Elevation Survey and Floor Level Measurement

Purpose

A floor elevation survey is a relative measurement. It says nothing about where the floor was built and everything about where it sits now compared to the rest of itself. That makes four things load-bearing: the datum you chose, the instrument and its resolution, the finish-thickness corrections you applied, and the closure check that proves the survey did not drift while you walked it. Drop any one and the numbers still look like data.

There is no code tolerance for the floor of an existing house. ACI 117 sets tolerances for new concrete construction and is not a repair acceptance criterion, so a shop that wants a threshold has to set one and print it. This procedure fixes the method so that a survey run today and a survey run after a lift, by two different techs, are comparable to each other.

Scope

Covers relative floor elevation measurement on slab-on-grade, basement slabs, and pier-and-beam floors, for diagnosis and for before-and-after comparison across a lift.

Does not cover crack mapping or monitor installation, which the initial assessment SOP owns. Does not cover the lift itself or its incremental stop rules, which the lifting and releveling SOP owns. Does not attempt to recover the as-built elevation, which is not knowable from a survey and must never be claimed to a customer.

Roles and responsibilities

Procedure

Step 1: Pick the instrument and write its resolution on the sheet

Match the instrument to the decision. A water manometer resolves to roughly an eighth of an inch in practice and is adequate for a coarse tilt. A precision altimeter-type gauge resolves to about a hundredth of an inch and is the tool for pre-lift and post-lift comparison. A rotary laser with a receiver resolves to the band you select on the receiver, commonly a sixteenth on residential work. Acceptance: instrument, model class and resolution written at the top of the point list before any reading is taken. Wrong looks like a survey with no instrument named, which cannot be compared to any other survey at that address. Send it back rather than guessing which tool produced it.

Step 2: Verify the instrument before the first reading, not after the last

Take two readings at one point, move the reference, and read the same point again. Acceptance: the two agree within the instrument's stated resolution. Wrong looks like a spread larger than that, which means the reference is unstable, the hose has air in it, or the receiver band is set finer than the head can hold. Re-seat and re-verify; if it fails twice, swap the instrument and note the swap, because a survey run on a drifting instrument is not salvageable by arithmetic afterward.

Step 3: Set a datum you can find again in a year

Choose a point that will still exist after flooring changes: a bearing wall base, a stair nosing, a garage slab corner. Mark it physically and photograph it in context. Acceptance: the mark is described in words precise enough for a different tech to stand on it without you. Wrong looks like a datum at the middle of a room with no reference to anything permanent. Hazard: never drill, core or saw a mark into a post-tensioned slab. Tendons are stored energy and cutting one releases it violently. Identify post-tension construction from the edge anchor pockets and the required placard at the garage or entry before any anchor goes in, and where the construction is unknown, mark with paint and a photograph rather than a drilled pin.

Step 4: Lay out the grid and number the points before you read any of them

Shop standard, tune to your housing stock: a grid no coarser than 8 ft on open floor, plus a reading at every exterior corner, at every interior bearing line, and at the center of every room, and no fewer than 20 points on a house under about 2,000 square feet. Number the points on the sketch first. Acceptance: the sketch is complete and numbered before the first reading. Wrong looks like reading as you wander and numbering afterward, which produces a point list nobody can re-walk. Hazard: clear the walking route first; a survey walked backwards past a floor register, a stair opening or an open crawl hatch is how a surveyor goes through a floor.

Step 5: Read on a fixed route with the same rod foot every time

Walk the numbered points in order and hold the rod or the module the same way at each: same foot, same pressure, same orientation. On carpet, seat the foot through the pile to the substrate rather than reading the pile top. Acceptance: every numbered point has a reading, in order, with no gaps. Wrong looks like a skipped point recorded later from memory. Go back and read it. Hazard: pier-and-beam readings taken from the crawlspace require the crawlspace entry controls, including air testing where a fuel-burning appliance, standing water or a sewer defect is present, and no crawling under visibly deflected framing or an unfooted jack post.

Step 6: Print the finish-thickness correction as its own line

Where the floor covering changes, the top surface you are reading changes with it. A tile bed sits proud of adjacent hardwood, an added underlayment sits proud of the original subfloor, a self-leveler poured by a prior contractor can hide the very dish you were sent to find. Measure the step at each transition and record raw reading, correction, and corrected value as three separate numbers on the same row. Acceptance: every point on a different covering than the datum carries all three numbers. Wrong looks like a single corrected figure with no raw value behind it, which nobody downstream can audit. An uncorrected tile reading manufactures a hump that is not there and sends the crew looking for a plumbing leak.

