Welcome back. Today I am going to continue with a few additional thoughts on water temperature. This is clearly a big topic, and there are numerous ideas to explore within it.
Last time I talked about my density measurement chart in correlation with choosing the ideal brewing temperature. What I didn’t talk about was how I go about measuring the density of whole bean roasted coffee. Some of you may have your own method or even a density reader to get this measurement, but since using density for the purpose of preparing coffee has not been a cafe standard in the past I will explain my method.
First I want you to understand the formula I use (g/cl), which results in a density number that can be used to determine the water set temperature. If you missed it, the reference chart is in my last blog post. In order to get a consistent volume measurement in centilitres, I use a 250ml/25cl graduated cylinder which is a common tool in an SCA certified lab. Simply fill the cylinder to 25cl and weigh the beans (not including the cylinder weight of course). Using a known volume of 25cl means that all I need is the weight of the coffee and the rest of the math is very simple. Let’s say your 25cl of coffee weighs 104.3g. To get the density number simply divide the weight by the volume (103.4/25) to get 4.172g/cl. Based on my chart you would set your water temp to 201 or 202°F (around 94°C). I would probably round up to 202°F since the measurement is close to the next temperature/density benchmark.
You might wonder why I use centilitres rather than millilitres or litres. This is because the resulting number is much more approachable; a whole number and a fraction/decimal (4.172). If we use ml as our volume the resulting number would be smaller (0.4172), and using litres would result in a much larger number (417.2). Both of these numbers create a more burdensome impression for the user and I wanted to avoid that.
I have some tips on other ways to measure density if you don’t have a graduated cylinder in the book as well.
The next topic of the day is an interesting one which recently came to light from reading Mr. Jonathan Gagné’s Instagram page, where he references water temperature and viscosity.
View this post on Instagram
I previously had the intuition that brew water temperature would slightly affect drawdown time in a V60 because water viscosity depends on its temperature, and @jimseven’s similar comment in his recent V60 video prompted me to look into this dependency in more details. I used data from IAPWS 2008 and Engineers Edge Machinery's Handbook to create the above figure that displays the dynamic viscosity of liquid water at atmospheric pressure versus its temperature. The variation in viscosity is *much* stronger than I expected; it approximately halves if you heat water from 20°C to 80°C ! Darcy’s law states that the flow rate in a percolation goes with the inverse of water viscosity, if everything else is kept constant. I used this to display the blue line above in the figure, which shows the relative flow rate for different water temperatures through the coffee bed. I also used temperature logging data from my previous Instagram posts to place little circles and triangles at the typical average slurry temperatures when using a plastic and glass V60s, respectively (in both cases using water right off boil). The difference in slurry temperature in these cases should create a change of flow by 10%, i.e. the hotter plastic V60 slurry would draw down 24 seconds earlier than a 4 minutes brew with the glass V60. This is not a small effect ! For me this is yet another reason to use hot water, because I agree with @jimseven that a faster drawdown is good; it reduces the risks of channeling, and potentially allows you to grind finer, unlocking more coffee flavors. #coffeescience #coffeegeeks #pourover #waterforcoffee
I particularly find this interesting, and while I disagree with the assumption that hotter water is best, what sticks in my mind is a cause and effect that I have observed for a long time now. This observation is basically that the cooler the brew water, the slower the water will flow through the filter and vice versa. I have been trying to wrap my head around why this may be happening for some time, and I believe our coffee loving astrophysicist has the answer.
In testing numerous brews I have witnessed this trend of cooler water filtering more slowly, and specifically in cases where excess fine particles cannot be blamed for the filter choking. Assuming Gagné is correct, this equates to the cooler water being more viscous and therefore less efficient at flowing through a paper filter (especially depending on the pores and thickness, which I will discuss in a later post). So if your brew has a very long drawdown and you don’t think it is because of fine particles, it could well be that your water is too cool. If you are using my temperature/density chart I believe you will already be in a very good range for your coffee, so this could bring to light any major temperature loss issues between water heating and the actual brew.
It is also important to note that your water temperature and resulting viscosity are only one variable in a complex system. Releasing fines from their static connection to larger particles through turbulence also compounds the impact of slower water flow due to temperature. So while you will get a slower water flow with a lower water temperature, you are also avoiding extracting less desirable flavours, and can maximise the extraction of the components which taste better. The key will be to avoid choking the filter, which may mean less turbulence than is often prescribed by others.
Last, I want to point out why I don’t believe that hotter water is always best.
It is easy to say something definitive if your only goal is to maximise a single data point, such as extraction yield. However I have rarely found a correlation between abnormally high extraction (higher than 21%) and better flavour. Similarly, I have found various coffees which express their best flavours with a lower extraction yield. There is no singular solution for coffee, likely because there are so many variables at play. Roasting style plays a huge role in this, and it seems that the approach of seeking ever higher extraction yields is valuing a certain style of roast (mostly very light) over the variety of styles which are present in the real world of coffee.
If you stop seeking maximum extraction and pay attention to the final results in the cup, the variability and versatility of coffee will be more apparent. The point is that you want enough extraction from your coffee, and not complete extraction. We all know that there are components of coffee that taste terrible.
If high extraction yield were the only important factor in coffee brewing, we would be ignoring time constraints, grinding to a Turkish grind, using boiling water, and everything would be easy. In the real world, if you have ever brewed coffee and tasted it comparatively should know that this is not what anyone wants to drink. So surely we need to either accept that we don’t want all of the compounds in coffee or else figure out how to make those compounds taste better. The latter is probably a conversation for an organic chemist, but I would be excited to have the discussion.
Stop chasing the numbers and start understanding your coffee extraction better by pushing the button below.