Case 0002 — Composite case study
Claim family: Quantity, shortage and outturn disputes
Cargo: Wheat in bulk
Stage: Warehouse receipt and delivery reconciliation
Research reach: Durban and South African inland corridors
An apparent shortage first detected at the warehouse is not automatically a warehouse loss
The receiver’s first exception described an apparent discharge shortage in wheat in bulk during warehouse receipt and delivery reconciliation. The warehouse receipt was lower than the quantity expected from the shipment account, and the difference immediately affected stock planning, supplier settlement, insurance notification and the receiver’s ability to release wheat into its milling programme.
The first subtraction looked simple. The expected shipment quantity stood on one side; the quantity entered into the warehouse account stood on the other. The difference was called a shortage. Yet the records beneath those two figures did not describe one uninterrupted measurement of one identified parcel.
The source chronology in photographs, tallies and statement-of-facts entries used operational references created during vessel discharge. The working record built from draft survey worksheets used a different parcel description and covered a vessel-measurement interval that did not close at the same time as the warehouse account. Between those records lay terminal receipt, storage, possible line retention, dispatch by truck or rail, rejected or redirected loads, recoverable sweepings, damaged wheat and the final warehouse intake process.
The survey therefore began with a narrower and more useful question:
Can the expected quantity and the warehouse-received quantity be tied to the same wheat parcel, the same measurement basis and a complete chain of custody?
Until that question was answered, the apparent discharge shortage entry showed where a discrepancy was detected. It did not necessarily show where the difference originated, whether wheat had physically disappeared, or which party might ultimately bear responsibility.
This composite case study explains how the records were reorganised into a defensible warehouse receipt and delivery reconciliation. It combines recurring claim patterns for professional education and does not identify an actual vessel, terminal, warehouse, receiver or casualty.
Why Durban matters as an operational location
Durban is not merely the name of the discharge port. It is the beginning of a sequence of marine, terminal, storage and inland custody stages that may use different equipment, records and closing times.
Transnet identifies Durban facilities that include Maydon Wharf and an Agri-Bulk Terminal, and its published port-development material describes dry-bulk activity at Durban as including the import of wheat and other agricultural products. The exact facility, berth and handling arrangement still have to be confirmed for every real appointment; a general reference to “Durban discharge” is not enough. See Transnet Port Terminals and the Port of Durban development framework.
After landing, wheat may enter a hopper, conveyor, transfer tower, silo, flat store or other terminal-controlled system. It may later be dispatched by road or rail to a mill, inland silo or warehouse. Transnet describes the Durban–Gauteng rail corridor as carrying grain and connecting the port with inland terminals and private sidings. That makes the inland leg commercially important, but it also adds custody boundaries at which quantity, condition and parcel identity must be recorded. See the Transnet Freight Rail corridor overview.
For the reader handling a live claim, the practical consequence is direct. A quantity recorded at the vessel, a quantity received into a port silo, a quantity dispatched on trucks and a quantity accepted at an inland warehouse are not interchangeable merely because they all concern the same shipment. Each figure answers a different question:
- what the vessel is estimated to have discharged;
- what the terminal system recorded as received;
- what left terminal custody;
- what travelled under identified dispatch documents; and
- what the destination warehouse accepted on its own measurement basis.
The survey must therefore identify the actual berth or terminal, the discharge route, the temporary storage point, the mode of inland transport and the named destination. It must also establish the opening and closing time of every account. Without that operational map, a Durban cargo survey may compare accurate figures that simply cover different parts of the movement.
The claim setting: two records, two parcel references
The receiver’s claim file initially contained two important record groups.
The first was the source chronology in photographs, tallies and statement-of-facts entries. It showed the progression of discharge, interruptions, equipment changes, hatch sequences and some visible recovery activity. These records were contemporaneous, but the parcel labels followed the vocabulary used at the working face. Some entries referred to a bill-of-lading parcel, others to a hold, a discharge line, a silo destination or a daily total.
The second was the working record built from draft survey worksheets. It contained the observations and calculations used to estimate the change in the vessel’s displacement. It was capable of testing the overall mass discharged during the measured interval, but its parcel reference did not match the warehouse receipt file, and the available worksheets required review before their final result could safely be used.
Neither record group was discarded. The contradiction was itself evidence. It showed that the file had not yet established a single, continuous quantity account.
The commercial temptation was to select the figure that best supported the preferred claim position. That would have produced a quick number but a weak case. The survey instead preserved both sources, marked what each could prove, and identified the period and parcel that remained unverified.
