Coffee and the Agricultural Exports of Eastern Africa: Pre-Shipment Inspection, Stuffing Supervision and the Record That Travels With the Cargo
Ethiopia, Kenya, Uganda, Tanzania, Rwanda, Burundi and eastern DRC ship coffee, tea, cashew, sesame, pulses and horticulture into markets that test what arrives. This paper sets out what is examined before the container is sealed, how it is sampled, the calculations behind a quality or quantity dispute, and the documents the cargo cannot travel without.
Moses Agwaro and Dickens Ouma Marine and Cargo Surveyor, and Cargo Claims Handler, Observater Surveys and Services
How This Paper Should Be Read This is a working paper on coffee and the other agricultural exports of eastern Africa, and on the inspection work carried out before those cargoes leave. It covers what is examined at the warehouse and at stuffing, how lots are sampled, the analysis that follows, the documents the cargo travels on, and the calculations behind a quality or quantity dispute. Findings on an individual shipment are made file by file, on the evidence at that attendance, which is what allows them to be relied on when a buyer, a bank or an arbitrator examines them.
The practices described here are those commonly observed across many attendances. On a live shipment they are read together with the contract, the certificates and the analysis, and it is that combination that carries weight. Whether a recorded condition or result amounts to a rejection, an allowance, a shortage or no claim at all is decided against the sale contract and the terms it incorporates. Those terms are applied to a dated record made before the cargo left, and supplying that record is the work. Grade names, screen sizes and moisture bands quoted here are typical trade practice. The contract and the national authority govern in any particular case. Figures in the worked examples illustrate method. They are not results for any actual lot. We accept one appointment per matter, act for the interest that instructed us, and the findings reach that party first. The Trade, Origin by Origin Eastern Africa exports agricultural product that is bought on description, paid for against documents, and judged again on arrival by a laboratory in Hamburg, Trieste, Antwerp, Dubai or Yokohama. The distance between those two judgements is where inspection earns its place.
Coffee Origin, and what the buyer is buying
Ethiopia Arabica from Yirgacheffe, Sidama, Guji, Jimma, Limu, Harar and Nekemte, washed and natural, moving through the exchange and through direct relationships. Traceability to washing station level is part of the value, so lot identity through the dry mill and into the container matters as much as the grade.
Kenya Washed Arabica graded by screen size and shape into AA, AB, PB, C, E and TT, sold largely through the auction and by direct contract. Cup quality and consistency across a mark are what the buyer is paying for, and lot separation is the practical risk.
Uganda Robusta from the central and western regions and washed Arabica from Bugisu and the west Nile, with screen and defect based grades. Moisture management through a long inland leg is the recurring exposure.
Tanzania Washed Arabica from the north and the south, Robusta from the lake regions, sold through the auction and direct. Bag condition and lot identity through several handlings are the practical questions.
Rwanda and Burundi Fully washed Arabica of high value, sold on cup. The particular risk is the taint associated with insect damage, which is not visible in the green and appears only in the cup, and which travels with a lot unless it is caught at origin.
Eastern DRC Arabica from the Kivus and Robusta from the west, moving through Uganda, Rwanda and Tanzania as well as through Matadi. Multiple borders mean multiple handlings and multiple opportunities for lots to be mixed.
The other exports that travel the same way Commodity, and what is examined before shipment
Tea Made tea in paper sacks, sold on liquor and leaf. Taint is the dominant risk: tea takes up odour from anything stowed near it or from a container that carried something else, so container selection and odour testing are the core of the inspection.
Cashew, raw nuts Bought on outturn: nut count per kilogramme, kernel outturn ratio and moisture. Weight, moisture and count at stuffing decide the contract price adjustment.
Sesame Purity, oil content, free fatty acid, admixture and moisture, with colour mattering to some markets more than others. Stone, sand and other seed are what the purity check finds.
Pulses and pigeon peas Moisture, foreign matter, damaged and discoloured grain, weevil activity and size grading, with a fumigation and phytosanitary position that has to be right before shipment rather than after.
Maize, soya and groundnuts Moisture, foreign matter and, above everything, mycotoxins. Aflatoxin sampling is the most error prone task in agricultural inspection and is done to a plan or not at all.
Cotton Bale weight and moisture, contamination by polypropylene and other foreign fibre, grade and staple. Contamination is the claim that recurs and it is a stuffing floor issue as much as a ginnery issue.
Tobacco Bale condition, moisture and mould risk, with cases handled and stowed to a standard the buyer sets.
Horticulture and avocado Cold chain from the pack house, pulp temperature, dry matter and maturity where the contract sets a figure, pallet integrity and the temperature record of the reefer unit.
Spices and vanilla Moisture, volatile oil, foreign matter and, on vanilla, both moisture and the risk of theft in transit.
