Baking Basics
Doubling a recipe fails because volume and surface do not grow together
Scaling up is arithmetic on the ingredient list and geometry on everything else, and the things that break are almost always the ones the recipe never wrote down.
By Clara Lindqvist4 min read
The ratio that everything else follows from
Double the linear dimensions of a cake and its volume goes up eightfold while its surface area goes up only fourfold. Heat enters through the surface and has to reach the centre, so the amount of material each square centimetre of crust is responsible for cooking has doubled, and the distance to the middle has doubled as well. That single relationship explains most of what goes wrong when a batch is scaled.
The consequence in practice is a bake that is dark at the edges before it is done in the middle, because the outside spent the extra time at oven temperature while the centre waited for heat to conduct inwards. No adjustment to the ingredient quantities addresses this, because the ingredients were never the problem.
It also explains why small things are so much more forgiving than large ones. A tray of biscuits is almost all surface, cooks through in minutes and gives you very little room to be wrong about anything except timing. A deep loaf tin is the opposite case, and the difference between them is geometry rather than skill.
Keep the depth and add pans
The reliable fix is to hold the geometry constant and multiply the number of vessels. Two tins of the original size bake in the original time because nothing about the heat path has changed. One tin of twice the volume doesn’t, and a tin of twice the diameter behaves differently again from one of twice the depth, since depth is the dimension heat has the hardest time crossing.
Where a bigger single item is genuinely wanted, the compensations are a lower oven and a longer time, so that the surface is punished less while the centre catches up. Cake strips wrapped around a tin, or the double-tin trick, do the same thing from the other side by slowing the edges. A large fruit cake baked at a low temperature for several hours isn’t a quirk of tradition; it is the only way to cook that much dense batter through.
Evaporation scales with surface and therefore does not scale
Water leaves a batter through its exposed surface, so a large batch loses proportionally less of it than a small one during the same bake. The result is a moister, sometimes underset centre from a recipe that was reliable at its original size. In sugar work and in anything reduced on a stove the effect is far stronger, because reduction is entirely a surface phenomenon and a doubled quantity in the same pan can take considerably more than twice as long.
Scaled-down batches meet the mirror image of the same problem. A quarter batch of caramel in a full-sized pan has an enormous surface for its volume, races through the sugar stages and burns while you are still reading the thermometer.
Anything involving a syrup taken to a temperature is particularly exposed, because the temperature is a proxy for concentration and concentration depends on how fast water can leave. Change the pan and you have changed the experiment. Matching the depth of liquid rather than the volume is the closest thing to a general rule here.
Mixing is not a quantity you can multiply
A stand mixer aerating a doubled quantity of butter and sugar has the same beater sweeping a deeper mass, so the material at the outside is worked far more than the material in the middle, and the total air incorporated per unit of batter falls. Creaming times taken from a small-batch recipe will under-aerate a large one, and scraping down matters much more than it did.
The same applies to folding, where a large bowl of whipped egg white takes many more strokes to combine evenly and deflates measurably while you do it. Professional kitchens deal with this by mixing in fixed batch sizes matched to the machine rather than by scaling the mix, which is a useful hint about how much of a problem it is.
What genuinely does scale linearly
Mass scales. Weigh everything, multiply, and the ratios hold. Chemical leavening scales with the batter it has to raise, and the common advice to reduce it in a large batch usually reflects the longer bake time rather than any non-linearity in the chemistry — a bigger cake spends longer in the oven, and gas produced early has more time to escape before the structure sets.
Salt, spices and strong flavourings scale linearly in quantity but not always in perception, and a scaled-up batch tasted at the end is worth more than a calculation. Yeast is the honest exception. Fermentation is a biological process running against a clock, and a bigger dough mass holds its own heat, ferments faster in the middle and needs watching rather than timing.
Common questions
Can I just bake a doubled cake for twice as long?
No, and the attempt is where most scaled bakes are lost. Heat penetration is not proportional to volume, so twice the time overcooks the outside long before the centre is set. Lower the temperature, extend the time, and check the middle with a probe rather than a timer.
Does halving a recipe with one egg work?
It works if you beat the egg and weigh out half of it, which is the only accurate method and takes ten seconds. Using a whole egg in a halved batter adds a substantial amount of water and protein, and in a small quantity that is a large proportional change.
Why does a big batch of dough ferment faster?
Yeast generates heat, and a large mass loses that heat to the room far more slowly than a small one because it has less surface per unit of volume. The interior warms, fermentation accelerates, and a dough scaled up without adjustment can over-prove while the recipe timing says it has half an hour left.
Deputy editor, The Last Bite
Clara writes about baking basics, chocolate & sugar, frozen, mostly the parts other people skip and is unreasonably interested in the detail nobody else checks.
