The documents described a coherent Houston-to-Tianjin cargo movement. We tested that story against the physical vessel universe, persistent IMO identity and voyage timing before treating it as established.
Submitted commercial documents remain redacted. The vessel-screening data below uses the actual public vessel names, IMO numbers, capacity and timing used in the verification, so the physical-voyage logic can be evaluated directly.
Each item is tied to a specific claim. Status describes what the evidence establishes — not how official the document looks.
| Ref | Evidence / source | Source class | What it tests | Finding | Status |
|---|---|---|---|---|---|
| E01 | Submitted Bill of Lading | Submitted Evidence | Route · quantity · loading claim | Houston → Tianjin and 44,000 MT were stated in the document. | SUPPORTED |
| E02 | Submitted Voyage Report | Submitted Evidence | Flag · departure · ETA | The report claimed a Turkey flag and a Houston departure around 5 June. | CLAIM |
| E03 | IMO-linked vessel identity | Independent Operational | Persistent physical identity | IMO 9320752 anchored the vessel across name and flag changes. | VERIFIED |
| E04 | Houston LPG / VLGC movement screen | Independent Operational | Capacity · timing · destination | The closest physical match departed before the submitted voyage narrative and did not support the Turkey-flag identity. | CONTRADICTED |
| E05 | Carrier / port / terminal confirmation | Required Evidence | Current voyage connection | No independently verifiable carrier / port / terminal confirmation was established from the reviewed package. | REQUIRES EVIDENCE |
| E06 | Current seller title / authority chain | Required Evidence | Commercial control of the cargo | The reviewed vessel evidence did not itself establish that the presenting party controlled the cargo. | NOT ESTABLISHED |
The submitted B/L described liquefied propane loaded at Port of Houston for Tianjin and stated a specific on-board date.
The voyage report stated “Flag State: Turkey”, with Houston departure around 5 June and revised Tianjin ETA around 3 July.
Using the IMO rather than the vessel name allowed us to track the same physical ship across name and flag changes and compare it with the transaction documents.
The submitted voyage material identified the vessel as Turkish-flagged. Independent vessel records tied to IMO 9320752 instead showed the same ship as BW ELM under the India flag around the relevant voyage period, later changing name to SHENG HAI LONG and flag to Marshall Islands.
Why this matters: vessel names and flags can change; the IMO number is the persistent identifier. Once the IMO was fixed, the transaction document's Turkey flag claim no longer matched the independent vessel identity history.
We did not search for a vessel that “looked right.” We built a broader candidate set and eliminated vessels by capacity, timing and destination.
| Candidate / IMO | Houston timing | Capacity / class | Post-Houston route | Assessment |
|---|---|---|---|---|
| BW ELM / 9320752 | 29 May → 3 Jun night / 4 Jun UTC | DWT 58,423 · 82,291 m³ | Houston → Tianjin | Closest physical match |
| BW TYR / 9346122 | 4 Jun → 5 Jun | 82,303 m³ class | Later Mundra | Timing close; destination mismatch |
| CONSTELLATION / 9734680 | 4 Jun → 8 Jun | 84,000 m³ | Houston → Dahej | Still in Houston after claimed departure |
| BW FREYJA / 9733325 | 5 Jun → 8 Jun | DWT 54,053 · 84,143 m³ | Houston → Zhanjiang | China-bound, but wrong timing / port |
| CLIPPER GALLEON / 1026829 | 7 Jun → 14 Jun | DWT 56,171 · 88,056 m³ | Later China-bound | Arrived Houston after claimed loading date |
We screened the relevant Houston window rather than assuming the vessel name provided in the POP was correct.
Smaller LPG carriers, late arrivals, early departures and incompatible destinations were excluded before deeper identity work.
The closest physical match was then tested through persistent IMO identity and independent vessel records, exposing the flag inconsistency.
Method note: public AIS and vessel databases can be delayed or retain different snapshots. We therefore do not treat any single database as conclusive; the value comes from reconciling multiple independent sources against the commercial documents and timeline.