Gum arabic and gum resins Grade, foreign matter, moisture and packaging condition.
Macadamia and other nuts Kernel recovery, moisture, insect damage and rancidity risk on a long voyage.
A parcel is bought on a description written before it was picked, and judged on a laboratory result produced after it lands. Everything in between is documentation.
What a Pre-Shipment Inspection Consists Of Pre-shipment inspection on this trade is not one visit. It is a sequence of attendances at the points where something can still be established, and it ends at the moment the seal is applied.
From the washing station to the ship, and where a surveyor standsWashing stationor farm gateDry milland gradingExporter’swarehouseStuffingand sealingInland hauland borderPort, terminaland vesselmoisture atdeliverygrade, screen,defect countstack condition,bag marks, countcontainer check,lining, seal, tallyseal integrity,gate weightscondition onreceipt, loadingThe moment the seal goes on is the last point at which anything inside can be established without argumentThe chain from the washing station to the vessel, and where an attendance stands.
At the warehouse, before the container arrives The lot identified: mark, lot number, bag count, and the stack it is standing in, photographed with the marks legible. Stack condition: dunnage under the bags, gangways, clearance from walls, roof condition above the stack, and the general state of the store. Bag condition and packing: jute, sisal, polypropylene, or an inner liner where the contract calls for one, and whether the liner specified is the liner present. Bag weights checked at random and the average net established rather than accepted from the label. Moisture across the lot, measured at several positions in the stack rather than at the face. Sampling to a plan, with increments spread through the lot. Infestation: live insects, webbing, frass, and the fumigation history where one exists. Odour: on tea, coffee and spices, the store itself is part of the finding. The container, before a single bag goes in This is the part of the work most often skipped and most often decisive. A container that is damp, tainted, damaged or previously loaded with something aggressive will spoil a cargo that left the warehouse in good order, and once the doors are closed the question becomes a matter of argument between a shipper and a line.
A container prepared for coffee, seen from the doorDKraft lining on walls, floor and roof. Desiccant hung clear of the bags.Bags stowed off the doors, ventilators taped where the contract calls for it.Checked before a single bag goes inContainer number against the booking and the seal setLight test with the doors closed, from insideFloor dry, no odour, no previous cargo residueNo repairs that let water in, no oil or chemical stainDoors, gaskets and locking bars soundVentilators clear or closed, as the contract requiresLining fitted before stowage rather than afterRecorded as the bags go inBag count, marks, lot numbers and tare, with theseal number photographed on the closed doors.What is checked before stuffing, and what is recorded as the bags go in.
Stuffing and sealing Bag by bag tally against the lot, with the marks recorded and any bag rejected noted with the reason. Stowage pattern, dunnage, and space left at the doors so that bags do not press against them. Kraft or other lining fitted to walls, floor and roof where the contract or the trade calls for it, and photographed in place before it disappears behind the cargo. Desiccant hung rather than laid on the bags, with the type and quantity recorded. Ventilators open, closed or taped according to the contract, and the decision recorded. Final weight established, and the verified gross mass figure derived on a stated basis. Doors closed, seal applied by the surveyor or in the surveyor’s presence, and the seal number photographed on the closed doors with the container number in the same frame. From that moment, the only thing that can be said about the contents rests on what was recorded before the doors closed. That is the reason the whole exercise exists.
Sampling, Which Is the Whole Case Every quality result on this trade is a result about a sample. A parcel of six hundred bags is judged on a few hundred grammes, and the way those grammes were collected decides whether the number means anything at all.
Sampling a bagged lot, and what the sample becomesIncrements drawn by spear from bags across the stackGrosssampleDivided tocompositebuyersellerretained, sealedEvery split sealed, numbered and signed for at the point of divisionA lot sampled from the top layer only is a lot nobody has sampledFrom increments to sealed splits, with nothing taken only from the top layer.
How the sample is built Increments drawn from bags spread across the lot and across the height of the stack, using a spear that reaches the middle of the bag rather than the mouth. The number of bags sampled stated, and the selection method stated: every nth bag, a random table, or a stated proportion. Increments combined into a gross sample, divided down by riffle divider or by quartering, and the method recorded. Splits sealed at the point of division: one for the buyer, one for the seller, one retained, and a further split for a named laboratory where one is instructed. Each seal numbered, each split labelled with the lot, the date and the seal number, and each handover signed for. Lots kept separate throughout. Two lots combined into one composite produce a result that belongs to neither. Where sampling goes wrong Sampling only the accessible face of a stack, which produces a result about the face. Sampling at the mouth of the bag, where the coffee has been handled and where moisture behaves differently from the middle. Sampling after the lot has been prepared for the surveyor, which is why an unannounced attendance is worth more than a scheduled one where the contract allows it. Dividing by hand into a bag without a stated method, which is the step most often left out of a report and most often challenged. Leaving a split unsealed, or sealing it and then not recording the number. Sampling aflatoxin sensitive commodities with too few increments. The distribution is so uneven that a small sample is close to meaningless, and the plan has to say how many increments were taken and from how many bags. Quality Work on Green Coffee Green coffee is graded on physical characteristics and confirmed in the cup, and both parts of that sentence appear in a proper report.
