Capacity Assessment: Retrofit or Replace?

By Mubashir · Senior Structural Engineer · May 2026

Every ageing steel structure reaches a point where the question is no longer "maintain" but "retrofit or replace?" The answer is not instinctive — it emerges from a structured engineering assessment that evaluates the structure's actual capacity against current requirements. Getting the answer wrong in either direction is costly: premature replacement wastes a serviceable asset, while delayed replacement creates liability and safety risk.

This article lays out the framework our team uses for structural capacity assessment, the typical retrofit options available, and the triggers that indicate replacement is the more rational path.

What Capacity Assessment Actually Is

A capacity assessment evaluates an existing structure's ability to carry current and future loads, measured against today's code requirements — not the code that existed when the structure was designed and built. This distinction matters because codes evolve. Seismic hazard maps are updated as our understanding of geophysical risk improves. Live load requirements change with occupancy. Wind provisions are revised after storm-damage studies. A structure that fully complied with its original design code may have a meaningful gap against the current code, particularly for seismic loading.

The assessment is not a pass/fail exercise at the first check. It quantifies the utilisation ratio of every demand-critical member and connection under each governing load combination. A member at 92% utilisation under the current code is not a failed member — it is a member with 8% reserve, which may or may not be acceptable depending on consequence of failure and code interpretation. Context matters.

The Assessment Process Step by Step

A rigorous capacity assessment follows a consistent sequence:

1. Visual inspection. A thorough walk-down of the structure to catalogue visible deterioration: surface corrosion, section loss, deformation, weld cracking, missing fasteners, paint system condition. The inspection forms the basis for deciding where destructive testing is warranted.
2. As-built drawing review. Original drawings are the starting point for the analytical model. When drawings are missing — common in structures built in the 1980s or earlier — field measurements and partial reverse-engineering are required. This adds time and uncertainty to the assessment.
3. Material testing. For older structures, mill certificates may not be traceable. Hardness testing (Brinell or Vickers) can estimate yield strength; core samples from non-demand-critical locations can provide material certification. For critical welds, ultrasonic testing (UT) may be specified.
4. Analytical model build. A 3D structural model is built (or updated from original design model) using current member sizes, accounting for any section loss identified in the inspection. Load cases are defined per the current governing code.
5. Code check against current standard. Each member and connection is checked against current provisions — for steel, this means AISC 360, CAN/CSA S16-19, or the applicable national standard. Utilisation ratios are extracted for every element.

Triggers for Reassessment

Not every ageing structure needs immediate assessment. The triggers that warrant commissioning a formal capacity assessment are:

• Change of use. A warehouse converted to a gymnasium, a roof deck now carrying mechanical equipment, a mezzanine added to an industrial building. Any increase in imposed loads above the original design basis requires verification.
• Code update affecting seismic or wind demand. Seismic hazard maps in Canada (NBC 2020), the USA (ASCE 7-22), and Japan have all been revised in the past decade. If an existing structure was designed to an earlier edition, the current seismic or wind demand may exceed the original design load.
• Detected deterioration. Corrosion visible on primary structural members, fatigue cracking at connections in dynamically loaded structures, or evidence of overload damage (permanent deflection, buckled members) all require assessment before continued operation.
• Addition of equipment loads. New HVAC units, conveyors, cranes, or process equipment mounted to an existing structure require verification that the structure can carry the added load.

Retrofit Options Available

When the assessment reveals deficient members or connections, retrofit rather than replacement is the first option to evaluate. Common interventions include:

• Section plating. Welding additional plate to the flanges or web of an under-capacity member to increase its cross-sectional area and moment capacity. Effective for individual members with modest deficiency; becomes uneconomic if many members require plating.
• Adding members. Installing new bracing members to reduce unbraced length, new columns to reduce tributary span, or supplementary purlins to reduce secondary member loads. This changes the load path and requires a full re-analysis after addition.
• Connection strengthening. Adding welds or bolts to under-capacity connections, replacing bearing-type bolted connections with slip-critical assemblies where required for seismic ductility, or replacing corroded gusset plates entirely.
• Base plate repair or replacement. Corroded base plates, oversized anchor rod holes from drift or construction tolerance, and inadequate anchor rod embedment are common findings. Replacement base plates with new epoxy-anchored rods are a standard intervention.

