Flatwork Cleaning Leaves Visible Lines or Stripes

The call

A strip-mall storefront apron and connecting walkway, about 2,400 sq ft of broom-finish concrete, cleaned midweek by a two-person crew running a 5.5 GPM belt-drive rig and a 20 in surface cleaner. The tenant called the property manager two days later, not the crew's shop directly, and the property manager forwarded a phone photo taken at the end of the day when the sun sat low across the lot: parallel bands running the full length of the walkway, evenly spaced, lighter and darker in alternating strips. Not a patch, not a corner, the whole field.

Before anyone touches the site again, one thing needs stating plainly. Any product tote still on that truck from the job, sodium hypochlorite or an oxalic-acid rust treatment, stays exactly where it is until someone confirms which one, if either, was near the walkway that day, because the two release chlorine gas on contact and this callback is not the time to be moving totes around by guesswork. If a stray container turns up unlabeled or unfamiliar near the SH tote, then treat it as unknown and keep it away from the SH rather than assuming it is something else. That housekeeping check costs thirty seconds and it happens before the diagnostic walk, not after.

Ruling out the equipment first

The crew's habit on this rig is to log nozzle installs against the surface cleaner's own chart, and that log showed a fresh, chart-matched nozzle set installed three weeks and roughly nine job-hours earlier, well inside the manufacturer's stated service interval. A worn or mismatched nozzle throws a wider, weaker pattern than its rated size, and a bar spinning on an undersized flow slows enough that a hot spot can etch into a single narrow arc, but neither of those mechanisms produces a straight, evenly spaced band running the length of a driveway. A bad nozzle shows up as an uneven or scalloped edge tracking the bar's own rotation, not a clean parallel stripe the width of a whole pass. Low hours, correct chart match, and a defect shape that does not fit the fault pattern together rule this branch out without needing to pull the housing apart on site.

Skirt height came next, since a skirt set too high vents pressure sideways instead of down and weakens the outer edge of every pass. The crew's setup habit is to test-fit the skirt on an inconspicuous corner before the first working pass, same as the trade standard, and this crew's tech remembered setting it in the same 1/8 to 1/4 in range they always use, with no chatter or drag reported anywhere in the field. Nothing in the fault pattern points here either: a skirt riding too high weakens a pass's outer band consistently on every single pass in the same direction, which would read as one dull edge running the length of the walkway, not multiple alternating bright and dark bands across the whole field.

Ruling out the chemistry

The organic-stain pretreatment on this job was a standard sodium hypochlorite mix run through the injector ahead of the surface cleaner pass, and an uneven dwell, part of the field flashing dry in direct sun while a shaded section stays wet longer, is a real, documented cause of patchy results elsewhere in this trade. It was worth checking before moving on. The job record showed an overcast midweek morning with no direct sun on the apron at any point during the pretreat window, full even coverage logged at application, and no re-wet entries, meaning nothing dried out mid-dwell. A dwell problem also would not explain the specific shape reported: flash-drying tends to leave one or two irregular patches following where the sun or wind hit hardest that hour, not a repeating, evenly spaced pattern covering the entire 2,400 sq ft field from one end to the other. Weather log and fault shape both point away from this branch.

What a perfectly even, repeating pattern is actually telling you

Three candidates ruled out, and the one fact that survives every branch above is the pattern's own regularity: not one bad patch, not a wandering streak that drifts wider or narrower across the field, but bands spaced the same distance apart, over and over, the whole length of the walkway. A pattern that repeats at a fixed interval is not the signature of an operator's attention drifting over a long job, that shows up as an irregular, wandering band that widens and narrows unpredictably as fatigue sets in. A perfectly regular repeat means something constant is producing it, applied the same way, pass after pass. That constant does not have to be a piece of hardware failing. It can just as easily be a decision the technician made once, at the start of the job, and then held consistently for every single pass after it, which produces exactly the same kind of regular, repeating footprint a bad nozzle or a stuck skirt would.

Worth naming what this rules out by contrast. If the property manager's photo had shown a band that wandered, wider in one section, nearly gone in another, drifting rather than repeating at a fixed interval, the next move would not have been remeasuring the housing at all. It would have been checking whether the crew swapped operators partway through the field, or whether the job ran long enough that pace naturally slows as an arm tires over a couple thousand square feet, both of which read as irregular drift rather than a constant, repeatable error. The evenness of the spacing here is what earns the equipment-and-math branch over the fatigue branch; a different photo would have sent this diagnosis somewhere else entirely.

