Study CSWIP underwater inspection by separating every scenario into three decisions: what was observed, whether the observation quality supports a conclusion, and what the report should state. Match each defect type to a technique whose underwater limitations you can name, and keep every written conclusion inside the evidence.
Splitting the Diver's Observation From the Inspector's Conclusion
CSWIP is a role-specific personnel certification scheme run by TWI Certification Ltd. Underwater inspection competence lies in turning observations relayed from below into documented, defensible conclusions rather than in diving skill alone.
CSWIP certificates exist to assure industry that a person is competent for a specific job role, assessed under the ISO/IEC 17024 personnel certification standard. The underwater inspector occupies a distinctive position in that scheme: the structure is examined by a diver or an ROV, while the inspector interprets what comes back. That split between who observes and who concludes defines the role, and it is where your study should begin.
Practise separating three decisions in every scenario. First, what was actually observed: the indication, its location, its appearance. Second, whether the observation quality supports a conclusion: visibility, cleaning, lighting, equipment condition. Third, what the report should state and what follow-up is justified. Rehearsing these as separate steps stops you blending a diver's description into your own unwarranted conclusion, which is the conceptual core of this credential rather than a memorisation problem.
Close Visual Inspection Underwater: A Procedure, Not a Look
Underwater visual inspection is a planned procedure: prepare the surface, remove marine growth, control lighting and camera position, then record findings together with the conditions that limited what could be seen.
Distinguish general visual and close visual as named levels of effort. A general visual sweeps a wider area for obvious damage, debris, corrosion, and anode condition; a close visual targets a defined location with cleaning and lighting to characterise a specific feature. Deciding which level a task requires, and stating that level in the report, is itself part of the inspector's judgment, and it makes a useful scenario drill because the two levels demand different preparation and produce different evidence.
The interpretation limit is the sea itself. Turbidity, biofouling, rust scale, and coating can hide crack openings or mimic them. Your report should name what was removed, how, and what remained obscuring the surface. A precise statement such as a weld toe confirmed clear of visible linear indications after cleaning carries more weight than a bare phrase like no defects found, because it discloses the evidence conditions alongside the finding.
Matching Each Underwater NDT Technique to What It Can Actually Confirm
Match the technique to the suspected defect: visual for surface condition, magnetic particle or ACFM for surface cracks, ultrasonics for wall thickness, flooded member detection for flooding, and CP surveys for protection state.
The scenario skill is matching method to failure mode. Fatigue cracking at welds calls for a crack-detection technique, not a thickness sweep; corrosion thinning calls for ultrasonics; the integrity of a sealed member is a flooded-member-detection question. When a scenario hands you a defect type, your first decision is which technique's physics actually address it, and your second is whether the reported conditions would let that technique work at all.
Also compare what each output lets you conclude. Visual finds and describes; magnetic particle confirms a surface indication as crack-like but gives no depth; ACFM adds sizing information; ultrasonics yields thickness numbers but does not discriminate crack types. When you write a phrase like crack detected with estimated length, know which method supports each word. Reports that outrun the capability of the method used are the reporting habit worth training out.
| Technique | Best suited to detect | Key limitation underwater |
|---|---|---|
| Close visual | Surface damage, corrosion, coating breakdown, obvious deformation | Line-of-sight only; cannot confirm or size subsurface flaws |
| Underwater magnetic particle | Surface and near-surface cracks in ferrous welds | Requires surface preparation and disciplined diver technique; locates but does not depth-size |
| ACFM | Sizing of surface-breaking cracks | Sensitive to geometry and operator technique; needs calibration control |
| Ultrasonic thickness | Wall loss from corrosion | Coupling and access govern reading quality |
| Flooded member detection | Water ingress into sealed structural members | Presence or absence only; gives no defect location |
| CP potential survey | Adequacy of cathodic protection | Validity depends on reference cell condition, contact, and position |
Scenario: A Linear Indication at a Weld Toe Reported by the Diver
Treat a diver-reported linear indication as an unconfirmed indication: verify surface condition, request a crack-detection method, and record location and limitations before anyone discusses repair scope.
Worked scenario: during a close visual at a nodal weld, the diver reports a sharp linear line along the toe that looks like a crack. The tempting write-up reads crack detected, repair required. That wording merges the diver's impression with your conclusion. The better sequence: confirm the surface was cleaned and by what method; request underwater magnetic particle inspection or ACFM at that location; log the indication with its position, orientation, and reported length as an observation awaiting confirmation.
