How to Read a Drainage Plan Back to the System That Got Built

Why this matters

A drainage plan is a statement of intent made before anyone put a pipe in the ground, and the ground gets a vote. That is not a criticism of the plan; it is what a plan is. The problem is that the plan is the only document anybody keeps, so ten years later it gets treated as a map of the building, and the tech who trusts it prices a job around a vent riser that was never built and a cleanout that is under a tiled floor. The divergences are not random, though. They cluster in five places, all of them places a drawing cannot constrain once construction starts. Work through those five deliberately and what you hand back is not a marked-up print, it is a register that says, field by field, what was verified, what was contradicted, and what nobody can check without opening something. The last of those three is a real answer and it belongs in the record.

The five places a plan and a building reliably part company

• Vents. Concealed the moment the walls close, so nobody ever sees the divergence. Vent runs get re-routed around framing, converted to wet vents, or replaced in part by an air admittance valve during a fit-out, and no revised drawing follows.
• Slope and invert elevations. The plan states a slope; the existing tie-in elevation, the footing, and whatever was found in the ground decide the one that got built. Bellies live here.
• Cleanout locations. Drawn at the spacing the code wants, built where there was room, then covered by a later finish or a landscaping change.
• Fixture count and type. Fixtures get added, deleted, upsized and re-purposed during construction and continuously after occupancy. The schedule on the plan is a day-one snapshot.
• The exit point and the route outside. Almost never as drawn, because it had to meet what was actually in the ground.

Step 1: Establish what document you are actually holding

Read the title block before the pipes. A permit set is intent at permit time. An issued-for-construction set is intent on the day the crew started. A record or as-built set is a claim about what was built, usually assembled at the end of a job by someone reconciling markups, and it is more reliable than a permit set and still not evidence. Write the set name, the revision and the date into the register as its own field.

Skipped: every later discrepancy gets argued as a mistake rather than understood as the normal gap between two different documents, and you have no way to say which of two prints in the customer's file is the newer one.

Step 2: Transcribe the plan's claims into the register before you go looking

Pull each of these out and write it down as a claim to be tested, not as a fact: fixture schedule and count, drainage fixture unit total, drain and stack sizes, stated slope, vent arrangement and the number of terminals, cleanout stations, and where the drain leaves the building. Six or seven lines.

Doing this first, on paper, is what stops the field walk from turning into confirmation. A tech who has not written down "plan shows two vent terminals" will count the terminals on the roof, see one, and think nothing of it.

Skipped: you find what you were looking at rather than what the plan promised, and the divergence in class one goes undetected for another decade.

Step 3: Inventory what can be established without opening anything

Fixtures, visible pipe sizes and materials, accessible cleanout covers, and vent terminals counted from the roof or from the ground with binoculars. Count the fixtures that are not on anybody's list: mop sinks, coffee-bar sinks, ice machine and condensate drains, a floor drain in a back room, an equipment drain in a mechanical space.

Where the count has to be done from a roof, fall protection is rigged first, under 29 CFR 1910.28 in general industry or 29 CFR 1926 Subpart M on construction work, and those two Parts do not share a trigger height. Where it can be done from the ground with optics, do that instead; a terminal count is not worth a January roof.

Skipped: your fixture unit total is the plan's total, which is the one number in this whole exercise that is guaranteed to be out of date.

Step 4: Reconcile the counts against each other before testing anything

Two comparisons, both cheap and both high yield: field fixture count against the plan schedule, and roof terminal count against the plan's vent risers. A mismatch in either names the divergence class and tells you what to test in step 5, which is the expensive step.

Skipped: you spend camera time on the drain and never find out that a vent riser on the drawing does not exist.

Step 5: Test what is left, and ask one question before you put anything in a line

Ask whether a chemical drain cleaner has been used, and put the answer in the register as its own row before you commit. Standing caustic in a line is what the next person to open it inherits, and that person is now you. If the answer is yes or unknown, flush with water first from a distance with a face shield and chemical-resistant gloves on, read the product's safety data sheet if the container is on site, and do not put bare hands, a bare cable or your face over the opening. This is the single most-skipped question in drain work and it is the reason for a large share of the chemical burns in this trade.

Then, in this order because each one narrows what the next has to cover: the non-invasive vent check covered by the sibling HowTo card, to establish whether the vent path is open without opening anything; tracing dye at each suspect fixture to establish which drain actually goes where; and a camera run for the slope and invert questions, which is the only one of the three that can see a belly.