Measuring the actual swath instead of trusting the housing number

That points straight at overlap, and overlap is set from two numbers: the surface cleaner's actual cleaning swath and how far each pass advances past the last one. The crew's habit, common on this trade and exactly the shortcut the setup and overlap guide warns against, is to calculate overlap off the housing's nominal diameter rather than measuring the swath the nozzles actually throw on a test corner. This unit's housing reads 20 in. The tech measured the real swath on a dry test corner back at the yard afterward: 15 in, narrower than the housing because the nozzle set's spray pattern does not reach the housing's outer edge.

The overlap target logged on the ticket was 15 percent, calculated against the 20 in housing figure: 20 multiplied by 0.15 is 3 in of intended overlap, which the crew read as a net advance of 20 minus 3, or 17 in, between the center of one pass and the next. That math is internally consistent, and it is built on the wrong number. Against the real 15 in swath, advancing 17 in between passes does not overlap the cleaning coverage at all, it overshoots it: 17 in of advance against 15 in of actual coverage leaves a 2 in strip, 17 minus 15, that neither pass ever touched. That 2 in gap repeats every single time the crew advances 17 in to start the next pass, which is every pass across the whole 2,400 sq ft field, producing exactly the evenly spaced banding on the phone photo.

The measurement that confirms this over guessing is checking the actual spacing on the walkway itself against the number the math predicts. Band center to band center on the finished job measured consistently close to 17 in, matching the crew's assumed advance distance, not the swath figure alone and not some other interval a nozzle or skirt fault would produce. When the spacing you find in the field matches the number a technician calculated rather than any number tied to a hardware defect, that is the tell the cause sits in the calculation, not the equipment.

Why the wet-sheen check did not catch it mid-pass

The setup guide calls for reading the wet sheen at a low, raking angle during the pass itself, precisely so a gap like this gets caught after one bad pass instead of at the final walk. It did not catch it here for a specific, avoidable reason: the crew's mid-pass checks that day were done looking straight down at the surface from a standing position rather than crouching to a genuine low angle, and a 2 in under-cleaned strip on a broom-finish surface is close to invisible from directly overhead while it is still wet, since a wet surface reflects fairly evenly from a high viewing angle regardless of how clean it actually is underneath. The gap only became visible once the concrete dried fully and the tenant caught it in the exact low, raking afternoon light the guide's verification step exists to simulate. The check that would have caught this on pass two existed in the procedure; it was performed at the wrong angle to actually work.

Fixing it and what changes going forward

The walkway needed a second full pass, this time with the swath re-measured on a test corner rather than assumed from the housing, set to a true 15 percent overlap against the real 15 in figure, which is a 2.25 in overlap and a 12.75 in net advance rather than 17. Re-running a fresh pass at a corrected, narrower advance closed every gap cleanly, since the gaps were narrow and untreated concrete rather than damaged concrete, and a second correctly-set pass covers untouched ground the same as a first one would.

Two things change for the shop, not just this crew. First, any time a skirt or housing gets swapped between rigs, even a same-model swap, the actual swath gets measured fresh on a test corner before the job starts; a housing's stamped diameter is a manufacturing spec, not a promise about what that specific unit throws once nozzles, wear, and skirt fit are all accounted for. Second, the mid-pass sheen check gets performed the way the setup guide actually describes it, crouched to a genuine low angle on a marked section of every job over a certain size, not eyeballed from standing height, because the whole value of a real-time check is catching a bad pass before it repeats across a field this size.

Job size is what turned a 2 in miscalculation into a callback instead of a shrug. The same wrong math on a typical two-car residential driveway, maybe 40 ft of travel instead of a couple hundred, produces the identical gap at the identical spacing, but it shows up as three or four faint bands rather than dozens marching the length of a commercial apron in a customer's phone photo. The underlying error does not scale with the job, only its visibility does, which is exactly why a shop that only catches this on the big, obvious jobs is still missing it on every smaller one where nobody happened to look at a low enough angle to see it.

References

• See related: Surface Cleaner Flatwork Procedure Standard, Surface Cleaner Setup and Overlap Technique, Nozzle Tip Color and Spray Angle Reference.
• Manufacturer's nozzle and flow chart for the specific surface cleaner model, since swath and overlap figures are model-specific.