The first operational decision: freeze the clocks and map custody
The urgent decision was not to recalculate the entire shipment immediately. It was to stop the next movement from erasing the evidence.
The surveyor requested the current terminal, dispatch and warehouse totals as they stood at a stated time. Later adjustments were to remain visible through corrected versions or system audit trails. The parties were asked not to overwrite the first warehouse exception, delete contradictory tally entries or combine recovered wheat with sound stock before it had been separately recorded.
A custody map was then built at the working face. It linked each handover to the document, instrument and responsible party that recorded it.
This map allowed operations to continue without forcing an early admission of liability. Each party could preserve its position while agreeing on the factual boundaries that still required proof.
What the attendance protected
The attendance protected four things that would otherwise have been lost as wheat continued to move.
First, it protected parcel identity. The working record built from draft survey worksheets was connected to the relevant vessel interval, while the warehouse receipts were connected to specific truckloads, rail wagons or delivery batches. Where the chain could not be completed, the gap was stated rather than hidden.
Second, it protected chronology. The source chronology in photographs, tallies and statement-of-facts entries was retained with its original timestamps and contradictions. A later summary did not replace the records from which the operational decisions had been made.
Third, it protected physical recovery evidence. Recoverable sweepings, wheat retained in the handling system, downgraded wheat, contaminated material and rejected deliveries were separately identified. This was essential to separating recoverable sweepings and damaged units from the alleged shortage. In a bulk-wheat file, “units” may refer to recorded lots, truckloads, wagons, scale transactions or other handling batches; it should not be allowed to imply bags or discrete packages that never existed.
Fourth, it protected settlement logic. A commercial compromise could later be recorded without deleting the technical trail. The survey file would still show how the net unexplained quantity had been reached and which amounts related to physical quantity, condition damage, mitigation, storage, transport or contractual adjustment.
Bulk wheat generates multiple quantities, each produced by a different method and for a different purpose. A reliable shortage analysis does not treat one figure as automatically superior. It tests whether the figures measure the same parcel and interval, then evaluates the uncertainty of each method.
The result is a ladder, not a single subtraction. Each rung must connect to the next before a net unexplained quantity can be attributed to a custody stage.
Parcel identity must be reconciled before quantity
The first technical defect in the composite file was not arithmetic. It was identity.
The warehouse account used a delivery or contract reference. The vessel records used hold and bill references. The terminal system used silo, line or stock codes. Dispatch records used truck, wagon or order numbers. These references had to be cross-matched through contemporaneous documents rather than assumed to describe the same wheat.
The reconciliation schedule therefore included an identity bridge:
Any quantity that could not pass through this bridge remained unverified. That did not mean the cargo was lost. It meant the evidence had not yet connected it.
Testing the working record built from draft survey worksheets
The draft survey was reviewed as a measurement process, not accepted merely because a final certificate contained a precise tonnage.
The surveyor checked the six draft readings where available, trim, list, hog or sag corrections, dock-water density, ballast and other tank soundings, fuel and freshwater changes, hydrostatic data, lightship or constant treatment, consumables, anchor or other unfactored weights, arithmetic and the timing of cargo and ballast operations.
Repeatability mattered. If drafts or density measurements had been repeated, the spread between readings was considered. If the vessel or water surface was moving, the record needed to show how an average had been obtained. If tanks changed between observation and calculation, the relevant time had to be established.
Industry loss-prevention guidance consistently warns that draft surveys are estimates rather than exact measurements. Gard notes that a properly performed dry-bulk draft survey is generally regarded as capable of accuracy around plus or minus 0.5 per cent, while West of England P&I discusses a typical variation of approximately 0.5 to 1.0 per cent and explains how drafts, density, ballast and unfactored weights can affect the result. These are useful technical orientations, not automatic allowances for every claim. See Gard’s dry-bulk shortage guidance and West P&I’s discussion of draft-survey inaccuracies.
The working record built from draft survey worksheets was therefore retained with its uncertainty and limitations. It could support the quantity estimated as discharged from the vessel during a defined interval. It could not, without the custody bridge, prove the quantity received at an inland warehouse or the location of a later variance.
Testing terminal receipt, storage and dispatch
The shore-side account required the same discipline.
Where a belt scale, flow meter, hopper scale or weighbridge generated the terminal quantity, the file sought the instrument identity, calibration or verification record, zero checks, operating logs, alarms, manual adjustments and the raw transaction export. A typed daily total was not enough if the underlying system record remained available.