Why the wording matters: report language drives mobilisation of repair spreads, which cost and carry risk far beyond a re-inspection. Labeling an unconfirmed visual impression as a confirmed crack can trigger unnecessary work; under-labeling a genuine fatigue crack can defer a safety-relevant repair. The defensible middle ground is a linear indication reported at the named location with confirmatory NDT requested, which is exactly as strong as the evidence in hand.
Scenario: One Cathodic Protection Reading That Looks Wrong
A single cathodic potential reading is an observation, not a verdict: check reference cell condition, electrical contact, position relative to anodes and coatings, and compare adjacent readings before reporting.
Worked scenario: a diver reports a potential reading on a jacket leg that seems low, and the draft report states CP system failure. The mistake is treating one number as a system verdict. A potential survey reading depends on the reference electrode type and its condition, the quality of electrical contact, the location relative to anodes and coating condition, and when the reading was taken relative to other work on the structure. Any of these can move the number.
The better decision: repeat the reading, note the electrode type used, verify contact, take adjacent readings on comparable members, and report the anomaly with its context, for example the value, the member, the cell type, and adjacent values, with a recommendation to review against the governing program criteria. Whether a value breaches the protection criterion is set by the applicable program and electrode type, not decided ad hoc underwater. Your deliverable is valid data plus a clearly flagged question.
Writing a Report in Three Layers: Observation, Limitation, Disposition
Write reports in three layers: the observation with a findable location reference, the conditions and methods that limit it, and the disposition or follow-up actually supported by the evidence collected.
Traceability is what makes an inspection record usable. Every indication needs a location a reviewer can find without you: structural member, elevation, clock position or weld number, plus date, personnel, vessel or ROV, and cleaning method. Keep fact and opinion in separate sentences. A sentence reporting a diver's measured indication at a weld toe is fact; a sentence saying it appears consistent with fatigue cracking and confirmatory NDT is requested is your stated judgment, and it should read like one.
Practical exercise with expected observations: take a written scenario, for example a platform node with reported coating breakdown and a suspect weld, and write the full report from scratch before consulting anything. Then score it against the five-point rubric below. A complete report earns five of five; each missing point tells you which layer to rework, and reworking the same layer twice is a signal to slow down and rebuild your template rather than rush another scenario.
- Rubric check 1: is every indication location findable by a stranger to the job (member, elevation, clock position or weld number)?
- Rubric check 2: are cleaning method, visibility, and lighting stated alongside the finding?
- Rubric check 3: is every conclusion tied to a named method and its capability?
- Rubric check 4: is the recommended follow-up proportionate to the evidence, neither inflated nor suppressed?
- Rubric check 5: does any sentence claim more than the technique that produced the data can support?
A Five-Pass Preparation Sequence With Self-Check Milestones
Prepare in five passes: map defect types to methods, drill close-visual procedure, rehearse paper scenarios with the three-decision split, build a report template, and finish by scoring yourself against the rubric.
You can treat the technical side of preparation as complete for your own tracking when, without notes, you can state why each technique in the table suits its defect type, write a limitation statement for a degraded-visibility close visual, separate a diver's observation from your own conclusion in two clean sentences, and say when a finding warrants escalation to confirmatory NDT. These are learning milestones for self-assessment, not predictions of any exam outcome.
Administrative items, including prerequisites, assessment arrangements, and renewal, are defined by TWI Certification Ltd and belong in one short check of cswip.com rather than in your technical study time; do not substitute forum hearsay for the issuer on those points. Keep the two streams separate: scenarios and reporting drills build the competence the certificate represents, while the issuer's pages settle the logistics.
- Pass 1: build a one-page map of defect type, detection method, and what that method can and cannot conclude.
- Pass 2: drill the close-visual sequence on inspection photographs, stating cleaning, lighting, and visibility limits before describing any feature.
- Pass 3: run two written scenarios daily using the observation, quality, and disposition split, writing the report before reading any suggested answer.
- Pass 4: draft a reusable report skeleton with fixed fields for location, conditions, method, limitation, and follow-up.
- Pass 5: score two full reports against the Section 6 rubric and rework any point below your target twice before moving on.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.