Should You Wait for the Water to Finish Flowing During Three-Stage Pour-Over Coffee Brewing?

FrontStreet Coffee often says during three-stage pour-over brewing: "Pour the next batch of water when the water level is about to run out!" This has naturally aroused the curiosity of many coffee enthusiasts. Why pour water when it's almost running out, rather than after it has completely drained? What's the difference between these two approaches?

Great question! The best way to prove this is through comparative experiments, so let's dive right into our experiment!

For this comparative experiment, FrontStreet Coffee will use a light roast coffee bean for two brews. Both will use the three-stage method, but with a difference in the third water addition: one brew will receive the third pour of hot water before the water completely drains, while the other will receive it after complete drainage. Let's see how this small change affects the final coffee flavor!

Experiment Parameters

Coffee beans used: Yirgacheffe Gedicho
Brew dose: 15g
Water ratio: 1:15 (225ml)
Grind size: 10 on Ek43
Water temperature: 92°C
Dripper: V60

For both brews, we started with twice the amount of water (30ml) relative to the coffee grounds for a 30-second bloom. Then, slowly pouring in large circles, we added the second batch of hot water (120ml), with both pours completed at the 46-second mark.

The first brew reached nearly empty water level at 1 minute 15 seconds, at which point we slowly poured in small circles the remaining 75ml of hot water. For the second brew, at 1 minute 25 seconds, when the water stream turned to slow droplets (indicating minimal water remaining in the dripper), we poured in the third batch of hot water, again slowly in small circles with the remaining 75ml.

Results

Brewing complete! The total time for the first brew was 2 minutes 15 seconds, while the second brew took 2 minutes 26 seconds. In terms of aroma, the first brew was more pronounced than the second. In flavor, the acidity of the second brew appeared fuller than the first, with other flavors being largely similar.

From this comparative experiment, continuous pouring versus interrupted pouring doesn't seem to make a significant difference. Some might ask: "But aren't the aroma and acidity quite different? How can you say there's no distinction?" The reason for these differences is mainly: the second brew required waiting for complete drainage, taking an additional 10 seconds. During this waiting time, the water in the server experienced some temperature drop, allowing us to perceive acidity more distinctly than in the first brew! Additionally, the hot water in the kettle also experienced about 1°C temperature loss, slightly reducing the extraction rate and thus making the aroma less pronounced! However, both brews remained within acceptable parameters, showing neither under-extraction nor over-extraction, with flavors well-expressed! That's why we say there's no significant difference.

Second Experiment with Different Beans

However, since we used Yirgacheffe Gedicho, which is a light roast bean and relatively tolerant to extraction, we'll now conduct another comparative experiment using a less extraction-tolerant medium-dark roast—"Papua New Guinea Paradise Bird"—with the following parameters:

Brew dose: 15g
Water ratio: 1:15 (225ml)
Grind size: 10.5 on Ek43
Water temperature: 90°C
Dripper: Kono

Similarly, both pours completed the second batch of water (75ml) at the 46-second mark, then one received the third pour before complete drainage, while the other received it after complete drainage.

The continuous-pour Paradise Bird received the third water addition at 1 minute 16 seconds, while the interrupted-pour batch didn't complete drainage until 1 minute 30 seconds, at which point we added the third pour. Brewing complete! The total time for the first brew was 2 minutes 14 seconds, while the second brew took 2 minutes 30 seconds. In terms of aroma, both showed distinct nutty cream notes with no significant difference! However, the flavor profile differed notably—besides the almond and caramel aromas characteristic of Paradise Bird, the second brew clearly exhibited slight bitterness!

The reason for this is that medium-dark roast coffee beans, due to their more porous structure, have less room for error. When extraction time extends, it easily reaches the point where bitter compounds are released in large quantities. Therefore, the second brew, due to increased bitter compounds, lost its sweet-sour-bitter balance, resulting in slightly over-extracted coffee! Besides beans with porous structures, light roast beans with excessive fine particles during brewing will produce similar situations! Too many fine particles not only lead to over-extraction but also cause blockages, making other normal-sized particles over-extract along with them, creating noticeable bitterness!

Conclusion

Therefore, FrontStreet Coffee has consistently adopted brewing methods with higher error tolerance, pouring the third batch of hot water when the water level is about to run out. However, this doesn't mean that interrupted brewing is inherently bad—it simply requires corresponding extraction parameters! The parameters FrontStreet Coffee used in this article were designed around fast brewing methods. Therefore, as long as we adjust them to suit interrupted brewing (such as the coarse grind 46 brewing method invented by Tetsu Kasuya), interrupted brewing can also produce delicious coffee!

- END -

FrontStreet Coffee
10 Bao'an Qianjie, Yandun Road, Dongshankou, Yuexiu District, Guangzhou, Guangdong Province

Important Notice :

前街咖啡 FrontStreet Coffee has moved to new addredd:

FrontStreet Coffee Address: 315,Donghua East Road,GuangZhou

Tel:020 38364473