D-Value (Decimal Reduction Time)

Heat kills microorganisms on the same log-linear pattern chemicals do, and the number that describes it is the D-value: the time at a fixed temperature needed to reduce the population by one log — a factor of ten. Fifteen minutes that delivered five logs implies a D of three minutes. Every further three minutes removes another factor of ten from whatever is left.

W. D. Bigelow set this out in 1921 in the Journal of Infectious Diseases, working on the thermal processing of canned food, and the framework has run food and beverage process engineering ever since. It is the same first-order idea Chick published for chemical disinfection thirteen years earlier, with temperature in the place of concentration. None of what follows is a clinical or medical calculation. This is thermal processing of a product — pasteurisation, retorting, a hold tube on a milk line — and nothing here concerns sterilising anything that will be injected into a person.

A D-value without its temperature is not a number. That is why the literature always writes it with one attached: $D_{121}$ is the decimal reduction time at 121 °C, and the same organism has a wildly different D ten degrees either side — the z-value page is about exactly how different. A D quoted bare is unusable, and a D copied from a paper without its temperature is worse than unusable, because somebody will assume one.

It is not only the temperature that has to travel with it. D depends on the substrate as much as on the organism. Fat, sugar and low water activity all protect microorganisms, sometimes by a large factor, so a D measured in phosphate buffer can badly understate what the same spore survives in a fatty, sugary food. Acidity cuts the other way — which is the reason the canning world splits its products at about pH 4.6 and gives high-acid foods an entirely gentler process. A borrowed D is a starting hypothesis, not a design input.

The trade tends to speak in whole D's, and it is a good habit. A 12D process — the classic botulinum cook for low-acid canned foods — is twelve decimal reductions and nothing more mysterious than that: twelve times the D-value of Clostridium botulinum spores at the process temperature. Counting in D's keeps the arithmetic in front of you.

Two things the simple form leaves out. First, as with Chick's law, the straight line is an approximation with a shoulder at one end and a tail at the other, and a D fitted across the whole curve splits the difference — fit it on the log-linear stretch. Second, the time this equation returns is time at temperature. A real vessel takes time to come up and time to cool down, and both of those deliver lethality: the product is being cooked all the way through the cycle, not only during the hold. Counting the hold alone means you are overprocessing, sometimes substantially, and the F-value page is how the whole cycle gets added up properly.

One last reminder that a reduction is relative. Twelve logs from a starting count of $10^3$ per container and twelve logs from $10^6$ leave very different numbers of survivors, and the process cannot tell the difference. Incoming load matters as much as the schedule, which is why sanitation upstream of the cook is part of the safety case and not merely good housekeeping.