What happens to a 350 gramme green sample350 gScreen sizes,sieved and weighedDefects counted,primary and secondaryMoisture, density,colour, odourRoast and cup,score and taint checkRecorded for each lotLot and mark, bag countScreen retention by sizeFull defect count with thecategories listed separatelyMoisture and water activityColour, odour and foreign matterCup notes, including any taintSample seal numbers and holdersNever averaged across lotsA parcel is as many results as ithas lots, reported separately.What is done with a green sample, and what is recorded for each lot.
Parameter, and what it decides
Moisture The parameter that governs both storability and weight, measured on a calibrated meter and, where disputed, by an oven reference method. Contracts commonly work in a band around nine to twelve and a half per cent.
Water activity A companion measurement to moisture that indicates how available that water is to mould, increasingly specified by buyers of high value lots.
Screen size Retention over a set of sieves, reported by size. Kenyan and Tanzanian grade names follow directly from it.
Defect count Primary and secondary defects counted from a sample of stated weight, with the categories listed rather than lumped into a single number: blacks, sours, pods, stones, insect damaged, broken, immature, withered, shells.
Density and bean colour Indications of altitude, processing and age, and part of what a buyer means by a good preparation.
Odour and foreign matter Recorded on the green before roasting, since a taint present in the green will be present in the cup.
Cup Roast, grind, brew and score, with taints described in the terms the trade uses. Potato taste, phenolic, fermented, mouldy, chemical and bag taste each point somewhere different.
Ochratoxin A Tested where the destination requires it, on a properly built sample, because the distribution in a lot is uneven.
The taint questions that matter most Bag taste from old or contaminated jute, phenolic and mouldy notes from wet or slow dried lots, fuel and chemical taint from a store or a container, and the potato taste defect associated with insect damage in the Great Lakes origins. Each is recorded as observed in the cup, with the physical evidence that accompanies it, and none of them is characterised beyond what was found.
Packing, Storage and the Physical Handling of a Lot Most of what a buyer complains about is decided by how the parcel was packed and where it stood, and both are cheap to get right and expensive to argue about.
Bags and liners Packing, and what it does and does not protect against
Jute or sisal bag The traditional pack. Breathes, which is an advantage while the parcel is stabilising and a disadvantage inside a sealed box. New jute carries its own odour, and old jute carries whatever it held before.
Polypropylene woven bag Cheaper and stronger, does not breathe in the same way, and on coffee raises the question of fibre contamination if it is cut rather than opened.
Multiwall paper sack Used for tea and some processed products, vulnerable to handling damage and to any water at all.
Grain and coffee liners inside the bag A barrier bag inside the jute, which holds the moisture the cargo left with and keeps out what it meets. Where the contract specifies one, the surveyor checks that the liner present is the liner specified and that it was closed properly rather than folded over.
Bulk liner in the container Used for some origins and commodities. Fitting, sealing and discharge arrangements are checked before filling, since a failed liner is a total handling problem at the destination.
Vacuum and barrier packs Used on high value specialty lots, where the pack is part of the product and damage to it is damage to the sale.
The store itself Roof condition above the stack, checked in daylight from inside, since that is where leaks are visible. Floor: sealed, dry, and with dunnage under every stack rather than under some of them. Walls: stacks clear of them, so air moves and so bags can be counted and probed. Separation between lots that is physical rather than notional, because a chalk line has never prevented a mix. Pest control: bait stations, records, and evidence of activity rather than an assurance about it. Neighbouring cargo, which on a taint sensitive commodity is part of the inspection. Fuel, fertiliser, hides, fish meal, soap and treated timber all have a history of reaching coffee and tea stored nearby. Temperature and humidity in the store, recorded rather than described, because they explain a moisture reading that has moved since the lot arrived. Stack building, and why it matters to the count A stack built in uniform blocks with a known number of bags per tier can be counted, probed and checked by anybody. A stack built to fill a corner cannot, and the count then rests on the warehouse ledger, which is exactly the document a financier or a buyer wanted verified independently.
Horticulture, and Cargo That Moves Under Temperature Kenya, Ethiopia, Tanzania and increasingly Uganda and Rwanda export fresh produce and flowers by air and by reefer container. The commodity is different and so is the inspection, but the principle is the same: the record has to exist before the cargo leaves.
What is examined on a temperature controlled export
Product condition at the pack house Maturity, defects, sizing and grading against the specification, with the sample drawn from the pallets actually being shipped rather than from the line.
