The drawings for a retaining wall describe a structure that may or may not exist. Reinforcement lengths, vertical spacing, backfill gradation, compaction and the drainage detail are all buried within days of being built, and after that the only account is the paperwork, where paperwork exists at all. When such a wall fails, a large share of the investigation is a controlled excavation of the backfill to establish what was actually built, what was placed behind it, and what was added above it since.
Specified backfill versus placed backfill
Reinforced-soil wall designs specify backfill by gradation and plasticity, and the reason is mechanical rather than cosmetic. The soil-to-reinforcement interaction that develops pullout resistance depends on granular material, and fines make the mass moisture-sensitive and slow to drain. FHWA's MSE wall guidance and the NCMA segmental retaining wall design manual both set limits for the reinforced zone.
Substitution is common and rarely documented, because importing conforming granular fill costs money and the soil on site is already there. Sampling recovered backfill and classifying it under ASTM D2487, with sieve gradation and Atterberg limits, establishes what was placed against what the specification allowed.
Compaction, the parameter nobody can see
Compaction governs the unit weight, strength and stiffness the design assumed, and it becomes invisible the moment the next lift goes down. Field density testing during construction, referenced to a laboratory Proctor curve, is the contemporaneous record, and its complete absence across a wall is itself a finding.
After the fact, in-place density can still be measured on undisturbed portions of the backfill and compared against the same laboratory maximum. Under-compaction tends to announce itself as settlement behind the facing, voids beneath pavements or slabs at the crest, and measured strength lower than the design assumed.
Over-compaction is a failure mode too
Heavy compaction equipment operating close behind a wall face induces lateral pressures well above design values and can push the facing outward during construction. Design guidance limits equipment size within a defined zone behind the facing for exactly this reason. A wall out of alignment before it ever carried service load, documented in construction photographs or survey, points here rather than at anything later.
Reinforcement length and spacing
Length controls pullout resistance and the external geometry of the reinforced mass; vertical spacing controls how much load each layer carries. Both are cheap to shorten in the field and hard to detect once buried. Excavation locates each layer, measures its embedded length and elevation, and identifies the product, all of which can be checked against the design submittal.
Recurring findings include layers truncated where an obstruction was met, spacing opened up toward the top of a wall, reinforcement of a different grade than specified, and layers omitted where a utility crossed. Each is a discrete departure from what the design required at that elevation.
Obstructions and the details nobody drew
Real sites contain utilities, catch basins, light-pole foundations and building corners inside the reinforced zone. Every one of them displaces reinforcement, and the correct treatment, whether splaying layers around the obstruction or a design revision, is engineering work that costs time in the field.
Where it was skipped, the excavation finds reinforcement cut, bent around the obstruction, or simply stopped short. These local defects explain local failures. A bulge at one panel while the rest of a long wall performs well usually has an explanation of this kind rather than a global one.
Surcharge added after the wall was built
A wall is designed for a stated surcharge at the crest. Material stockpiles, a parking area, a new building foundation, a pool, fill placed to level a yard, or equipment tracking above the wall all add load the design may never have contemplated. AASHTO's provisions treat vehicular surcharge as an equivalent height of soil; a wall designed with no traffic above it that later got traffic carries an added load continuously.
Aerial imagery, permit files, listing photographs and grading records date those changes. A wall that performed for years and failed after something changed above it is a materially different case from one inadequate from the first day.
Running an excavation that holds up
The excavation is destructive, which means the documentation produced during it is the only version anyone will ever have. Practice is to survey and photograph the wall before disturbance, open test pits sampling both distressed and apparently sound sections, log each lift and reinforcement layer with elevation and measured length, sample soil at known depths, and cut reinforcement coupons under documented chain of custody.
Sampling the sound sections matters as much as the failed one. A defect found only where the wall came down proves less than the same defect found throughout, and a control section is what turns a single observation into a pattern.
What the paper record adds
Design submittals, material certifications, delivery tickets, inspection reports, daily construction logs, change orders and progress photographs supply the sequence the excavation cannot. Delivery tickets establish whether conforming fill was imported and in what quantity; certifications establish which reinforcement product was supplied; inspection reports establish who was present as the lifts went in.
Where the physical evidence and the paper record disagree, the physical evidence generally governs, and the disagreement itself is often more informative than either source alone.
Preserving the option to do this at all
Every one of these findings depends on the failed wall and the backfill behind it staying where they are. Emergency stabilization is sometimes unavoidable, but there is a wide gap between shoring a hazardous wall and hauling the reinforced zone to a spoil pile. Documenting condition before equipment arrives, keeping removed material separated and identified, and scoping the excavation before repair begins preserve the record cheaply.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.