The opening and closing state of the system also mattered. Wheat could remain in hoppers, transfer towers, conveyors, dust systems or other parts of the handling route when one account closed. That quantity might enter a later terminal total. Treating it as immediate physical loss would create a timing error.
Silo and stock records were checked for opening stock, receipts, transfers between bins, blending or commingling, dispatches, returns, rejected loads, cleaning, fumigation-related movements, sweepings, waste and closing stock. If wheat from more than one shipment entered the same storage system, parcel allocation required a stated method and could not be reconstructed by labels alone.
Dispatch reconciliation then matched each truck or wagon to a gross weight, tare, net weight, dispatch order, seal where applicable, release time and destination. Duplicate ticket numbers, manual tickets, cancelled loads, changed destinations and vehicles reweighed after an exception were investigated separately.
Testing the warehouse receipt account
The warehouse receipt was treated as an operational record, not as proof of origin.
The destination file identified the receiving instrument, calibration status, gross and tare procedure, vehicle or wagon identity, time of arrival, queue or holding period, intake bin, rejection or downgrade decision and the time at which the quantity entered the stock ledger.
Cut-off differences were a common source of apparent shortage. A terminal could close its dispatch account at midnight while the warehouse posted late arrivals on the following business day. A truck could have left terminal custody but remained in transit when the warehouse exception was calculated. A rail consignment could be split across receipt periods. Those quantities were “in transit” or “not yet posted,” not automatically missing.
The survey also checked whether the warehouse receipt used the same weight basis as dispatch. Different tare practices, estimated wagon tares, unverified vehicle modifications, scale rounding and weighbridge drift could accumulate across multiple movements. If moisture content was commercially relevant, the claim file had to distinguish actual mass measured at each point from any contractual adjustment based on moisture or quality.
Separating recoverable sweepings and damaged wheat from the alleged shortage
Bulk-cargo files often become distorted when all non-standard material is placed in one shortage line.
Recoverable sweepings are still physical wheat unless evidence shows that they were irretrievably lost, lawfully disposed of or rendered valueless. Wheat delivered wet, contaminated, mixed, heated, infested or otherwise damaged is also not automatically missing. It may support a condition or diminution-in-value claim, but its delivered quantity must not be claimed again as shortage.
The survey used the following classification:
This separation protected mitigation. It also prevented double recovery by ensuring that damaged wheat, recoveries and physical shortage were not valued twice.
Competing explanations tested
Analysis began with the evidence, not with the party first associated with the discrepancy.
The report distinguished supported, not excluded and contradicted explanations. Sequence alone was not treated as causation. The fact that a difference appeared after vessel discharge did not prove that it arose during discharge; similarly, detection at the warehouse did not make the warehouse responsible.
Ordinary trade tolerance is not the same as physical loss
The reported apparent discharge shortage was tested against ordinary trade tolerance rather than a physical loss. That phrase required careful handling.
There is no responsible basis for inserting a universal percentage merely because the cargo is bulk wheat. A relevant tolerance may come from the sale contract, charterparty, bill of lading terms, terminal agreement, warehouse contract, insurance policy, established trade usage or an agreed measurement standard. Instrument uncertainty and unavoidable handling variation may also be relevant, but neither should be converted into an automatic deduction without evidence.
The reconciliation therefore separated four concepts:
- Measurement uncertainty — the range within which a method may reasonably vary when properly applied.
- Contractual tolerance — a quantity or settlement mechanism expressly or validly incorporated into the governing contract.
- Known operational adjustment — a verified quantity such as recoverable sweepings, line contents, goods in transit or a corrected tare.
- Physical loss — wheat proved, on the available evidence, to have left the accounted custody chain without recovery.
A variance falling within a measurement range is not proof that no physical loss occurred. Equally, a variance larger than an indicative range is not proof of theft, negligence or carrier liability. It is a signal requiring closer investigation. The conclusion must remain proportional to the quality of the underlying evidence.
Rebuilding the warehouse receipt and delivery reconciliation
Once the parcel and time windows were aligned, the loss schedule was rebuilt around the affected quantity rather than the entire exposed shipment.
Every quantity retained a source reference. If a line depended on an estimate, the schedule stated the assumption and sensitivity. If the quantity could not be linked to the documented part of warehouse receipt and delivery reconciliation, it was excluded from the firm conclusion and shown as an unresolved item.
Valuation without losing the recovery trail
Quantum followed the technical findings; it did not replace them.
The unit value was taken from the legally relevant contract or policy basis, not automatically from a later spot price. Any freight, duty, handling or other component was included only if the governing basis permitted it. Salvage, retained use, sale proceeds, reconditioning value and other recoveries were credited to avoid double recovery.
