Analyzing a new structure begins with a design. Analyzing an existing one begins with a question: what is actually there? Nobody hands you a model. The drawings may not exist, and where they do exist they describe what somebody intended in 1974 rather than what got built and what has happened to it since.
So the analysis is the second half of the job. The first half is characterizing the structure, and an assessment that gets that half wrong produces confident numbers about a building that does not exist.
The site visit decides everything downstream
I measure and record the framing that is accessible: member sizes, spans, spacing, connection types, species or grade where it can be determined, and the bearing condition at each end. Bearing is worth its own attention, because a joist with an inch and a half of bearing on a rotted sill is a different structure from the same joist drawn on a plan.
Then I trace the load path from the area of concern down to the ground. The symptom and the cause are frequently in different places. A floor that slopes in one room has often been caused by something two levels down: a post on no footing, a beam with a failed splice, a column bearing on soil rather than on concrete. Starting at the complaint and following the load downward finds that. Starting at the complaint and analyzing the joists does not.
Distress and damage are different
A large fraction of what worries people is cosmetic, or is the record of movement that finished long ago and is now stable. Some of it is a structure carrying more than it should. Telling those apart is most of the value of the engagement, and the evidence for it is collected on site rather than derived afterward.
Crack width, direction, and whether the edges are sharp or weathered. Whether paint or patching across a crack has re-cracked, which tells you something happened after the repair. Floor slope measured rather than felt, and measured at enough points to distinguish a local dip from a whole-building tilt. Whether a wall is out of plumb uniformly or bowed, since those have different causes. Doors and windows that bind, and whether the binding is consistent with the direction of the measured movement.
None of that is difficult. It is just a discipline about writing down numbers instead of impressions, because six months later the report is all that survives and “significant cracking” means nothing to the next reader.
Inferring the original design basis
Where it can be established, the era and the original code basis change the conclusion. A structure built to a 1974 code is not required to satisfy a 2026 one unless the use is changing or the work triggers an upgrade. Reporting an existing building as deficient because it fails a current code check, without saying that, is technically true and practically misleading.
Framing dimensions, fastener types, lumber grade stamps where visible, and the construction detailing itself all give a reasonable estimate of the vintage. That informs what allowable stresses are defensible and what the structure was originally designed to carry.
Recording what you could not see
This is the part that separates a report that holds up from one that does not.
Framing behind finishes cannot be evaluated without opening the wall. Foundations below grade cannot be evaluated without excavation. Soil behavior underneath a footing cannot be determined by looking at the building above it, and where settlement is the suspected cause, a geotechnical investigation is the next step rather than a guess.
Every assessment has these boundaries. A report that quietly omits them reads as more certain than the work supports, and the first person to discover the gap is whoever relied on it. So the limitations go in writing, and where a limit prevents me from reaching a conclusion, I say which conclusion I could not reach and what would be required to reach it.
Why the report is written for a stranger
The reader is rarely the person who hired me. It is a contractor pricing the repair, an insurer evaluating a claim, a buyer deciding whether to proceed, or a building official deciding whether to issue a permit. Sometimes it is me, three years later, when something changes and somebody asks whether the original opinion still applies.
That means the report has to contain the description of the structure, the measurements, the photographs, the reasoning, and the limitations, in a form that lets a competent reader follow the argument rather than take the conclusion on faith. The same standard I apply to a shop drawing review response applies here: state what was checked, what was found, and what remains open.
It is often a small job
Most of these are bounded work, which is exactly the kind of engagement I built the practice to take. A site visit, measurements, an analysis of the specific question, and a written opinion. Frequently the conclusion is that the condition is not structural, and that answer is worth paying for, because it ends the uncertainty and usually saves a repair that was never needed.
If something in an existing building looks wrong, or an inspector has written the sentence recommending evaluation by a structural engineer, the scope and what to send me are set out on my condition assessment page.