04 · Basic · Intermediate
Brewing Technique — Tools, Flow Rate, Methods
Percolation and immersion, ribs and paper, and how a recipe is designed
Written by Chanho Hong
What this chapter covers
Why cones and flat bottoms are both percolation brewers; how hybrids such as the Hario Switch and Clever separate grind size from contact time; what ribs and filter papers actually do to flow rate; and the design logic behind the 4:6 and Chanho-Tornado recipes.
If theory (Chapter 02) is "what and why", this chapter is "how, with your hands". Even with the same roasted coffee · the same variable targets, which dripper, which water stream, in which order you pour still changes the cup.
Brewing tools — percolation · immersion · hybrid
Drippers are first divided by extraction principle. Extraction in which the water passes through the coffee bed is percolation, and extraction in which the coffee stays submerged in water and is drawn out by time is immersion. To these is added the type that uses a valve to switch between the two principles inside a single tool — hybrid — making three branches in all.
"Cone / flat bottom / immersion" are commonly placed side by side as three of a kind, but the cone and the flat bottom are both percolation devices. The principle is the same in that water keeps flowing through; what differs is only the geometry of the bed. In a cone the water converges towards the centre and the bed is deep; in a flat bottom it passes through broad and shallow. Immersion sits at a different level, because extraction there is governed not by flow but by steeping time. Get this distinction right and you can predict the character of a tool the first time you see it.
| Family | Examples | Extraction behaviour | Tendency |
|---|---|---|---|
| Percolation · Cone | Hario V60 (spiral ribs), UFO dripper (80° cone · very open base), Kono | Convergence to the centre · deep bed · fast flow rate, control over variables↑ | Bright · articulate, sensitive to pouring skill |
| Percolation · Flat | Kalita Wave (three holes), Fellow Stagg [X], Origami (octagonal · takes either filter) | Flat bed · even penetration, high resistance to flow | Balance · sweetness · reproducibility↑, beginner-friendly |
| Immersion | French Press, Aeropress | Coffee stays submerged in water and extracts with time | Body↑ (press) · forgiving · consistent |
| Hybrid | Hario Switch, UFO dripper + Hario Switch base, Clever, Bonavita immersion dripper | Close the valve for immersion, open it for percolation. Both principles used in sequence within one cup | Contact time controlled directly — strength achievable even at a coarse grind |
In percolation, the only real way to lengthen contact time is to grind finer. But the finer you grind, the more fines you create, and over-extraction and clogging follow. The hybrid closes the valve and thereby separates contact time from grind size. Since time can be bought with the valve, you may grind coarse, and a coarse grind produces fewer fines, so extraction uniformity rises. This is exactly why the Chanho-Tornado modern recipe seen earlier grinds almost twice as coarse as the classic and still comes out stronger.
Starting closed puts immersion first; starting open and closing partway gives percolation followed by immersion. The adjustment handle is the length of time you keep it closed — increase it and contact rises, shorten it and contact falls. The greatest practical benefit is being able to shift strength without changing the grind. The drainage rate once the valve is open is still decided by the base structure and the paper, so even in a hybrid the physics of the percolation phase applies unchanged.
What do the ribs do?
Ribs are ridges raised on the inner wall of the dripper. They stop the paper filter from sealing against the wall, creating channels for the water to drain and a path for displaced air to escape. Without ribs, wet paper sticks to the wall, a vacuum forms, and drainage stops or becomes irregular.