Cleanouts on a line that may be under head get opened with a face shield on, the body clear of the plug's path, and the plug backed off slowly so it vents before it clears the threads. Camera reels and any other corded equipment used in a wet basement run on GFCI protection.

Skipped: the camera run becomes the whole survey, and a camera answers exactly one of the five classes.

Step 6: Record three states, not two

Every field closes as confirmed, contradicted or unverifiable, each with its evidence. The third state is the one that gets dropped, and dropping it is how a register becomes a lie: a cleanout that exists on the plan and cannot be reached is not the same as a cleanout that is missing, and the difference decides whether the next job is a cover replacement or an excavation.

The register, filled in: a two-storey retail with an office fit-out above

Document: permit set, revision 2, dated at original construction. No record set exists.

Working the fixture unit line. Field total 55 against plan total 46 is 9 DFU more, which is 19.6 percent above the design figure. The obvious next sentence is that the drain is overloaded, and it is wrong here: checked against the adopted local table's allowance for a 4 in building drain at 1/8 in per foot, 55 DFU sits well inside it. The finding is a record problem, not a capacity problem. Say that plainly rather than letting a 19.6 percent number carry an implication it does not support. The sibling card on what a drainage fixture unit is actually counting covers why the allowance is what it is.

Working the slope line, and why the average is the wrong statistic. Invert depth below the slab measured at the two accessible cleanouts, 60 ft apart:

• Station 0: 3.10 ft
• Station 60: 3.90 ft
• Fall = 0.80 ft = 9.6 in over 60 ft, which is 0.16 in per foot.

Each invert reading carries about plus or minus 0.05 ft of probe and tape uncertainty, and those two are independent random spreads rather than one shared offset, so the uncertainty on the difference is the square root of two times 0.05, about 0.07 ft, which is plus or minus 0.9 in on the fall. Measured fall is 9.6 plus or minus 0.9 in against a designed fall of 60 x 0.125 = 7.5 in. The line has 2.1 in more fall than the plan called for, and the margin is bigger than the uncertainty, so that is a real difference rather than measurement noise.

And it tells you nothing about whether the line drains. The camera run found standing water over about 9 ft beginning at 34 ft from the cleanout. A line can carry more than its designed fall on average and still hold water in the middle, because average slope is a difference between two endpoints and a belly is a local reversal between them. Two invert readings can never see it. This is the reason the camera step exists and the reason it comes after the cheap steps rather than instead of them.

What the register changed. The belly at 34 ft is the actual defect and the actual quote. The air admittance valve becomes a documented device with a service interval and a code question for the jurisdiction, because the adopted code decides where such a valve is permitted at all and it is never permitted as the sole vent for an entire system. The tiled-over cleanout becomes a scheduled access item. The corrected exit point saves the next crew a trench on the wrong side of the building. And the next permit application starts from 15 fixtures and 55 DFU instead of a schedule that has been wrong since the fit-out.

Failure mode. The version that goes wrong is a tech who runs the camera first because that is the interesting tool, finds the belly, quotes the repair, and never counts anything. The belly gets fixed. Six months later the coffee bar's air admittance valve fails closed, the upstairs office traps get pulled, and the smell complaint is worked from scratch by somebody who has the same plan showing two vent risers.

Verifying the register before you hand it over

• Count the contradicted rows and check that number against your summary sentence. If you wrote "four discrepancies", count them in the table, starting from the last row and working up, because the bottom of a table is where a count goes stale.
• Confirm every unverifiable row names what would verify it and roughly what that costs in access, not in words like "further investigation". A row that cannot be closed is useful; a row that cannot be closed and does not say how is not.
• Re-read any percentage against its own base and its own implication. A 19.6 percent increase in fixture units means something entirely different depending on where the total sits against the allowance, and the sentence has to say which.
• Date the register and name the document it was read against. A register with no reference to a specific plan revision is the same problem you started with, one generation later.

References

• 29 CFR 1910.28, walking-working surfaces and fall protection in general industry; 29 CFR 1926 Subpart M, the construction counterpart, with a different trigger height
• The model plumbing code as adopted and amended by your local jurisdiction, which is the only version that binds, for fixture unit allowances at a given size and slope, cleanout spacing and access, and where an air admittance valve may be used
• Safety data sheets for any drain cleaning chemical identified on site, per the OSHA hazard communication standard at 29 CFR 1910.1200
• See related: What a Drainage Fixture Unit Is Actually Counting; How to Tell Whether a Vent Is Blocked Without Opening Anything; Sewer Camera Inspection Reference; Air Admittance Valve (AAV) vs Traditional Vent Stack Reference