Last month I spent a few weeks playing with the rate of extraction, and came up with some interesting findings. Notably, I learned that the majority of extraction happens within the first 90 seconds of brewing; by that point most of the solids that are going to wash out of the coffee bean and into the brew already have. This experiment also drove home the importance of sensory analysis to dialing and brewing; relying on a refractometer will not give much indication of how delicious the cup of coffee is.
With this in mind, I began to wonder: could we manipulate brewing so as to maximize on the extraction that happens in the first 90 seconds? What if we tried to keep all the best parts of a coffee without bogging it down with the watery or bitter effects of the later stages of brewing?
Can perceived cup quality improve by utilizing a bypass brew technique?
I was curious what the effects of bypass brewing would have on an innately delicious coffee, like a high quality washed Ethiopian, versus a coffee that tasted fine but might not be anyone’s favorite – in this case, an Organic North Coast Peru. Would I be able to improve both? Did the high scoring coffee benefit from the long extraction of a traditional brew bed? Did the heavier coffee need aggressive extraction in order to present a complex cup?
First things first: What is bypass brewing?
Bypass technique involves adding water directly into the brew, literally bypassing the filter. Rather than running hot water through the grounds, water is added to the liquor itself, diluting it. This can prevent over extraction, limit brew time, and prevent brew bed overflow on automatic brewers.
This technique is most commonly used when brewing large volumes of coffee on batch brewers. Putting all the necessary water to brew a huge batch through the grounds might extend brew times well beyond the desired window, raising extraction percentage and strength. If the brew bed is filled with coffee, there’s possibility of an overflow in the filter, too. Limiting the amount of brew water pouring into and through the filter can help avoid or alleviate all of these issues, and the remaining water can be added directly to the brewed coffee.
Bypass brewing lives in the shadows of our industry. Everyone knows about it, and most people have done it at least once or twice, but probably only as a last resort. There’s not much information on best practices or varied techniques, and it’s generally assumed that bypass just isn’t the best way to brew coffee. Why not? It’s a useful technique and can be entirely necessary in certain situations. Why not explore and perfect it?
I wanted to know a little more about the variables at play in a bypass brew. As mentioned above, I was curious whether I could improve a coffees perceived cup quality by playing around with bypass parameters.
There was some specific information I was looking for:
- What is the preferred ratio for brew water to use for this technique?
- What is the ideal ratio of bypass water to add?
- What is the ideal grind size to use to brew the concentrate?
At this point, it’s important to note that we work in less than ideal lab conditions at the Crown. I’m looking forward to being able to execute repeatable, reliable experiments that produce relevant data, but until then, I do my best to keep things as consistent as possible. Nonetheless, we should all be fully aware that none of this data should be considered valid scientific findings.
The first order of business was to get a baseline brew, properly dialed in using a traditional method. This would be the standard against which my group of tasters would judge the other coffees for preference. I brewed 20g grams of coffee on a Kalita with 360g of brew water, resulting in a final beverage volume of 320g (coffee absorbs at least its weight in water, so final brew volumes will always be less than the total brew water added).
Then I began brewing concentrates by using only a percentage of the original brew water, weighing the final beverage volume, and adding enough bypass water to match the 320g volume of the original dial. My small group of cuppers preferred the cups that used only 40% of the brew water from the first brew, or about 150g of water to 20g of coffee (instead of the original 360g water).
It was essential that I keep things as consistent as possible. Brewing coffee comes with a minefield of variables, and it’s important to control these as much as possible in order to collect valuable data. For this reason, I chose to use Clever Brewers for this part of the experiment, allowing me to add all the brew water at once, limiting agitation, and control extraction times. I set the Clevers to drain at 1:30 minutes across the board, because my previous experiments had shown that most of the important stuff had already been extracted by this point.
I began experimenting with two different coffees. One was a delicious coffee from the Gedeo Zone of Ethiopia, farmed by Desta Gola. It’s been one of my favorite coffees this year. The other was a certified Organic North Coast Peru, chocolaty and full bodied but not very delicate or elegant. I wondered: would the Ethiopian coffee lose some of the complexity offered by the end stages of brewing? Would the Peru lighten up and become more acidic by utilizing bypass?
I brewed a 40% concentrate of each coffee on a Clever brewer, which yielded me about 50g of brewed coffee liquor. To this is added 20% bypass water (10g), then 30% bypass, then 40%. Again, I had my cuppers taste and evaluate each coffee on a 0-5 hedonic scale and include qualitative descriptions of each so I could see how they changed.
Here I got some really interesting data. Most of my cuppers scored the 20% and 30% bypass brews higher than they scored the traditional brew method! I had successfully improved both coffees by using bypass technique.
There was one more variable I wanted to wrangle with: grind size. Some folks feel that grinding coarse is really ideal for bypass brews, and I wanted to test this theory. After all, wouldn’t it make more sense to grind fine so as to expedite extraction?
I stuck with the same concentrate recipe, and used the 20% bypass recipe that my cuppers seemed to prefer. I experimented with four different grind sizes on the EK43, ranging from super fine to super coarse. The coarse ground coffee scored overwhelmingly higher than the finer grind size. We found that the fine grind brought out some really unusual, unexpected (and for the most part, undesirable) flavors in the brew. Meanwhile, the coarse settings scored the almost universally higher than the traditionally brewed dial.
Experimenting with grind setting brought up some really interesting results, but I wasn’t able to collect enough data to really solidify my findings. You’ll recall that one of the questions I wanted to answer was how a high quality coffee would respond in contrast to how a blender quality coffee responded to bypassing. While I found early on in the experiment that I could improve preference scores on both coffees by using bypass, it wasn’t until I got to the grind size experiment that I found differences in relative improvement between the two coffees.
Unfortunately, the only relevant data I was able to collect with my flawed scientific method was that the extreme coarseness caused the coffees to be more polarizing. Although on average, the coarse-ground-bypass-brew scored higher than the traditional brew dial, there was less consensus among my cuppers over which was better.
Can the perceived cup quality of a coffee be improved by using bypass technique?
The short answer is yes. I was able to increase my cupper’s hedonic preference scores by brewing a concentrate and diluting it. However, this technique produced only half the total brew volume that traditional brewing would. In other words, I was using the same dose of ground coffee to make half the amount of drinkable coffee. Given the amount of work it takes to produce each bean, and the volume of coffee we are wont to drink on any given day, this doesn’t seem like a great solution. On the other hand, utilizing bypass method cuts brew times in half, so from a flow perspective in a busy cafe this might actually be more efficient. It also makes the cup subjectively more delicious.
The question then is: what is our goal when brewing? Is it to get the most possible out of the bean, or is it to get the bean taste the best it possibly can?
In any case, there’s room for more exploration here.