Dry matter and maturity indices Where the contract or the destination sets a figure, particularly on avocado, measured and reported by lot.
Pulp temperature At the pack house, at loading and after loading, taken with a calibrated probe and recorded per pallet position rather than as a single figure.
Pre-cooling Whether it was carried out, for how long, and to what temperature. A container is not a cooling machine and cargo loaded warm stays warm in the middle of the pallet.
The reefer unit Set point, supply and return readings, fresh air exchange setting, and the unit’s own data record downloaded rather than described.
Stow and air flow Pallet pattern, gaps, the T-floor left clear, and cargo not stowed above the red line, because blocked air flow produces temperature failures that look like unit failures.
Packaging Carton strength and ventilation, corner posts, palletisation, strapping and corner protection, since a collapsed pallet in transit is both damage and a cooling failure.
Time The interval between pack house, plug in at the depot, plug in at the terminal and loading, recorded in hours, because that interval is where most of these claims are made.
On arrival questions, the download from the unit and the pulp temperatures at loading answer most of what is asked. Where neither exists, the discussion runs on assertion, and the party without documents carries it.
The Calculations Behind a Claim on This Trade
1. Bag count, average weight and the net of a lot Lot said to be 320 bags of 60 kg, nominal 19.2 t
Physical count: 318 bags
Twenty bags weighed at random: average net 59.6 kg
Net = 318 x 59.6
18,953 kg against a nominal 19,200 kg, a difference of 247 kg before the cargo has moved. Both the count and the average are established physically, because a lot accepted on its label is a lot nobody has checked.
2. Moisture and weight, the arithmetic behind a shrinkage claim Shipped weight 19,000 kg at 11.8 per cent moisture
Arrival moisture measured at 10.9 per cent
Dry matter is constant, so W2 = W1 x (100 – M1) / (100 – M2)
W2 = 19,000 x 88.2 / 89.1
18,808 kg. A loss of 192 kg, about one per cent, from moisture alone. A shortage reported without this calculation is a shortage reported against the wrong figure.
3. Container capacity, weight against volume Twenty foot container, payload commonly around 21 to 22 t depending on the box and the road limit
Internal volume about 33 m3, usable with bagged coffee about 28 m3
Coffee in 60 kg jute bags occupies roughly 0.085 m3 per bag as stowed
Volume limit = 28 / 0.085 = about 329 bags
Weight limit at 21.5 t = 21,500 / 60 = about 358 bags
The box fills before it weighs out, so the practical figure is around 320 bags and the loading plan is set by volume rather than by payload. A container planned on payload alone arrives either short loaded or with bags pressed against the doors.
4. Verified gross mass Cargo net established at stuffing: 19,080 kg
Dunnage, lining, desiccant and pallets weighed: 46 kg
Container tare from the plate: 2,250 kg
Gross = 19,080 + 46 + 2,250
21,376 kg, derived by the calculation method with each element weighed or taken from the plate, and the basis stated. A figure produced by weighing the loaded box on a calibrated bridge is reported instead where that is available, and the two are never mixed.
5. Sampling plan on a bagged lot Lot of 320 bags
Plan agreed at increments from the square root of the number of bags, rounded up
Square root of 320 = 17.9, so 18 bags sampled
Increment about 200 g from each, drawn at depth
A gross sample of about 3.6 kg, divided down to the composite the contract requires. The number of bags sampled and the selection method appear in the report, because a result without its plan cannot be compared with anybody else’s.
6. Cashew, kernel outturn and its price effect Sample of raw nuts: 1,000 g, nut count 190 per kg
Kernels recovered after cutting and grading: 232 g whole and half kernels of merchantable grade
Outturn expressed in pounds per 80 kg bag: 232 / 1000 x 80 = 18.56 kg per bag
Converted at 2.2046: about 40.9 lb per bag
Reported with the nut count, the moisture and the defective kernel proportion beside it. The contract scale is then applied by the parties to a figure established by cutting rather than estimated by eye.
7. Defect count converted to a grade position 350 g sample of washed Arabica
Full black 2, full sour 1, pod 3, large stone 1, insect damaged 6, broken 14, shell 5
Primary defect equivalents from the contract scale: black 2 x 1, sour 1 x 1, pod 3 x 1, stone 1 x 5 = 11
Secondary equivalents: insect damaged 6 x 0.2, broken 14 x 0.2, shell 5 x 0.2 = 5
16 defect equivalents in 350 g. The count, the categories and the scale used are all reported, so the grade position follows from arithmetic the buyer can repeat rather than from a conclusion.