In general, the more numerous, taller and longer the ribs, the less contact area there is between paper and wall, so venting and drainage run more freely and the effect works in the direction of a faster flow rate. Conversely, when the ribs are few or confined to the base, the paper in the upper section clings to the wall, resistance rises, and the flow rate slows. What is interesting is that each tool solves the same problem differently.
| Tool | Structure | Solution |
|---|---|---|
| Hario V60 | Spiral ribs over the full height, right to the rim | Venting along the whole path via the ribs — open, on the fast side |
| Origami | 20 vertical ribs | Many ribs. But with a cone filter it is fast, and with a wave filter it interlocks with the ribs and becomes slow |
| UFO dripper | Smooth wall + recessed air channels | Grooved channels instead of protruding ribs, fixing bypass at an intended location |
| Kalita Wave | No wall ribs | The paper's own 20 waves reduce contact with the wall |
| Melitta · traditional Kalita | Ribs only near the base | The paper seals against the wall at the top — the narrow outlet is the main resistance |
| Chemex | No ribs | The paper is folded into three layers and hooked over the spout to create an air passage |
The direction of "more ribs, faster" is consistently supported by manufacturer explanations and field observation, but no controlled experiment isolating rib count · height alone and measuring drawdown time has been published. What the literature confirms about the role of ribs goes only as far as preventing the paper from sealing and a vacuum from forming. In practice, once sealing is prevented, what governs flow rate is the permeability of the paper and the geometry of the outlet, and that influence is far larger — one measurement found a 2.2-fold difference in flow-through speed from changing the paper alone in the same V60 (Hario unbleached 15.3 mL/s vs Hario untabbed bleached 6.89 mL/s). It is more accurate to understand ribs not as a flow-rate dial but as a condition that must be met for normal operation.
The raw material of the paper filter changes extraction
Even with the same dripper and the same recipe, changing the paper changes the result. Four variables determine a paper: bleached or not · fibre source · crepe (wrinkle) structure · thickness.
| Variable | Detail | Effect on extraction |
|---|---|---|
| Bleached or not | Bleached (oxygen · ECF) vs unbleached (brown) | Unbleached gives more papery taste from lignin. Measured, the unbleached leachate was 5 ppm and the bleached 0–1 ppm. Rinsing with hot water reaches 0 ppm by the third rinse — it is a problem that rinsing resolves |
| Fibre source | Wood pulp / abaca (Manila hemp) / bamboo / eucalyptus | The longer the fibre, the larger the pores and the higher the permeability. Abaca and bamboo are long; eucalyptus is the shortest. Long fibres keep their structure better when wet, so the flow rate is stable |
| Crepe structure | Single- or double-sided wrinkling, height of the wrinkles | More decisive than thickness. A double-sided, high-crepe paper creates passages between paper and wall and is in fact faster than a thin paper |
| Thickness | Basis weight (g/m²) · calliper | On its own it does not predict flow rate. Thicker papers do tend to hold back more fines and lipids, but it has to be read together with structure |
In CAFEC's published specifications, the thinnest T-92 (0.15 mm) is slow and the thickest T-90 (0.28 mm) is fast. What made the difference was not thickness but crepe structure (T-92 single-sided wrinkling / T-90 double-sided high wrinkling). In independent measurements too, the order by thickness did not match the order by flow rate. "Thick = slow" does not hold.
The explanation that a paper filter filters out lipids and makes a clean cup points in the right direction but is exaggerated in magnitude. Measured, of the cafestol present in the roasted coffee, the paper catches 12.4%, 87.5% remains in the grounds, and 0.15% reaches the cup. In other words, hot water barely extracts the lipids in the first place. It is closer to reality to understand the paper not as a "lipid removal device" but as a fines barrier — a substantial part of the difference in texture comes not from lipids but from suspended fines.
First match tool and paper (Origami + wave = slow, Origami + cone = fast). Then, if the flow rate is faster than your target, move to a denser, low-crepe paper; if slower, to a long-fibre, high-crepe paper. If you use unbleached paper, rinse thoroughly with hot water. When you change paper, change one thing at a time and re-set the grind — a change of paper changes the flow rate, and once the flow rate changes the previous grind no longer fits.
Every tendency recorded here is a relative direction. With the same dripper and the same paper, results can be reversed entirely once the density of the coffee · the degree of roast · the grind particle size · the water composition · the pour change. In fact, no peer-reviewed study yet exists that changes filter brand or material and measures the resulting change in TDS · extraction yield · sensory score — the account above rests on measurements of the physical properties of the paper itself and on field consensus. Tool comparisons must always be verified directly with your own coffee and in your own environment.
