Diagnosing a Fault That Is Really a Bad Original Design
Why this matters
Telling a customer their system was designed wrong is a serious claim, and you had better be able to prove it. A hunch will not survive a second opinion or a warranty dispute. This card is the method: how to build the evidence that a fault is a design problem, and how to classify which kind, so your recommendation stands up. Get this right and you save the customer from paying for repairs that can never work. Get it sloppy and you are the guy who blamed the design because you could not find the fault.
Five families of design fault
Most design faults fall into one of five buckets. Name the family before you write it up.
| Family | What it means | A generic example |
|---|---|---|
| Undersized | Capacity is below the real load | AC that cannot keep up on design days; a pump that never hits required flow |
| Layout or routing | The pieces are right but arranged wrong | Long dead-leg runs, a drain with no fall, a return in the wrong place |
| Component mismatch | Correctly sized parts that fight each other | A coil paired to the wrong-capacity air handler; a pump curve that does not match the head |
| Missing component | A required piece was never installed | No expansion tank, no equipment ground, no trap, no isolation valve |
| Code or spec violation | Built below a written minimum | Undersized conductor for the load; insufficient combustion air |
Build the evidence, do not assert it
A design claim needs three things on paper before you say it out loud.
- The target: what the system was supposed to deliver (from the nameplate, the spec, the code minimum, or a load estimate).
- The measured reality: what it actually delivers, measured, not guessed.
- The healthy-parts finding: proof the individual components are working, so the gap is not a failure.
When you have a real target, a measured shortfall, and healthy parts, you have a design fault. Missing any one of the three and you are still guessing.
Measure against the spec, not against "seems low"
"Seems weak" is not evidence. Put a number on the demand and a number on the delivery and compare them to the design target.
- Cooling or heating that never satisfies on a peak day, with clean components and correct charge, points to undersized capacity.
- A circuit that drops voltage or trips under a load that is within its intended use points to an undersized conductor or breaker.
- Fixtures that lose pressure only when several run at once points to undersized supply, not a failed part.
- A pool or spa that never clears with a clean filter and correct runtime points to a pump or plumbing sized below the turnover the water needs.
The pattern across trades is the same: the shortfall shows up under load, and the parts are fine.
Rule out the ordinary failures first
Do not reach for "bad design" because the real failure is hard to find. That is the lazy version of this diagnosis and it is often wrong. Confirm the boring causes are clear first: correct charge or pressure, clean filters and coils, tight connections, no leaks, correct settings, no fouling. Only when the ordinary failure list is genuinely ruled out does a design fault become the leading answer. A design verdict reached because you gave up is worse than no verdict.
Document so it survives a second opinion
Write it as if the next tech, the manufacturer, or a lawyer will read it, because one of them might.
- State the target, the measured reality, and the gap in numbers.
- Note which family the fault is (undersized, layout, mismatch, missing, or code).
- Photograph the evidence: the nameplate, the routing, the missing component, the meter reading.
- Separate what a repair can do from what only a redesign fixes.
A clean file turns "he blamed the install" into "he showed me the numbers." That is the difference between a claim and a diagnosis.
References
- Trade-standard practice for load calculation and equipment selection
- Manufacturer application/design data and installation requirements
- Applicable mechanical, plumbing, and electrical code minimums
- See related: Telling a Design Flaw Apart From a Failure; Reading the Design Intent to Know What Should Be Happening