8. Condensation risk in a sealed box Beans loaded at 24 C and 11.8 per cent moisture
Air in the box at loading 28 C, relative humidity 72 per cent, dew point about 22.6 C
Sea temperature on the northbound passage brings the roof to about 17 C
The roof falls about 5.6 C below the dew point of the air sealed inside, and water forms on it. That is the mechanism behind most top layer wetting, and it is addressed with lining, desiccant and loading temperature rather than after arrival.
Why a dry cargo arrives wet, and where the water comes fromLoaded warm at originBeans at 11.5 per cent moisture carry waterWarm air in the box carries moreCooled on passageWater forms on the cold roof and runs downTop bags wet, floor dry, walls streakedTop layer wet with a dry floor points inside the box. A wet floor with dry upper bags points somewhere else.Which is why the position of the wetting is recorded before a single bag is moved.Warm air sealed in at origin, a cold roof on passage, and where the water ends up.
9. Desiccant, sized rather than guessed Water carried by the cargo above the safe level: estimate from moisture and weight
19,000 kg at 11.8 per cent against a target of 11.0 per cent
Excess water = 19,000 x 0.008 = 152 kg, of which only a fraction becomes free vapour
Desiccant capacity: commonly around 30 to 40 per cent of its own weight
The calculation shows why a handful of sachets in a coffee container achieves very little, and why lining, loading temperature and ventilation practice do most of the work. The quantity actually used and its type are recorded either way.
10. Aflatoxin sampling, why the plan is everything Lot of 500 bags of groundnuts
Contamination typically concentrated in a very small proportion of the kernels
A plan taking 100 increments produces a far narrower result than one taking 10
Increments combined, ground and divided to a test portion
The report states the number of increments, the bags they came from, the grinding and division method and the laboratory. A result reported without them cannot be defended and cannot be compared with the arrival result.
11. Purity and admixture converted to tonnes Sesame lot of 26,000 kg, contract purity 99 per cent
Composite of 1,000 g separated by hand and by sieve
Foreign matter and other seed recovered: 17 g, so purity 98.3 per cent
Shortfall against contract = 0.7 per cent
On 26,000 kg = 182 kg
Reported as the purity found, the contract figure, and the tonnage the difference represents, with the separated material photographed and retained.
12. Cold chain, the interval that decides a reefer claim Pack house dispatch 14:20, pulp temperature 7.8 C
Plug in at the depot 19:05, ambient in transit about 28 C
Interval unplugged: 4 hours 45 minutes
Container pull down from the unit record: reaches set point 9 hours after plug in
Nearly fourteen hours between dispatch and set point, on a product loaded above its carriage temperature. The times are recorded to the minute because the argument on arrival is always about hours.
Where a green coffee moisture reading sits, and what follows from itunder 99 to 12.512.5 to 13.5above 13.5brittle, weightloss, cup thinthe band mostcontracts sit instorage risk rises,watch the stackmould and taintrisk, and weightRead on a calibrated meter and, where the number is disputed, by oven reference methodBands shown are typical trade practice. The contract governs.Where a moisture reading sits, and what usually follows from it.
How These Cargoes Are Lost and Damaged
1. Condensation in the container The dominant mechanism on bagged agricultural exports from this region. Warm humid air is sealed into a box at origin, the box crosses cooler water, and water forms on the roof and runs down onto the top layer of bags. The signature is wet upper bags with a dry floor, streaked walls and, frequently, mould on the bags rather than in the middle of the parcel.
2. Moisture that travelled with the cargo A lot dried too fast, dried unevenly, or stored before it had stabilised carries moisture that redistributes on passage. The picture on arrival is patchy rather than uniform, and the moisture record at origin is what separates it from condensation.
3. Mould and mycotoxin development Follows from the two above, given time and temperature. Ochratoxin on coffee and aflatoxin on nuts and grains are the parameters the destination tests, and the sampling plan at origin is what makes the origin result comparable with the arrival result.
4. Taint Tea, coffee, spices and cocoa take up odour. Fuel, chemicals, previous cargo residues, treated timber, a container that carried fish meal or hides, a store next to a fuel depot. The finding is in the cup or the nose and the cause is in the box or the building, so both are examined.
5. Infestation Insects present at origin, insects picked up in a store, and insects that arrive with the container. Live activity engages plant health authorities at both ends, and the fumigation record and the phytosanitary certificate are checked as part of the inspection rather than at the port.
6. Contamination and admixture Stones, soil, other seed, polypropylene fibre in cotton, metal from a mill. Purity is measured and reported as a percentage with the foreign matter separated and photographed rather than described.
7. Shortage and substitution Bags short at destination, bags of a different lot, or bags refilled with lower grade material. The controls are the tally at stuffing, the seal, the seal number recorded and photographed, and the gate weight. Where all four exist, this argument is short.
8. Losses on the inland leg The corridors from Ethiopia, Uganda, Rwanda, Burundi and eastern DRC to Mombasa, Dar es Salaam and Djibouti are long, and a container is out of anyone’s sight for days. Seal integrity at the port gate, checked against the number recorded at stuffing, is the single most useful record on that leg.