Costs were classified separately:
- value of verified physical quantity loss;
- quality or condition diminution for wheat actually delivered;
- reasonable segregation, recovery, cleaning or reconditioning expense;
- additional transport or storage caused by rejection or diversion;
- survey and mitigation costs where recoverable; and
- contractual tolerance, franchise, deductible or other adjustment.
Amounts unrelated to the identified apparent discharge shortage were shown separately. A warehouse demurrage charge, production interruption or market-loss allegation might be commercially important, but it required its own causation and recoverability analysis.
The file retained the working record built from draft survey worksheets and the source chronology in photographs, tallies and statement-of-facts entries. A commercial settlement could then resolve the account without erasing the evidence needed for contribution or recovery against another party.
Documents that should remain together
Original exports and superseded versions should remain available. A clean spreadsheet created for settlement is useful, but it should be traceable back to the evidence produced at each custody point.
Legal and contractual orientation in South Africa
The technical reconciliation does not decide legal responsibility.
South Africa’s Carriage of Goods by Sea Act 1 of 1986 is part of the legal framework that may govern the sea-carriage stage. The actual application of that Act and the incorporated cargo regime depends on the contract, voyage and legal position of the parties. The warehouse and inland transport stages may be governed by different contracts, standard terms, statutes, insurance provisions and forums.
The real file therefore needs to identify:
- the claimant’s title and right to sue;
- the bill of lading, charterparty and incorporated terms;
- the terminal, storage, road, rail and warehouse contracts;
- where delivery under each contract legally occurred;
- notice provisions, survey clauses and time bars;
- contractual quantity and tolerance wording;
- exclusions, limits, deductibles and jurisdiction clauses; and
- the governing law and agreed dispute forum.
The survey report should provide the facts needed for that analysis: what quantity was supported, where each custody handover occurred, what exceptions were observed, what remains unexplained and how quantum was calculated. Legal advisers can then address responsibility without asking technical evidence to answer a legal question it was never designed to decide.
Lessons for cargo owners, receivers and millers
Do not wait for the warehouse month-end stock account before investigating a material variance. Preserve terminal dispatch and destination receipt records while vehicles, wagons, system logs and witnesses remain available.
Require the claim schedule to separate wheat not received from wheat received but damaged or downgraded. Confirm that goods still in transit, rejected loads, diversions and recoveries have been credited before a physical shortage is alleged.
When a commercial tolerance is invoked, ask where it comes from and what it applies to. A contractual tolerance, instrument uncertainty and policy deductible are different adjustments and should not be combined casually.
Lessons for carriers, vessel interests and P&I correspondents
Preserve the complete draft survey working record and not merely the final certificate. Record the time of every observation and any cargo, ballast, bunker or freshwater operation occurring around it.
If vessel and shore figures differ, issue a reasoned reservation while the evidence can still be checked. But avoid treating the vessel’s estimate as proof of the inland warehouse receipt. Support the transition from vessel to terminal with the statement of facts, discharge sequence and jointly observed closing condition.
Lessons for terminals, silo operators and stevedores
Define where terminal receipt occurs and which instrument controls the commercial account. Preserve raw transaction data, audit logs, corrections and the opening and closing condition of the handling system.
Connect sweepings, dust, waste, retained line contents, transfers and reworked wheat to the relevant parcel. If parcels are commingled, state the allocation method and its limitations before an apparent shortage is assigned to a particular shipment.
Lessons for road, rail and warehouse custodians
Every dispatched movement should have a unique identity that can be matched at destination. Gross, tare and net weight records should remain linked to the vehicle or wagon, seal where used, dispatch order and receiving note.
Warehouses should distinguish rejected, diverted, pending and accepted deliveries. The first stock exception is valuable evidence of detection, but it should not be described as proof that loss occurred inside the warehouse.
Lessons for insurers, brokers, adjusters and recovery teams
Set early reserves with an uncertainty explanation. State which quantities are verified, in transit, recovered, damaged, disputed or unsupported. Update the reserve as the custody bridge closes.
Preserve settlement and recovery logic separately. A pragmatic payment may be commercially justified even when technical causation remains incomplete, but the settlement document should not silently convert an unresolved variance into an admission about origin or responsibility.
Practical response checklist for an apparent wheat shortage at Durban
During discharge
- Identify the exact berth, terminal, discharge route and destination storage point.
- Obtain the cargo plan, bill references and hold sequence.
