How to Record a Drainage System So the Next Camera Run Is Shorter

Why this matters

The camera work on a repeat visit is fast. What is slow is everything before it: finding the access, deciding which way is downstream, working out what the fittings are, and then discovering that the last tech's distances cannot be compared with yours because nobody wrote down where zero was. That is the hour you are paying for, and it gets paid again on every visit for the life of the property.

A record that fixes the datum, the direction and a handful of permanent features turns the next run from a survey into a comparison. The difference is not documentation for its own sake; it is that a comparison can detect change and a survey cannot.

Before the reel comes off the truck

• A manhole, wet well or interceptor is a permit-required confined space, and the entry permit, atmospheric testing and attendant obligations attach under 29 CFR 1910.146 in general industry or 29 CFR 1926 Subpart AA on a construction site before anyone's shoulders pass the plane of the opening. Reaching in with a camera from outside is not entry; leaning in far enough that your head is inside is.
• Power the reel and monitor through ground-fault circuit-interrupter protection whenever you are working in a wet basement, at an outside cleanout or anywhere the cord can lie in water, which is required for temporary power on construction sites at 29 CFR 1926.404(b)(1) and is the right practice everywhere else.
• If the access sits in a street or driveway apron, set the work zone before you set the camera, with traffic control conforming to the manual your state department of transportation has adopted, and wear high-visibility apparel. A tech kneeling at a cleanout is below the sight line of most vehicles.
• Locating with the sonde means walking the line with a receiver, and marking the surface with paint or a flag rather than a stake driven into ground with unlocated utilities in it.

What actually costs the time

Time a repeat run and the camera is a small share of it. The three expensive activities are all forms of re-deriving something the previous visit already knew:

Re-establishing the datum. Which access point, and where on it the counter reads zero. Two techs entering the same cleanout can differ by more than a foot depending on whether they zeroed at the cap face, at the top of the riser, or at the point the head turned into the horizontal.

Re-establishing direction. Which way is downstream, which branch is the building and which is the street, and which of the two cleanouts in the yard is the one that reaches the segment in question.

Re-identifying the fittings. Every wye, bend, transition and tap has to be recognised again from scratch, and on an old line with mixed materials that is most of the run.

Each of those is a fact that does not change between visits. Any fact that does not change should be written once.

The record: nine fields and what each one removes

Each field exists because of one specific ambiguity, and the ambiguity is the reason to fill it in rather than the field itself.

1. Access point, named and photographed. Not "the cleanout" but which one, with a photo showing it in the context of two permanent features, a corner of the building and a fence post. Removes: arriving and not knowing which lid.
2. Zero convention. The physical point the counter was zeroed at, stated in words. Removes: the single largest source of disagreement between two runs, and the one nobody suspects.
3. Direction of travel and which way is downstream. Removes: a whole run performed up the wrong branch.
4. Instrument identity. Which reel and counter, by the shop's own tag. Removes: the ambiguity about whether a difference between runs is real, for the reason worked below.
5. Feature list with stations. Every wye, bend, transition, coupling and tap, each with its counter reading. Removes: re-identifying the run, and provides the fiducials the next field depends on.
6. Material and size, with the station of every change. Removes: choosing the wrong cutter or the wrong jetting pressure on the next visit.
7. Condition observations with station and clock position. Twelve o'clock is the crown, six is the invert. Removes: "there was a crack somewhere around forty feet."
8. Sonde marks. Any station physically located from the surface, recorded with its depth and with offsets to two permanent surface features. Removes: relocating from scratch before a dig.
9. Line state at the time of the run. Cleaned or not, by what method, how long before. Removes: comparing a run through a dirty line with a run through a clean one and calling the difference deterioration.

The rule that outranks all nine

Record every position twice: once against the counter, and once against the nearest permanent feature.

A counter reading depends on the instrument, the entry point and the zero convention, all three of which change between visits. A distance from the second wye does not. "Root entry at the transition joint" survives a new tech, a new reel and a different cleanout; "root entry at 34 feet" survives none of them.

This is also what makes the record useful to somebody quoting a repair without a camera in hand, because a feature-relative position can be checked against the building's own layout.

The filled-in record

One residential building sewer, first run:

Access:  outside cleanout, south side, 4 ft east of the
         hose bib, 3 ft off the foundation. Photo 01.
Zero:    counter zeroed at the face of the cleanout cap.
Travel:  toward the street. Downstream = away from house.
Reel:    shop tag R-2. Counter: 2 percent of reading.
State:   run AFTER cabling, 20 minutes after, line clear.

