset advances along the run. With 45 degree fittings the setback equals the true offset, 20 in. The travel is the diagonal through the box:

• Square root of (16 x 16 + 12 x 12 + 20 x 20) = square root of (256 + 144 + 400) = square root of 800 = 28.28 in.

Same answer by a different path. Then the cut:

• 28.28 in - 2.25 in = 26.03 in, cut at 26 in.

This time he dry-fit both elbows on the piece without cement, rolled the assembly to line up with the outlet, and marked the fittings and the pipe with a pencil line across each joint so the roll angle would come back when it was cemented. It landed on the outlet. Then he cemented it, matching the marks, with the area ventilated and nothing lit nearby, since solvent cement and primer vapors are flammable and irritate the airways and eyes.

The same box with a different fitting angle

The true-offset step does not belong to 45s. It comes first whatever the fittings are, and only the constants after it change. With the same 20 in true offset:

• 60 degree fittings: travel = 20 x 1.155 = 23.10 in, setback = 20 x 0.577 = 11.54 in. A shorter piece and a tighter box, useful where the run has little room along its length.
• 22-1/2 degree fittings: travel = 20 x 2.613 = 52.26 in, setback = 20 x 2.414 = 48.28 in. A long, gentle piece that needs four feet of run to make the move.

Figured from the 16 in rise alone, the 60 degree piece would have come out 16 x 1.155 = 18.48 in, short by 23.10 - 18.48 = 4.62 in, and the 22-1/2 degree piece 16 x 2.613 = 41.81 in, short by 52.26 - 41.81 = 10.45 in. The shortfall is always a fifth of the true travel here, because 16 is four fifths of 20, so the gentler the fitting the more pipe the missing roll costs.

Why being right twice did not help

The fitter's first diagnosis, "the travel is short", was correct, and so was the follow-up thought, "something in my calculation is off". What sent him in a circle was where he looked. He audited the three things he had calculated, found them sound, and concluded his tape work was sloppy, so he made it more careful. Careful measurement of the wrong quantity produced a second answer just as wrong as the first, only more precise.

The piece did more than fall short. It arrived swung to one side of the outlet, and that was the clue nobody read. A simple offset that is too short arrives short and in line. One that arrives off to the side is telling you the offset has a second direction. If the first failed piece had been held up and the gap measured both ways, along the run and across it, the roll would have shown up then and only one piece would have been scrapped.

What would have changed the outcome

Measuring from a datum in both directions before figuring anything. A rolling offset has two dimensions across the run, and both have to come from fixed references: a plumb line from the starting centerline and a level line to the arriving one. Measuring the drop alone, however carefully, gets one of them and hides the other. See related: How to Figure a Rolling Offset, which owns the full method.

Solving the true offset first. The step that was missing is one line: true offset = square root of (rise squared + roll squared). Once that line is on the joist, the travel follows from the same 1.414 the fitter already trusted.

Dry-fitting before cement. A dry fit costs a minute and nothing else. Solvent-welded joints are permanent, so every cemented mistake scraps the fittings along with the pipe. Two dry fits would have cost two minutes and both sets of elbows would still be in the box.

Reading the failed piece. A wrong piece is information. Measure where it lands against where it should, in both directions, before cutting another.

What changed afterward

The shop added two lines to its offset card, under the constants table every fitter carries:

• Rolling offset: measure rise AND roll from a plumb line and a level line. True offset = square root of (rise squared + roll squared). Figure travel from the true offset, never from the rise.
• Dry-fit and mark the roll before cementing any offset.

It also asked fitters to write all three numbers on the work, rise, roll and true offset, rather than one. A note that reads "offset 16" invites the next person to trust it. A note that reads "rise 16, roll 12, true 20" shows its own working and can be checked from the floor.

References

• The fitting manufacturer's dimension sheet for take-off values
• The solvent cement and primer manufacturer's instructions and safety data sheet for ventilation and cure times
• See related: How to Figure a Rolling Offset
• See related: Pipe Offset Constants for Every Fitting Angle
• See related: Write the Calculation Down: The Field Math Habit