Grain bin hot spots

Why does one corner of a bin of dry, sound grain heat up and spoil while the rest of it keeps for a year?

A bin is filled with grain that tested dry and sound at every sample point. Months later most of it is still perfect, and one region — often near the top centre, sometimes a wall or corner — is warm to the hand, caked and mouldy. Nothing was added or taken away.

Two things need explaining, and they are separate. Why does a sealed mass of dry grain heat at all, and why in one place rather than everywhere?

Take the heat first, because this collection has already done half of it. The walk-through on compost heat establishes the source: living things oxidise carbon, and the energy not captured as chemical work leaves as heat. Grain is alive — seeds respire — as are the fungi and insects riding on it. But the fixed point of difference matters: a compost heap is built to reach that state, made big, wet and rich in easy carbon, the operator wanting a core at fifty-something degrees. A bin is the identical physics fought rather than farmed, its defence being to keep the biology too slow for heat to accumulate.

What sets that rate? Not the total water, but how available it is — moisture in equilibrium with the air between the kernels. Below roughly sixty-five percent relative humidity in that air, even the most drought-tolerant storage moulds cannot grow; that figure is recalled and approximate, and the corresponding grain moisture content differs by crop, temperature and storage life and is set by trading standards as much as by biology, so it is not quotable as a constant.

Now the second half. Why should a uniformly dry bin develop a wet patch? As autumn turns to winter, grain near the wall and roof cools while the core, insulated by tens of tonnes of a poor conductor, stays at harvest temperature for months. Bulk grain conducts heat at roughly a tenth to two-tenths of a watt per metre per kelvin — recalled, but firmly in insulator territory. The bin holds a warm core and a cold shell.

Warm air between the kernels at the core is less dense, so it rises. Meeting colder grain near the top, its capacity to hold vapour falls, its relative humidity climbs, and the grain there takes up the moisture; the drier, cooler air sinks down the walls and warms again at the base. That cell runs for months, pumping water from warm core to cold surface; in summer the gradient reverses and the wet zone forms low and central instead. The bin redistributes its own water, and a uniformly safe mass develops a region that is not.

That explains a wet patch. Now the runaway, worth deriving rather than asserting. Take a small warm region and ask two questions. How fast does it make heat? Biological rates rise steeply with temperature — roughly a doubling per ten degrees is the usual rule of thumb — so generation climbs like a compounding curve. How fast does it lose heat? By conduction into the surrounding grain, in proportion to the temperature difference, through a material that barely conducts. Loss is a straight line, generation a curve, and the two cross.

Below the crossing point, a warm patch loses more than it makes and subsides. Above it, the patch makes more than it loses, gets hotter, and makes more still — nothing further is needed for a runaway. Two extra loops worsen it: respiration produces water as well as carbon dioxide, so the spot wets itself, and heat drives more moisture toward it. It stops at the same brake as compost, in the fifties to low sixties, where the organisms are killed by their own heat.

Why one corner, though? Because bins are not uniform. Fines and broken kernels segregate toward the centre as the bin fills, forming a dense core that resists airflow and holds moisture tightly; a leaking roof panel or pocket of green material supplies the perturbation. Any excursion past the crossing point grows, so the least uniform spot decides where.

Which gives the test. If this is right, the hot spot should show three things: locally higher moisture than the bin average, a local rise in carbon dioxide, and a temperature signal arriving late. All three hold, and the third is the useful one — because grain conducts heat so poorly, a temperature cable a metre from a developing spot may register nothing for weeks, while carbon dioxide, a gas moving through the whole bulk, rises measurably in the exhaust air far earlier.

The refuting observation would be a hot spot at bin-average moisture with no carbon dioxide rise, or one that keeps heating after the bulk is dried and cooled through. Either would say the source is not respiring life.

The phrase "dry grain" does damage here. Grain is not dry or wet; it sits in equilibrium with the humidity around it, and that equilibrium shifts with temperature. Grain safe at harvest warmth can be unsafe at the same water content once cooled, because cooling raises the relative humidity in the pores without adding water.

Note too that aeration does the opposite job to turning a compost heap. Turning a heap supplies oxygen so the biology runs faster. Aerating a bin equalises temperature so no convection cell forms, and removes the heat a small excursion makes — precisely how you keep the loss line above the generation curve. Oxygen is not limiting in a bin, so air accelerates nothing.

One honest simplification: the curve-versus-line argument is drawn for a single spot at a single moisture. Real bins have insects generating heat in their own right, mite and fungal successions changing the substrate, and moisture moving in three dimensions. The crossing point shows why the behaviour is a threshold rather than a gradient; it is not read off a chart.

• How temperature cable spacing is chosen against the conduction distance, and what that means for what it misses.

Every term the collection defines is gathered in the glossary.