  Station  Feature / observation        Clock
  ------   ---------------------------  -----
   12.0    wye, branch from the west       9
           = 3 ft downstream of the SE building corner
   27.0    45 degree bend, left
           = at the fence line
   34.0    transition, cast iron to clay
           = 7 ft past the bend
   34.0    root entry at the transition    4
   52.0    tap, active, small branch       2
           = 18 ft past the transition
   68.0    connection to the main
           = at the curb face

Sonde:   34.0 ft located and marked. Depth 4.5 ft.
         Offsets: 11 ft from the SE building corner,
         6 ft from the fence line.
Video:   file 2 of the visit; stills at 34.0 and 52.0.

Fourteen months later, a different tech runs the same line with reel R-5, and zeroes the counter at the point the head turns into the horizontal, which is 1.4 feet downstream of the cap face. The transition reads 32.7 feet.

Reading the second run against the first

Raw, the transition appears to have moved 34.0 - 32.7 = 1.3 feet upstream, which is not a thing pipe does. The record is what lets that get resolved instead of argued about.

Correct for the zero convention. Run two's zero is 1.4 feet downstream of run one's, so run two's readings are all short by that amount relative to run one's datum:

• 32.7 + 1.4 = 34.1 feet on run one's datum
• Residual against run one: 34.1 - 34.0 = 0.1 feet

Now decide whether 0.1 feet means anything. The counter's error is stated as a percentage of reading rather than a fixed count, so it grows with distance, and on a single run with a single instrument it behaves as a systematic offset. Two different instruments were used here, so the two offsets are separate worst-case bounds and they add linearly rather than cancelling:

• 2 percent of 34 feet on R-2 = 0.68 feet
• 2 percent of 34 feet on R-5 = 0.68 feet
• Combined bound on the difference: 0.68 + 0.68 = |difference| <= 1.36 feet, written as a bound with one inequality sign rather than as a plus-or-minus interval, because worst-case bounds are not a spread

The observed residual of 0.1 feet sits well inside 1.36 feet, so there is no evidence of anything having changed. Note what would have happened with the same instrument on both runs: the offset would have been common to both readings and would have cancelled in the difference, leaving 2 percent of the 0.1 foot difference, which is nothing. Field four exists so you know which of those two cases you are in.

And the feature-relative reading needed none of that. Both runs recorded the root entry as being at the transition joint, so the observation "the root entry is at the material change" is identical on both runs with no correction, no instrument, and no convention. That is the reason for the rule that outranks the fields.

What the record bought, in labor hours. Both figures are field labor hours for one tech so they are the same kind of hour:

• Run one, including finding the access, establishing direction and identifying every fitting: 2.0 hours
• Run two, with the record in hand: 0.6 hours
• Ratio: 2.0 / 0.6 = 3.3 times faster

The 0.6 hours is not a shorter camera run. The camera portion was about the same both times. What disappeared was the re-derivation.

What the record does not fix

It does not survive a line that was altered. A record made before a remodel that added a bath describes a system that no longer exists, and the tell is a feature on the run that is not on the list. Treat one unexplained feature as a reason to re-record the whole line rather than to patch the old record.

It does not make an undocumented condition appear. A crack at 8 o'clock that the first tech did not notice is not in the record, and the second run finding it is not evidence that it is new. Only conditions actually recorded can be trended, which is an argument for recording the boring observations too: a joint noted as sound at the first run is what makes the same joint admitting water at the second run a finding.

It does not settle a repair location on its own. The counter tells you how long a defect is; the sonde tells you where it is. The record carries both because they answer different questions, and the arithmetic separating them belongs to the sibling card on choosing between cleaning and excavation.

References

• 29 CFR 1910.146, permit-required confined spaces, general industry; 29 CFR 1926 Subpart AA, confined spaces in construction
• 29 CFR 1926.404(b)(1), ground-fault protection for temporary power on construction sites
• The temporary traffic control manual adopted by your state department of transportation, which reaches the shop through the permit or the right-of-way authority
• Camera and locator manufacturer documentation for footage counter accuracy and sonde depth range
• See related: Sewer Camera Inspection Reference; How to Decide Between Jetting, Cabling and Excavation
• See related: Why a Drainage Fault Is Usually a Slope Fault in Disguise