① A busy shop with several staff. To reduce variance, a forgiving flat bottom (Kalita) or a hybrid (Clever · Hario Switch) carries a lower probability of failure. The hybrid in particular has its immersion phase absorb the influence of pouring technique, so results do not swing greatly even when skill levels differ.
② Competition · omakase, maximum expression. To maximise bright acidity and control variables finely, a cone (V60 · UFO) is advantageous. Small differences in pour and flow rate show up directly in the cup, so a skilled hand has a wide expressive range.
③ When you want a coarse grind and a clean cup. If fines mean the muddiness will not go away, move to a hybrid such as UFO + Hario Switch base. Take the grind coarse (700–800 µm) to reduce fines, and recover the lost contact time as time with the valve closed. The point is to separate grind size from contact time.
Flow rate — fast · slow · turbulence
Flow rate is the speed at which water passes through the coffee bed. How fast · how high · where you pour the water creates the flow rate and the turbulence (the degree to which the water stirs the coffee), and this regulates extraction yield through contact time · agitation · uniformity.
A thick, fast stream → contact time↓, but agitation (turbulence)↑. The bed can be gouged out, with a risk of channelling. Pair it with a coarse grind to keep the balance.
A thin, slow stream → contact time↑, agitation↓. The bed is stable but there is a risk of over-extraction · stalling. Watch for clogging if paired with a fine grind.
Turbulence is a double-edged sword. Moderate turbulence wets the coffee particles evenly and helps uniform extraction, but excessive turbulence digs out particular spots and creates channelling · stalling from fines. Advanced technique therefore focuses not on making turbulence "strong" but on making it "even" (→ the spiral pour of the Chanho-Tornado).
① It drains too fast and tastes watery. Pour low and gently to reduce turbulence, or grind finer to slow the flow rate. Buying contact time raises the extraction yield.
② It stalls, muddy · over-extracted. Grind coarser and raise the stream slightly to restore flow and agitation. Releasing the stalling from fines returns the brew to uniform extraction.
③ Securing reproducibility in the shop. If results swing because each member of staff pours differently, standardise stream height · speed as an SOP (e.g. "low stream · 8 g per second") or change the tool to a flat bottom · immersion device that is insensitive to turbulence. Flow rate · turbulence · grind size always move as a set, so fixing one simplifies the tuning of the rest.
Brewing methods — 3 Pour · 4:6 · Chanho-Tornado · intervals
A method is a design for how the bloom and the main pours are divided. From approaches that break the pour into several bursts (pulse) to those that pour all at once (continuous), each gives a different degree of control · reproducibility · taste.
Pulse vs Continuous — the two families
| Item | Pulse (broken into several) | Continuous (all at once) |
|---|---|---|
| Control of variables | Fine adjustment possible at each pour | Simple · repeatability↑ |
| Uniformity | The bed is re-agitated at every pour (double-edged) | Decided by a single agitation |
| Operational efficiency | Requires time · concentration | Fast and consistent |
| Late-stage temperature | The bed can cool between pulses | Favourable for holding temperature |
The principal methods
① 3 Pour (three pours) — after the bloom the main pour is split into two, for three pours in total. Example: 0:00 bloom 60 g → 0:30 up to 180 g → once it has drained, up to 300 g. The standard of balance and control.
② 4:6 Method (Tetsu Kasuya) — the water is divided broadly into 40% + 60%, with the first 40% (two pours) controlling the balance of acidity · sweetness and the last 60% (several pours) controlling strength. An intuitive frame that separates the variables into "taste" and "strength" for independent control.