9. Cold chain failure on horticulture Pulp temperature at loading, set point, the reefer unit’s own record, and the interval between pack house and plug in. Fresh produce claims are decided on the temperature record, and where there is none the argument runs on assertion.
10. Documentary failure A phytosanitary certificate with the wrong description, a fumigation certificate that does not match the treatment, weights that disagree between documents, a certificate of origin that does not match the contract. None of these damages the cargo and all of them can stop it, delay it or cost the seller his payment under the credit.
The Documents the Cargo Travels On On this trade the documents are the cargo, as far as payment is concerned. A lot that is perfect in every physical respect will not be paid for against a credit if the paper does not match.
Document, and what it has to be right about
Commercial invoice and packing list Description, marks, bag count, net and gross weights, consistent with every other document.
Bill of lading Description and marks as the credit requires, and any remark that a shipper did not expect.
Certificate of origin Issued by the authority the contract names, matching the description.
Phytosanitary certificate Issued by the national plant protection organisation, with a description and treatment statement that matches the fumigation certificate.
Fumigation certificate Product, dosage, exposure period, date and place, and the name of the operator. A certificate that does not match what was actually done is a serious problem at the destination.
Weight and quality certificates By the party the contract names, on the sampling method the contract names.
ICO certificate of origin and export permits Where the coffee authority of the country requires them, in the form and sequence that authority sets.
Verified gross mass declaration With the method used stated.
Insurance certificate Where the seller carries it, matching the terms of sale.
Warehouse receipt or collateral documents Where the parcel is financed, matching the stock the surveyor actually counted.
A pre-shipment attendance is worth more when it checks the documents against the cargo, and the cargo against the documents, before the container leaves. Corrections at origin cost a morning. Corrections after arrival cost a great deal more.
Working in the Origin Countries
Country, and what an attendance there involves in practice
Ethiopia Coffee moving through the exchange and through direct contracts, with dry mills and exporter warehouses concentrated around Addis Ababa and the regional centres, and the corridor to Djibouti carrying the bulk of the export. Lot identity through the dry mill and onto the truck is the practical work, and our surveyors are stationed in the country rather than travelling to it.
Kenya Auction and direct sales, milling and warehousing around Nairobi and Thika, and stuffing at inland depots as well as at Mombasa. Marks and lot separation through the mill are the recurring point, and the horticultural export trade runs alongside on its own cold chain discipline.
Uganda Robusta and washed Arabica, processors and exporters concentrated around Kampala and the regional buying centres, with the long haul to Mombasa or Dar es Salaam. Moisture on arrival at the port is frequently different from moisture at the store, which is why both are recorded.
Tanzania Auction and direct sales, curing works in Moshi and the southern highlands, and stuffing at Dar es Salaam or up country. Cashew, sesame and pulses share the same warehouses and the same trucks, so cross contamination and lot mixing are live questions.
Rwanda Fully washed coffee of high value from a large number of small washing stations, consolidated for export. Traceability and taint screening carry more weight here than bulk quantity work.
Burundi Similar structure and similar risks, with the added complication of longer road distances to a port and more border crossings.
Eastern DRC Coffee and other agricultural product moving out through Uganda, Rwanda, Tanzania and, for the west, through Matadi. Multiple handlings, multiple borders, and documentation that has to satisfy more than one authority.
Malawi, Zambia and Zimbabwe Pulses, groundnuts, soya, tobacco and macadamia moving to Beira, Nacala, Dar es Salaam and Durban. Aflatoxin management and warehouse condition are the dominant technical questions.
Mozambique Cashew, sesame, pulses and pigeon peas, with Nacala, Beira and Maputo as the outlets and a Portuguese speaking documentary environment.
Attendance at any of these places is confirmed in writing before work begins, and a file that starts at an up country store is followed to the port and, where the instruction calls for it, to the destination.
What Each Interest Needs From the File
Exporters and producers The exposure is a rejection or a price allowance on a parcel that has already been paid for at origin. A pre-shipment record with sampling, moisture, grading and photographs, and a sealed retained split, is what allows an arrival claim to be answered instead of absorbed. Without a retained split, an arrival result cannot be tested against anything.
Importers, roasters and processors The concern is what arrives against what was bought, and whether it can go into a blend or a production run. An origin inspection to the buyer’s own specification, reported before shipment, gives the option of stopping a lot at origin rather than discovering it at the mill.
Traders Sitting between the two, with a contract on each side and different terms on each. What is needed is a report written against the contract that governs, sampled by the method that contract names, with splits that both counterparties can test.
