Two reductions, not one
A damage factor measures the wall. A confidence factor measures your file on it. They enter the calculation at different points, and lumping them together is both unconservative and unauditable.
When you assess an existing masonry wall, you reduce its capacity. Most of us apply a single reduction. There should be two — and collapsing them into one is what stops the report answering the only question the client actually asks: so what do we do about it?
The two reductions are not the same kind of thing.
The first is about the wall. It is physical: rot, cracking, weathering, insect attack, loss of section. Aligned with ICOMOS ISCARSAH guidance and the ASCE 41 strength-reduction ranges, a condition rating maps to a capacity factor — for example 1.00 undamaged, 0.90 minor, 0.75 moderate, 0.50 severe. It multiplies the member's capacity.
The second is about your file. EN 1998-3 §3.3 (and the Italian Circolare 2019 in the same spirit) says the completeness of your investigation defines a knowledge level, and the knowledge level sets a confidence factor: KL1 limited → CF 1.35, KL2 normal → CF 1.20, KL3 full → CF 1.00. It divides the characteristic strength.
One describes the condition of the fabric. The other describes the quality of your own information about it. A perfectly sound wall you have never tested and a visibly cracked wall you have cored and tested are different problems, and a single factor cannot represent both.
They do not even enter the calculation at the same point
The confidence factor sits with the material safety factor:
f_d = f_k / (γ_M · CF)
The damage factor derates the resulting member capacity. They are applied to different quantities, which is the practical reason they cannot legitimately be merged into one number.
There is a detail here worth stating, because it is easy to get wrong: the confidence factor is not applied to stiffness. In the EN 1996-1-1 Annex G slenderness route the f_k/E ratio keeps the unreduced values, because E is an estimate of stiffness, not of strength. CF expresses uncertainty about strength. Dividing E by it as well is a reduction with no clause behind it.
The number
Take the ordinary case: normal knowledge (KL2), moderate observed damage.
0.75 / 1.20 = 0.625
You are designing on 62.5% of nominal capacity. If instead you had reached for a single "conservative" 0.7 — which is roughly what gets written when the two are collapsed — you would be about 12% unconservative, while believing you were being careful. That is the worst available outcome: less safe and less defensible.
Why the separation earns its keep
Keep them apart and the assessment tells you what to do next.
| Action | Effect | Net factor | Gain |
|---|---|---|---|
| (baseline) KL2, moderate damage | — | 0.625 | — |
| Extended testing → KL3 | CF 1.20 → 1.00 | 0.750 | +20% |
| Repair fabric → minor damage | 0.75 → 0.90 | 0.750 | +20% |
The two routes recover exactly the same capacity. Which means the decision is now purely a cost comparison — a testing campaign versus a physical intervention — and that is a conversation a client can actually have, and an engineer can actually justify.
With one lumped factor, you cannot even pose the question. You have a number, and no way to say which part of it you could buy back, or how.
And it is the conservation-correct answer
Derating an existing member to reflect its real condition, rather than condemning and replacing it, is the structural expression of minimum intervention — Venice 1964, Burra, Nara. A single pessimistic lumped factor quietly pushes the other way: everything fails, so everything gets replaced, and authentic fabric is lost to a modelling convenience.
A reduction you cannot attach a clause to is the one that gets questioned when the report is challenged. Two factors, each with its source, survive that conversation. One factor, chosen conservatively, does not.