③ Chanho-Tornado — a one-directional continuous spiral pour developed by Chanho Hong in a busy Australian coffee bar in 2014. After the bloom the water is poured in a vortex shape without breaks, agitating the whole bed evenly. Minimising the number of pours makes it fast and consistent, and it draws out the bright acidity · sweetness of light-roast specialty coffee well. The Chanho-Tornado is not a single fixed recipe but is run in two branches matched to the tool — the classic (V60) for percolation only, and the modern (UFO + Switch base), which alternates immersion and percolation.
| Variable | Classic — V60 | Modern — UFO + Switch base |
|---|---|---|
| Dose | 15 g | 15 g |
| Grind size | 300–400 µm | 700–800 µm |
| Water temperature | 96 ℃ | 93 ℃ |
| Bloom water | 30–45 g | 80–100 g |
| Bloom time | 30–60 s | 60 s immersion (closed) → 30 s open |
| Agitation | 8 times | 4 times |
| Total water | 225–240 g | after the 90 s bloom, close again and 200–220 g |
| Ratio | 1 : 15–16 | 1 : 13.3–14.7 |
| Stirring | 8 times | 8 times |
| Brew time | 1:30–2:00 | 2:00–2:30 |
In the classic V60, extraction happens only while the water is passing through. To buy contact time you have to grind fine (300–400 µm) and raise the temperature (96℃) to increase extraction per unit of time. The modern closes the Switch valve to create a phase where the water is held in place. Because immersion buys the contact time instead, you may grind almost twice as coarse (700–800 µm), and a coarse grind reduces fines and lowers the risk of over-extraction · clogging. Dropping the temperature to 93℃ is for the same reason — the long contact time is offset with temperature.
In the classic, grind size is the main handle. If the time comes in faster than 1:30, grind finer; if it exceeds 2:00, grind coarser. In the modern, bloom immersion time is the main handle — lengthen the 60 seconds it stays closed and contact rises, shorten it and contact falls. The difference in agitation — 8 times for the classic · 4 for the modern — follows the same logic. Since the immersion phase already wets the bed evenly, the modern needs less agitation.
How do intervals change extraction?
The interval between pours changes two things. ① Bed temperature — the longer the interval, the more the bed cools and the lower the extraction power in the later stages. ② Number of agitations — the more pours, the more re-agitation, which cuts both ways for uniformity. So more pours is not always better. Design interval · count to match the taste you are after.
① Competition vs shop, separating methods. For competition use 4:6 to tune taste · strength separately; in the shop use the Chanho-Tornado to reduce variance between staff. The purpose (expression vs reproduction) decides the method.
② Steering direction with 4:6. For "brighter", increase the water in the first pour of the front 40%; for "stronger", increase the number of pours in the back 60%. Use taste (front) and strength (back) as independent dials.
③ Optimising for high-volume service. In a busy bar that has to produce dozens of cups a day consistently, the Chanho-Tornado (continuous spiral · single pour) minimises the number of pours to secure speed · consistency, and avoids the problem of late-stage temperature dropping because of long intervals. Interval · water volume · number of pours are the adjustment handles.
Questions this chapter answers
- How do cone and flat-bottom drippers differ?
- Both are percolation brewers — water passes through — and what differs is the geometry of the bed. A cone converges the water toward the centre, so the bed is deep and flow is fast, giving more control over variables but making the result sensitive to pouring skill. A flat bottom lets water pass broad and shallow, which is more even, more forgiving and more repeatable.
- Why can a hybrid immersion dripper taste strong at a coarse grind?
- In pure percolation the only real way to extend contact time is to grind finer, and finer grinding brings more fines, over-extraction and clogging with it. A hybrid closes a valve and so separates contact time from grind size: the valve buys the time, which lets you grind coarser and cut the fines.
- Does a thicker filter paper slow the brew down?
- It does not follow. CAFEC's own published specifications show the thinnest paper (T-92) as slow and the thickest (T-90) as fast. Flow is governed by the crepe structure and the fibre stock rather than by thickness — in one measurement, changing only the paper in the same dripper more than doubled the flow rate.
Figures and regulations are verified against a source hierarchy: Tier 1 in-house knowledge base → Tier 2 SCA·WBC and peer-reviewed literature → Tier 3 web. Anything not adequately supported is marked [to verify] in the text.