Harmattan and the Dry Season: Dust, Sub-5% Humidity and Wide Diurnal Swings

The harmattan changes three storage variables at once, and they do not carry equal weight. From roughly the end of November to the middle of March, the dry north-easterly flow off the Sahara raises the airborne particulate load, drops relative humidity to values most storage guidance never contemplates, and widens the gap between the coldest and the warmest hour of the day. The humidity is the one that alarms people and the one least likely to reach your material. The particulate and the diurnal amplitude are the ones that quietly degrade the equipment and the records that keep material inside its stated condition — which, on an audit trail, amounts to the same failure.

What follows is written for a small storage footprint: one refrigerator, a logger or two, a bench, and mains power that cannot be relied on, in a country where the harmattan is an annual event rather than an anomaly.

What the harmattan actually changes

The harmattan is a ground-level stream of dry continental air moving south and west out of the Sahara and the Chad Basin during the boreal winter. It carries mineral dust with it and simultaneously reduces visibility, relative humidity and air temperature along its track ^[1]. Reported relative humidity commonly sits below a quarter of saturation, with minima under five per cent recorded in the interior, and ambient temperature within a single twenty-four hour period can move from single digits before dawn to the high twenties or low thirties by mid-afternoon.

The dust is not uniform across the region. Deposition work in Ghana that trapped harmattan dust across the country's agroecological zones found northern sites retaining several times the mass of southern sites, alongside a clear latitudinal sorting of particle size: a median diameter slightly above 15 micrometres in the north against slightly above 5 micrometres in the south ^[2]. Coarse material settles out on the way; the fine fraction travels. That distinction matters more than the headline dust figures, because an enclosure or a filter that copes with grit does not necessarily cope with what reaches the coast.

• Particulate: elevated for roughly four months, finer the further south you are, and accumulating preferentially on horizontal and electrostatically charged surfaces rather than distributing evenly ^[2].
• Humidity: relative humidity in the teens or lower for days at a stretch, punctuated by abrupt recovery whenever the flow breaks down ^[1].
• Diurnal amplitude: twenty degrees or more between the pre-dawn minimum and the afternoon maximum, which loads a refrigerator's control loop far more than it loads the contents.

Dew point, not relative humidity, decides condensation

Relative humidity is a ratio, not a quantity of water. What determines whether a cold surface wets is the dew point, and during the harmattan the dew point is startlingly low. The table below applies the Magnus approximation to ambient conditions typical of the season. It is arithmetic rather than measurement, so treat it as a sanity check against your own hygrometer, not as a substitute for one.

The operational consequence inverts the usual advice. In the humid season a coastal bench at 28 °C and 80% relative humidity has a dew point near 24 °C, so anything lifted out of a 2-8 °C refrigerator begins condensing on contact with the room and has to be equilibrated inside a sealed bag before the closure is broken — the standing containment step in handling procedure for lyophilised material. In the depths of the harmattan, with a dew point at or below freezing, the same item does not wet at all. The season removes a familiar hazard and substitutes two less familiar ones: electrostatic charge, and the slow drying of every elastomer and adhesive in the room.

Do not generalise that relief across the whole season. The flow establishes and breaks repeatedly, and a coastal site can move from a dew point near zero to the low twenties within a day when moist monsoon air pushes back inland. The defensible rule is to keep the equilibration step in the written procedure year-round and to measure the room rather than assume it.

Dust is an equipment failure before it is a contamination problem

The most likely harmattan casualty in a small setup is the refrigerator's ability to reject heat. A compression appliance dumps heat through a condenser, and a condenser blanketed in fine dust runs hotter, runs longer, and sits closer to its ceiling. WHO performance specifications for cold-chain refrigerators require an appliance to hold its internal band while operating in a test chamber at +43 °C ambient, and current verification protocols require every model to pass in that hot zone rather than in a milder one ^[5]. That is generous headroom on paper. A fouled condenser and a wall-hugging installation spend it before the weather does.

Enclosure ratings are worth reading literally rather than as marketing. Under the IP code in IEC 60529, the first characteristic numeral 5 means dust-protected — ingress is not prevented, only limited to a quantity that does not interfere with satisfactory operation or impair safety — while 6 means dust-tight ^[6]. The qualifying test circulates talc with 95% of its mass below 75 micrometres. Harmattan dust arriving in the southern half of the region has a median around 5 micrometres ^[2], sitting well inside the fine tail of that distribution, so a dust-protected rating tells you the appliance will keep working, not that its interior will stay clean.

• Move condenser and grille cleaning from a quarterly task to a monthly one for the duration of the season, and log each clean with a date and an initial.
• Respect the manufacturer's stated clearance behind and above the appliance; where none is stated, treat 100 mm as a floor and put nothing on the lid.
• Site the unit away from open louvres, doorways and unpaved approaches. Filtering or sealing the room is more tractable than filtering the appliance.
• Clean generator intakes and inverter ventilation filters on the same monthly cadence, because they foul from the same air.
• Do not blow a condenser clean with compressed air indoors; that relocates the dust onto the logger, the bench and the ventilation grille.