Banks, trade financiers and collateral managers Where a parcel is financed against warehouse receipts or against documents, the questions are existence, quantity, quality and condition, verified independently and repeated while the stock is held. We provide stock counts against the ledger, condition records, and confirmation that what is on the schedule is in the shed.
Cargo underwriters Condensation, mould and shortage claims on this trade all turn on what was recorded before sealing. Moisture at stuffing, container condition, lining and desiccant, seal number and tally are the facts that settle them. Where the origin record exists, these files close quickly.
Shipping lines, forwarders and NVOCC The allegation is usually that the box was unfit or that the cargo was wet when it was loaded. A container inspection record made before stuffing, with the light test and the seal photograph, protects the line as often as it exposes one.
Coffee and agricultural authorities Where a national authority licenses exports, sets grades or issues certificates, the inspection record is prepared so that it can be shown to that authority without rewriting, and nothing in it characterises a party beyond what was found.
Buyers’ quality departments What they need is a report structured the way their own laboratory reports are structured: lot by lot, parameter by parameter, with the method named, so that origin and arrival results sit side by side on the same page.
Lawyers and arbitrators Contemporaneous, sampled to a stated plan, sealed with numbers recorded, and reasoned in the file. Trade arbitrations on this cargo are decided on sampling and documentation far more often than on the cargo itself.
What to send us The commodity, the lot size, the origin and where the cargo is standing. The contract terms that govern quality, weight, sampling and any arbitration provision. The buyer’s specification if it differs from the contract. The intended stuffing point and date, and the port of loading. What you need: quality, quantity, container and stuffing supervision, documents, or all of them. What comes back before anyone travels is the name of the attending surveyor, the time he will be there, and the fee in writing.
Reducing the Loss
At origin Dry to the contract band and let the lot stabilise before bagging rather than bagging to meet a truck. Store off the floor, clear of walls, with gangways, and keep lots separated by more than a chalk mark. Reject an unfit container at the gate. It is a five minute decision at origin and a six month argument afterwards. Line the box, hang the desiccant, and record both with photographs. Stuff in the cooler part of the day where the schedule allows it, because the air sealed in is the water that later falls on the cargo. Seal in the surveyor’s presence and photograph the number on the closed doors. Retain a sealed split of every lot shipped, and keep it somewhere it can be found in a year. On the inland leg Seal numbers checked and recorded at every handover, including at the border and at the port gate. Trucks tarpaulined and the container inspected for damage before it leaves the yard. Delays in the sun recorded, because time and temperature are what turn a moisture reading into a mould claim. At the port and on the vessel Container condition on receipt at the terminal recorded where the instruction extends that far. Stowage away from heat sources and, on breakbulk, away from cargo that taints. Ventilation practice for the voyage agreed and logged where the cargo is carried under deck in bulk or in bags rather than in containers. Disputes That Recur
The argument, and the record that decides it in practice
The coffee arrived wet Moisture at stuffing, the container inspection record, the position of the wetting on arrival, and the lining and desiccant photographs. Wet top bags over a dry floor point inside the box.
The quality is not what we bought The origin sample, sealed and retained, tested by the same method as the arrival sample. Where no split was retained, the comparison cannot be made at all.
The lot was mixed Marks, lot numbers and the tally at stuffing, with photographs of the marks as the bags went in.
Bags are short The physical count and the average bag weight at stuffing, and the gate weight. A count taken from the label settles nothing.
There is a taint The cup notes, the container inspection record including the odour check, and the store condition. Taint is a chain question and it is answered by looking at the whole chain.
Live insects were found on arrival The infestation check and the fumigation record at origin, with dates. Treatment applied and treatment certified are not always the same thing, and the inspection is where that is caught.
The mycotoxin result is too high The sampling plan at origin: number of increments, bags sampled, division method and laboratory. Two results from two different plans are not comparable and should not be treated as though they were.
The seal was different on arrival The seal number photographed on the closed doors at stuffing, and the seal checks at each handover. This is the cheapest control on the whole trade.
The Attendances This Trade Calls For
Attendance, and when it is the right one
Warehouse and lot inspection Condition, count, marks, moisture and sampling before the container is booked.
Container inspection and stuffing supervision Container fitness, lining, tally, weight, seal and photographs. The core of pre-shipment work on this trade.
Sampling and analysis To the contract method, with splits sealed and a named laboratory.
Quality inspection to a buyer’s specification Green grading and cup, reported lot by lot in the form the buyer’s own laboratory uses.
Stock and collateral inspection Existence, count and condition against a warehouse ledger, repeated at intervals for a financier.
Loading supervision at the port Where cargo moves breakbulk or where containers are received and loaded under supervision.
Destination attendance Devanning, condition on arrival and comparison against the origin record, arranged on the same instruction wherever the cargo lands.
Two of these are worth ordering when nothing has gone wrong: the container inspection before stuffing, and a retained sealed split of every lot. Neither can be recreated afterwards, and between them they answer most of what is later disputed.
