Tuesday, August 20, 2013

Hypotheses Part 2

Fancy bars these days like to serve their cocktails with a single big chunk of ice. Less surface area means slower cooling means less dilution of your expensive elixir over time. I don't buy it.

Here's my reasoning. Ice serves two purposes in cocktails. First is to chill liquids, second is to keep those liquids cold. Let's consider these in turn.

At what temperature is ice? Zero degrees Celsius, right? Sometimes. Your freezer is actually quite a bit colder, so fresh ice from the freezer is just as cold. Pick up an ice cube with a damp hand, and it will freeze the water on your fingers and stick. It's not very pleasant. But do it anyway. For science. Ice sitting in a bucket for a while will be pretty close to zero. You'll know it if the ice is wet.

Let's ignore melting for a moment. We can talk about whiskey stones if you prefer. Once you drop a cold stone in a room-temperature liquid, the two will equilibrate in temperature. The liquid chills, and the stone warms. But the stone doesn't warm uniformly. The outside warms up first and passes this heat inwards. This means the inside of a whiskey stone or block of ice can be substantially colder than the outside. Every bit of cooling of the liquid has to come from the stone/ice, so with a single large chunk, it takes longer to get to the cooling power of the center, meaning it takes longer to cool the liquid.

Let's talk about wet ice, so that the whole chunk is at zero or so. Now all cooling of the liquid results in melting of the ice. Heat from the liquid causes the ice to transition into a liquid state until the two reach thermal equilibrium. But for a given volume and initial temperature of liquid that you're chilling, the exact same amount of ice will melt regardless of its shape. So again, less surface area means the liquid takes longer to chill.

Put these two effects together, and using one large chunk of ice will leave more cooling power locked in the center of the ice, and more of the chilling will have to come from melting rather than simple heat exchange. Result: bigger ice cube = slightly more dilution.

How about keeping a drink cold? Now the liquid is presumably at the same temperature as your ice already. Also presumably, you're a distinguished imbiber and sip your drink slowly. We can ignore simple heat exchange between ice and liquid here since that will occur quickly compared to the lifetime of your drink. Now, let's compare a glass filled with crushed ice vs. one big chunk. As the heat enters your glass, it eventually makes its way to your ice and melts it. Equal amounts of heat means equal amounts of melting. However, with one big chunk, your drink can get a little warmer in regions further from the ice. Since the volume of liquid that needs cooling becomes effectively smaller, less ice has to melt. Score.

On the other hand, if most of your ice is sitting above the surface of your drink it's chilling not only the liquid, but also the air. Most of your scotch's lifetime will be in this state. Yes, even if the ice were not sticking shamelessly out, the liquid would be chilling the air, and end up melting the ice anyway. But it would also have less surface area and sit lower in the glass resulting in more insulation from warmer air.

These are two small, competing effects, and it's not clear without experiment which wins. My guess is that a big, but not huge ice cube is better here. In part because most people won't care much about the difference between a drink at 0C and 3C.

My recommendation: always chill drinks with small ice cubes fresh from the freezer if possible. For serving, ice cubes from a standard ice cube tray are probably big enough. Bigger than that, and you're just showing off. If you really hate dilution, get whiskey stones and/or double-walled glasses. If your glass has a stem, use it so you're not pumping heat from your hands into your drink. If I've left something out of my analysis, please let me know!


  1. I remember the heat equation from my partial differential equation class! I hated that class. I was following you until you mentioned that wet ice is at 32. It took me a moment to think how did it jump from 0 to 32? Oh yeah, stupid Fahrenheit.

    For me to understand this, one of the critical concepts I remember from my chemistry class: At constant pressure, a material will remain the same temperature during a phase change. So, when ice is solid, it can get as cold as the environment allows it to. The minute it starts to melt, it will stay at 0 (at standard atmospheric pressure) until all the ice is water. I don't remember if the heat equation applies when a material is undergoing a phase change, would these two principles seem to contradict?

    Oh, I use whiskey stones.

  2. Oops, I switched from F to C and failed to change it everywhere. Sorry. It's fixed.

    You're right that the temperature of the ice is fixed while it's melting, but we have to separate the notions of temperature and heat. As heat from the drink passes into the cold ice, first the ice's temperature will rise until it reaches 0C. But since the liquid is still warmer, it will still pass heat into the ice. But at this point, the heat causes the ice to melt instead of changing its temperature. So the resulting heat flow means the ice melts and the liquid gets colder. When does this stop? When the temperatures are the same. Or when the ice is gone. Whichever comes first. Hope this clarifies things.

    1. That does make sense, but only if I change the word "heat" in your reply to "energy." Is that fair? If I think of the system a a flow of energy, it takes energy to melt ice and the energy has to come from somewhere, so it takes it from the surrounding libation, lowering the temperature. This is Kurt, by the way. I don't know why it doesn't pick up my name.

    2. Hey Kurt! Yes indeed. Heat is a form of energy and is measured in the same units. Well few people measure energy in BTUs, but Joules on the other hand....