Replacement Triggers

Retrofit is not always the right answer. The following conditions typically indicate that replacement is the more appropriate path:

• Section loss exceeding 30% from corrosion. At this threshold, the remaining section is geometrically compromised in ways that plating cannot efficiently restore. Corrosion at the web-flange junction, in particular, is difficult to plate effectively.
• Fatigue cracking in demand-critical members. A crack in a primary member subject to cyclic loading is a fracture mechanics problem, not a welding problem. Grinding and re-welding does not restore the fatigue life of a cracked detail; the member or connection must be replaced.
• Structure cannot meet current seismic provisions without complete redesign. This is common in structures originally designed before modern seismic codes — the lateral system is fundamentally non-ductile, and retrofitting for ductility would require modifying nearly every connection in the structure.
• Retrofit cost approaches or exceeds 60–70% of replacement cost. When the engineering estimate for retrofit scope reaches this threshold, the economics of replacement become compelling, especially when replacement also delivers improved performance, reduced maintenance, and extended service life.

Ontario Case Study: P-2022-044

The Ontario steel replacement project illustrates the assessment-to-replacement pathway clearly. The existing steel tower, originally designed in the early 2000s, was assessed under NBC 2020 following a change in occupancy classification that required re-examination of the seismic design basis.

The NBC 2020 seismic hazard maps, updated from the previous 2005 edition, showed higher spectral accelerations at the Ontario site than were used in the original design. When the existing structure was checked against these updated demands, multiple primary members and base connections were found to be over-utilised under the governing seismic load combination.

The retrofit option was scoped: plating of primary columns, replacement of base connections, and addition of new bracing. The retrofit cost estimate reached 68% of the replacement cost for equivalent new construction. Combined with the fact that replacement would deliver a fully code-compliant structure with a 25-year design life from completion — compared to a retrofitted structure with documented residual deficiencies and a shorter effective service life — the client elected replacement.

The replacement structural design was developed concurrently with the assessment, so the decision to replace did not restart the project timeline from zero. The design documentation package was submitted for permit within weeks of the replacement decision.

Frequently Asked Questions

How long does a capacity assessment take?

A typical capacity assessment for a single steel structure of moderate complexity takes 3–6 weeks from receipt of existing drawings and inspection access. This includes the inspection, drawing review, model build, code check, and preparation of the assessment report. Structures with missing drawings, extensive corrosion requiring material testing, or complex geometry take longer. For structures requiring permit submission, allow additional time for the authority's review cycle.

What does structural capacity assessment involve?

A capacity assessment involves four phases: physical inspection of the structure (visual and potentially NDT); recovery or reconstruction of as-built documentation; analytical modelling and code checking against the current governing standard; and preparation of a report that states the utilisation ratios of critical members, identifies deficiencies, and recommends remediation. The output is an engineering basis for the retrofit-or-replace decision, not simply a condition rating.

When is replacement more cost-effective than retrofit?

Replacement becomes cost-effective when: the scope of required retrofits is extensive (plating many members, replacing most connections); the structure has fundamental system-level deficiencies that cannot be corrected by member strengthening; the remaining service life of the retrofitted structure is significantly shorter than a new replacement; or the retrofit cost estimate approaches 60–70% of replacement cost. In each case, the engineering and financial analysis should be presented transparently so the client can make an informed decision.

Can you assess a structure where the original drawings are missing?

Yes. Where original drawings are unavailable, we conduct field measurements to determine member sizes, connections details, and geometry, and build an as-found model from that data. Material properties are estimated from hardness testing or assumed conservatively per the likely steel grade available in the construction period. The resulting assessment has wider uncertainty bands than one based on verified drawings, which is reflected in the report's confidence ratings and may result in more conservative recommendations.