Person

The Four Weeks After Harvest

DA Dr. Alvin Boateng

Almost everything I work on happens in the month between a crop coming out of the ground and it reaching its first buyer. That is where a third of Ghana's smallholder harvest is lost, and it is where almost none of the money and attention goes. This is a walk through what we build, what we have learned in the field that contradicts what we published, and the storeroom of failed designs I now show students in their first week.

01 What we build

Equipment designed under one constraint: a fabricator in a district town must be able to build it in four days and repair it with what is already in the shop.

The mixed mode dryer, design nine

A solar dryer for maize, cowpea and cassava chips built from angle iron, plain sheet and polycarbonate, designed around two interventions a day rather than constant attendance.

Twelve square metres of collector feeding a timber and mesh cabinet. It brings maize from about 20 percent down to 13 in two good days, or three in the minor season. The important number is not the drying curve. It is two: the number of times a day somebody has to touch it. Design six was more efficient and assumed the trays would be turned every forty minutes, which means it assumed a person with nothing else to do. There is no such person in the households we work with. One fabricator builds a unit in four days. Materials run about 4,800 cedis. Sixty one units are in the field and the oldest is now five years old and has had one hinge replaced.

Hermetic bags, tested where they are actually used

Four years of trials showing that hermetic storage works well and fails almost entirely for reasons that have nothing to do with the bag.

The bags do what the manufacturers say. Grain at the correct moisture goes in, oxygen falls, weevils die, and six months later you have your crop. Every failure we have recorded came from one of three things. Grain put in too wet, which turns the bag into a very effective mould incubator. Bags opened repeatedly for small sales and never resealed properly. Or a bag placed directly on a concrete floor against a wall, where condensation collects. So our training spends twenty minutes on the bag and a day and a half on moisture, resealing and stacking. The technology was never the gap.
02 What the field taught us

Findings from trials in Ejura, Techiman and the Upper East that changed our designs, including several that contradicted our own earlier published work.

Who actually dries the grain

Drying is overwhelmingly managed by women who are simultaneously managing several other tasks, which makes attendance time the binding design constraint.

We spent our first two years optimising for thermal efficiency and cost per unit. Both matter. Neither predicted which units were still in use after two seasons. What predicted it was how many times a day the thing needed a person. Once we started counting that, our design decisions changed completely, and we accepted roughly fifteen percent longer drying times to get there. This was in front of us the entire time. We did not see it because we were measuring the machine and not the day it has to fit inside.

The aggregator sets the moisture standard, not the ministry

Farmers dry to the moisture their buyer will actually test for and pay for, which in most markets is not the published standard.

There is a standard. There is a moisture meter at the depot. Whether that meter is used, and whether a lower moisture lot earns a better price that day, varies by buyer and by week. Where the aggregator tests and pays a differential, adoption of drying equipment is straightforward and we barely have to argue. Where the aggregator pays a flat price by weight, we are asking a farmer to spend labour to reduce the weight of what they sell. No amount of good engineering fixes that. This is why I want to reach aggregators here, not only farmers. The incentive lives with them.

Full construction drawings, cut lists and material specifications for our dryer designs, released openly so any fabricator can build one without contacting us.

🔗Drawings and cut lists, publishedadompostharvest.org↗
03 The storeroom of failures

Ten designs that did not work, kept in a room behind the workshop with a card on each one. I now take students there in week one instead of week twelve.

Design four, the roof mounted collector

Thermally excellent and impossible to install, because the collector could not be lifted onto a roof by two people or serviced by one.

It performed better than anything we had built. It also weighed enough to need four people and a ladder we did not supply, and once it was up there nobody was going to climb up to clean it. We installed six. Within a year four had a film of dust on the glazing that cut output by roughly a third and none of them had been cleaned. The card on it in the storeroom reads: efficient, unmaintainable. Students always assume the failed ones performed badly. Almost none of them did.

Design eleven, and the goat

A ground level dryer with thin glazing that a goat stood on within three weeks of installation, twice, at two separate sites.

We specified 4mm polycarbonate to bring the cost down and we set the unit at waist height in an open compound. In a Ghanaian compound there are goats. Everyone in our lab knows there are goats. We simply did not think about them at the design stage, because a compound in a drawing has no animals in it. We are now at 6mm and a raised frame, and the card on design eleven has a photograph of the goat on it. It gets the biggest laugh of the year, and it is the lesson students actually remember.