Step 7: Close the loop back to the datum and apply the closure rule

Return to the datum point as the last reading of the survey and compare it to the first. Shop standard, tune to your instrument: closure within 0.10 in for the altimeter-type gauge, within 0.25 in for a water level or a laser set to a sixteenth band. Acceptance: the closure figure is written on the sheet whether it passed or failed. Wrong looks like a closure larger than the limit. Do not distribute the error across the points, because you do not know where it entered. Discard the run, find the cause, and re-survey.

Step 8: Reduce to relative elevations and state the convention in the same breath

Convert to relative values against the highest reading, which becomes 0.0, with every other point negative and expressed in inches below it. Say that convention on the sheet and in any heading, because a survey read under the opposite convention inverts every conclusion drawn from it. Acceptance: the point list, the convention statement, and a contour at a stated interval. Wrong looks like a contour drawn without the point list behind it; the contour is a picture, the point list is the record.

Step 9: Interpret against written thresholds, and name what would change them

Shop defaults, no code number exists here, tune them to your soils and your market: under 1 in of differential per 20 ft with no corroborating cue is monitor-only; 1 to 2 in per 20 ft is investigate and correlate against cracks, drainage and vegetation; over 2 in per 20 ft is scope for repair. Separately, any local slope reversal exceeding 0.5 in across one 8 ft grid interval is a heave signature and gets a moisture and plumbing investigation before any pier is scoped, because piering a heave drives the problem the wrong way. Acceptance: the band and the trigger written on the record. Wrong looks like a repair scoped off a tilt number with a reversal sitting unexplained in the same data set. Stop and route to a design professional where the reversal crosses a bearing line.

The record this produces

One survey sheet per visit: instrument, resolution and the step 2 verification result; datum description and photo; the numbered grid sketch; a row per point carrying raw reading, correction and corrected value; the closure figure and whether it passed; the relative reduction with its convention stated; the contour; and the threshold band with its trigger.

The estimator prices from the corrected column. The lift crew uses this sheet as the pre-lift baseline and re-runs the identical grid afterward, which is only possible because the datum and the point numbers are on it. The office schedules the re-survey. Where a claim follows, the closure line is the field that decides whether the survey is evidence or an opinion.

One worked pass

Slab-on-grade, 32 ft by 40 ft, altimeter-type gauge resolving to about a hundredth of an inch. Step 4 laid out an 8 ft grid, five points by six, thirty points, plus corners and room centers. Datum set at the garage slab corner, marked and photographed; the slab carried no post-tension placard and no edge anchor pockets, so a pin was acceptable.

Step 7 failed. Closure came back at 0.22 in against the 0.10 in limit for this instrument. Under the stop rule the run was discarded rather than adjusted. The cause was found in five minutes: the reference module had been nudged when a chair was moved past it in the dining room. It was re-set, the datum re-established, and the survey re-run in full. Second run closed at 0.03 in, inside the limit, and that figure is on the sheet.

Step 6 then produced the finding. The high point sat at the front left corner. The rear right corner read 4.0 in below it over the 51 ft diagonal, which is about 1.55 in per 20 ft and lands in the investigate band. Across the kitchen the raw readings ran K1 at 1.0 in below datum, K3 at 0.4 in below, K5 at 1.1 in below, a 0.6 in rise across the 8 ft interval from K1 to K2. That is above the 0.5 in reversal trigger and would have opened a plumbing investigation under the kitchen slab.

K3 is on tile over a mortar bed measured 0.5 in proud of the adjacent hardwood. The row reads raw 0.4 below, correction 0.5 for added finish thickness, corrected 0.9 below. Against corrected values the kitchen runs 1.0, 0.9, 1.1 below datum, a 0.1 in variation across the 8 ft interval, well under the 0.5 in trigger. There is no heave. The kitchen is simply part of the global tilt, and the investigation went to the perimeter drainage and the tree line rather than under the slab.

Two numbers carried this survey: a closure figure that condemned the first run, and a correction line that would have sent a crew to the wrong side of the house if it had been folded silently into a single value.

References

• ACI 117, Specification for Tolerances for Concrete Construction and Materials, cited here only to note that it governs new construction and is not an acceptance criterion for an existing floor.
• ACI 562, Code Requirements for Assessment, Repair, and Rehabilitation of Existing Concrete Structures, in the edition your jurisdiction has adopted, which reaches you through the permit.
• Post-tensioned slab identification and the prohibition on cutting tendons: the slab manufacturer's or the Post-Tensioning Institute's published guidance, and the placard required at the structure.
• See related: the initial assessment SOP for crack mapping and monitors; the lifting and releveling SOP, which re-runs this identical grid before and after a lift.