Sub-10% humidity at the bench: static, seals and labels

The USP chapter on weighing with an analytical balance is explicit that fine powders acquire static charge, that charged particles scatter, and that the magnitude of the charge depends on the relative humidity of the room, with operator clothing a documented source of large weighing errors. It recommends siting a balance in a temperature- and humidity-controlled space, and lists antistatic weigh vessels, screens and ionisers among the remedies ^[7]. Harmattan conditions are close to the worst case that chapter contemplates — the conditions under which static charge and ambient moisture make a weighed figure unreliable — and a balance that behaved perfectly in October will drift and scatter in January without anything having been done to it.

The same dryness reaches things that never touch the balance. Elastomer door gaskets stiffen and lose contact; self-adhesive labels curl at the edge and lift; direct-thermal paper, the receipt stock in many small printers, is the least durable record in the building and the first to fade past reading. None of this is dramatic on any single day, which is precisely why it survives four months unnoticed and then surfaces during a stock reconciliation.

• Ground the bench and the operator, and verify the balance against a calibrated check weight at the start of each session rather than only at calibration intervals ^[7].
• Treat electrostatic discharge as a live risk to the data logger, the laptop and the barcode scanner, not only to the material being handled.
• Print labels on synthetic stock with a laser or thermal-transfer printer, keep the authoritative record digital, and treat any thermal-paper slip as disposable.
• Check refrigerator door seals monthly during the season with a paper-strip drag test, and replace a gasket that has gone hard rather than trying to adjust it.

The diurnal swing is what your mapping has to catch

A storage area mapped in the rainy season does not describe the same room in January. WHO's technical supplement on temperature mapping asks for a minimum of seven consecutive days for warehouse and ambient storage areas, deliberately spanning five working days and two weekend days, and between 24 and 72 hours for cold rooms and freezer rooms, with sensors arranged on a three-dimensional grid at high, medium and low levels ^[4]. Seven days is a minimum, not a season. The honest reading for a room with a twenty-degree diurnal amplitude is that it needs at least one mapping exercise conducted inside the dry season, or a continuously logged channel that runs straight through it.

For judging what a swing actually cost, the reference is the USP chapter on mean kinetic temperature, which sets MKT out as a tool for evaluating temperature excursions during storage and transportation and is equally clear that it cannot be used to normalise storage that was simply uncontrolled ^[8]. Controlled room temperature is defined as 20-25 °C with excursions permitted between 15 and 30 °C. An unairconditioned room in the harmattan will spend its afternoons outside that band no matter how reassuring the daily mean looks.

It is also worth knowing where a stated storage condition comes from. ICH Q1A(R2) places long-term stability storage at 25 °C with 60% relative humidity or, alternatively, 30 °C with 65% relative humidity, and uses 30 °C/65% RH as the intermediate condition ^[9]. Harmattan ambient humidity sits far below either figure, so the humidity axis of that testing is not your stress case. The temperature axis is — and because the chemical routes by which a peptide degrades accelerate disproportionately as temperature rises, the afternoon peak rather than the daily average is the number worth defending.

Power: the dry season is when the backup is weakest

Solar-backed refrigeration and the harmattan are badly matched, and the mismatch has been quantified. A 2023 analysis of decentralised solar across West Africa found soiling losses that can exceed 50% at the most affected locations during the harmattan, with year-to-year variation between 26% and 53% at Niamey, and projected that cleaning costs could approach one billion US dollars annually across sub-Saharan Africa by 2030 if decentralised solar scales as planned ^[3]. An array sized on clean-panel output is therefore short of input during exactly the months when ambient heat is pushing the compressor hardest — a compounding failure, not two independent ones.

• Measure array output weekly through the season instead of trusting the design figure; a soiled array degrades quietly and continuously ^[3].
• Set the cleaning interval from that measurement rather than from habit. In high-deposition zones a weekly clean is not excessive.
• Prefer an appliance verified against the hot-zone ambient condition, and check that its verification covers intermittent or absent mains rather than continuous supply ^[5].
• Keep an independent temperature record that depends neither on the appliance's own display nor on grid power.

What to record while the season lasts

The dry season is a four-month experiment run on your storage arrangements whether you observe it or not. A short, boring record is what converts it into evidence you can act on next November.

One limitation, stated plainly: none of this rests on a study of research compounds under harmattan conditions, because no such study exists in the published literature. What exists is dust meteorology ^[1][2], cold-chain equipment verification ^[5], enclosure testing ^[6], weighing metrology ^[7] and photovoltaic soiling data ^[3] — each solid within its own field, and each collected for another purpose. The transfer is reasonable, since a condenser does not care what is inside the cabinet and a balance does not care what is on the pan, but it is a transfer and should be labelled as one. The only evidence that describes your storage is the record you keep through the season, which is the argument for keeping one.