Timing and the Records That Expire
What is running from the moment of shipment
Contractual claim periods Trade contracts set their own periods for quality and condition claims, frequently short and frequently measured from discharge or from the analysis certificate.
Arbitration provisions Where a trade association’s rules are incorporated, the timetable and the sampling requirements are set by those rules and are not negotiable afterwards.
Retained samples They age. A retained split has a useful life, and where a contract sets a retention period it is diarised.
The cargo Coffee is roasted and nuts are processed. Once a parcel has been used, the only record of its condition is what was made before it went.
Phytosanitary and fumigation validity Certificates have dates and the treatment they describe has a life. A delay at origin can put a shipment outside them.
Letter of credit dates Latest shipment and presentation dates run whatever is happening to the cargo, and a documentary discrepancy discovered late is expensive.
Questions We Are Asked Before Every Shipment
How many bags do you sample? Whatever the contract or the buyer’s plan specifies, and where neither says, a plan agreed in writing before we start. The number of bags, the selection method and the increment size all appear in the report.
Can you reject a container? We report it as unfit and record why, with photographs. The decision to load or not is the shipper’s, and where a container is loaded against our finding we record that too.
Who holds the retained sample? We do, sealed and numbered, and the seal number is in the report. Splits go to the parties the contract names, signed for.
Is desiccant enough to stop condensation? It is one control among several. Lining, loading temperature, the moisture the cargo carries and ventilation practice all matter, and the calculation in this paper shows why desiccant alone rarely carries the whole load.
We ship from up country. Do you attend there or at the port? Wherever the cargo is. The most useful attendance on this trade is at the stuffing point, because that is the last moment anything inside can be established.
Do you cup as well as grade? Yes, where instructed, and the cup notes are reported alongside the physical result rather than in place of it.
The buyer’s laboratory disagrees with our result. What now? The two sampling plans are compared first, because most of these differences are sampling differences. Then the retained splits are tested by an agreed method at an agreed laboratory, which is what the retained split exists for.
Can you carry out inspection to our own company specification? Yes. Send the specification and the report format you use, and the findings will be reported against it rather than against a general standard.
Do you handle the documents as well? We check the cargo against the documents and the documents against each other, and report any discrepancy before shipment. The issuing of certificates rests with the authorities and the parties who issue them.
What is the single most useful thing we can do? Inspect the container before stuffing and retain a sealed split of every lot. Between them they answer most of what is later argued about.
Terms Used in This Paper
Working definitions
Increment A single portion drawn at one point and one time, combined with others to form a gross sample.
Gross sample All increments combined before division.
Composite and split The sample obtained by dividing the gross sample, and the sealed portions issued from it.
Screen size Retention over sieves of a stated aperture, reported by size.
Primary and secondary defects Categories of defect in green coffee, each carrying a stated weight in the count used by the contract.
Water activity A measure of how available the water in a product is to microbial activity, distinct from total moisture.
Kernel outturn ratio The recoverable kernel from raw cashew nuts, expressed against a standard bag weight.
Admixture and purity Foreign matter and other seed present in a lot, expressed as a percentage.
Verified gross mass The gross mass of a packed container established either by weighing the packed box or by calculation from weighed components.
Taint An odour or flavour not belonging to the product, picked up from a container, a store or a neighbouring cargo.
Potato taste defect A taint found in some Great Lakes coffees, associated with insect damage, detectable in the cup rather than in the green.
Ochratoxin A and aflatoxin Mycotoxins tested on coffee and on nuts and grains respectively, with sampling error dominating the result.
Where We Attend, Country by Country The desks below cover the places the group attends. Origin work on this trade is carried out at washing stations, dry mills, warehouses and stuffing points inland as readily as at the port, and each desk carries the local detail and the way an instruction is taken.
Instructions for any of these places are taken on one line, and attendance is confirmed in writing before work begins. Where a file needs attendance beyond them, at a destination anywhere in the world, it is arranged on the same instruction and reported to the same standard.
The Lot Left the Store Sound, and the Laboratory in Europe Says Otherwise. Whichever Side of That You Are On, It Is Answered With the Record Made Before the Seal Went On Exporters carry the allowance they cannot answer because no split was retained. Importers and roasters carry the lot that cannot go into the blend. Traders carry a claim on one contract they cannot pass along the other. Underwriters carry a wet top layer nobody photographed at stuffing. Banks carry stock financed on a count somebody else made. Lines and forwarders carry the allegation that the box was unfit. Producers carry the price that moves against them for a defect found far away. Whichever seat you hold, send the commodity, the lot, the stuffing point and the date. What comes back is the name of the attending surveyor, the time he will be there, and the fee in writing. One interest per matter: when we take your appointment we act for your side only, and the findings reach you first.