- Preserve arrival and completion draft survey worksheets and independent checks.
- Record stoppages, line changes, spills, recoveries and sweepings in the statement of facts.
- Confirm which shore instrument records terminal receipt.
Before terminal dispatch closes
- Freeze the terminal stock and dispatch totals at a stated time.
- Export raw scale, flow-meter, silo and weighbridge transactions.
- Reconcile opening stock, receipts, transfers, dispatches, recoveries and closing stock.
- Identify goods in hoppers, conveyors, transfer towers or other system hold-up.
- Match every truck or wagon to an order, net weight and destination.
At warehouse receipt
- Preserve destination gross, tare and net records and instrument identity.
- Record accepted, rejected, diverted and pending loads separately.
- Align warehouse posting time with terminal dispatch cut-off.
- Segregate and weigh damaged wheat and recoverable sweepings.
- Invite relevant parties to a joint reconciliation without prejudicing rights.
Before claim submission
- Build the parcel-reference bridge across all custody stages.
- Separate measurement uncertainty, contractual tolerance and physical loss.
- Credit sound receipt, damaged delivery, recovery and known goods in transit.
- State the net unexplained quantity and the evidence supporting its location.
- Present quantity, condition loss, mitigation cost and deductibles separately.
- Review title, governing contracts, notice, time bars, law and forum.
Questions frequently asked after a bulk-wheat shortage
Does a lower warehouse receipt prove that the vessel discharged short?
No. It proves that the warehouse accepted or recorded less than the selected comparison figure at a stated time. The terminal receipt, storage and inland dispatch stages must be reconciled before the difference can be located.
Can a draft survey replace terminal and warehouse records?
No. A draft survey estimates the change in the vessel’s displacement. It is valuable evidence of the vessel-side quantity but does not measure truck, rail or warehouse receipt. The methods should be compared only after parcel and timing differences are resolved.
Should recoverable sweepings be included in the shortage?
Verified recoverable wheat should be credited against the physical quantity account. Reasonable recovery, cleaning or diminution costs may be assessed separately where the governing basis allows them.
Is damaged wheat also a shortage?
Not merely because it is damaged. Wheat physically delivered must normally be credited as delivered quantity. Any loss of value, reconditioning cost or disposal consequence belongs in a separate condition and valuation schedule.
Is there a standard percentage tolerance for bulk wheat?
No single percentage should be assumed for every file. The relevant contract, measurement method, policy wording and proven trade usage must be identified. Indicative technical ranges are not automatic contractual allowances.
What if the vessel, terminal and warehouse all have different figures?
Build a quantity ladder and reconcile each adjacent custody stage. Record the method, parcel, time window and uncertainty behind every figure. The location of the first unsupported difference is more useful than choosing whichever total appears most favourable.
Who is responsible for an apparent shortage detected inland?
The detection point does not decide responsibility. First establish whether a real net loss exists and where the custody chain stops reconciling. The governing contracts, law, cargo regime and evidence can then be applied to the supported facts.
Closing insight
The decisive step in this composite wheat-in-bulk case was not selecting a preferred tonnage. It was connecting the vessel, terminal, dispatch and warehouse records to the same parcel and closing time.
The first apparent discharge shortage entry remained in the file as evidence of detection. The source chronology in photographs, tallies and statement-of-facts entries was preserved alongside the working record built from draft survey worksheets. Contradictory entries were not deleted. Recoverable sweepings and damaged wheat were separated from alleged missing quantity. Ordinary trade tolerance was analysed separately from evidence of physical loss.
That approach produced a settlement-ready account without forming a premature view on liability. It allowed cargo interests, carriers, terminals, inland custodians, insurers and recovery teams to see what was received, what was recovered, what was damaged, what remained in transit, what depended on measurement and what quantity—if any—remained genuinely unexplained.
For an independent Durban cargo survey, wheat-in-bulk discharge survey, tally and shortage investigation, or warehouse receipt and delivery reconciliation across South African inland corridors, Observater Surveys and Services Group can coordinate evidence preservation, custody mapping, quantity reconciliation, loss control and settlement-ready reporting. Visit observater.com or contact ops@observater.com.
Editorial and legal note
This is a composite research case prepared for professional education. Facts and issues have been combined and anonymised; it must not be represented as the report of a named casualty. External links provide operational, technical or legal orientation and should be checked against the requirements of the live file. The legal section is general information, not legal advice. Every actual claim requires review of its own evidence, contracts, policy wording, governing law, cargo regime, limitation provisions and forum.
