If you already know that water chemistry affects coffee flavor — that magnesium pulls brightness, calcium adds body, and bicarbonates buffer acidity — then the next question is obvious: which recipe should I actually use?
The specialty coffee world has produced at least eight well-known water formulas, each with different mineral profiles, different costs, and different levels of effort. Some are backed by peer-reviewed research. Some come from champion baristas. One is literally just a single ingredient in a jug of distilled water.
This article compares them all. Not the theory — the practical reality of making each one, what it costs, and when it makes sense to choose one over another.
For the underlying science of why water chemistry matters at all, start with why water quality matters more than your coffee beans.
The Recipes at a Glance
Before diving into each one, here’s the full side-by-side comparison. All hardness (GH) and alkalinity (KH) values are in ppm as CaCO3.
The Eight Recipes, In Detail
1. RPavlis — The One-Ingredient Wonder
What it is: 0.38g potassium bicarbonate (KHCO3) dissolved in one gallon of distilled water. That’s it. One ingredient, no concentrates, no scale.
Mineral profile: Zero hardness. About 50 ppm alkalinity. No calcium, no magnesium. The only mineral doing anything is the potassium bicarbonate buffer.
Invented by: Robert Pavlis, a chemistry professor who argued that hardness minerals aren’t necessary for good extraction — the buffer alone handles enough.
Cost: Under $0.01 per liter in minerals. A 1 lb bag of food-grade potassium bicarbonate runs $10-15 on Amazon and makes roughly 1,200 gallons.
Difficulty: Trivial. Weigh 0.38g on a $15 jewelry scale, dump it into a gallon jug of distilled water, shake.
Flavor profile: Mild, clean, slightly soft. Lacks the brightness and complexity that magnesium delivers. Won’t reveal the full potential of a $25 Ethiopian natural. But it’s a genuine upgrade over most tap water, and it’s completely scale-free — zero calcium means zero limescale, ever.
Best for: Espresso machine owners who want safe water without fuss. People who want the simplest possible upgrade from tap. Dark roast drinkers who don’t need aggressive acidity extraction.
2. Melbourne — The Ultra-Soft Approach
What it is: A very low-mineral recipe from the Melbourne competition scene. 24 ppm GH, 12 ppm KH. Using the Barista Hustle Two Bottles method: 24 mL of Bottle 1 (hardness) and 12 mL of Bottle 2 (alkalinity) per liter.
Mineral profile: Low everything. Minimal Mg from Epsom salt, minimal buffer. TDS around 50 — barely above distilled.
Flavor profile: Delicate, transparent, tea-like. Extremely clean cup with almost no mineral-driven body. The coffee’s origin character dominates completely — for better or worse. Under-extracted notes (sourness, thinness) can appear with darker roasts because the water lacks extraction muscle.
Best for: Competition cupping. Very light roasts where you want nothing between you and the bean’s terroir. Not a daily driver for most people — the cup can feel hollow.
3. Holy Water (Cameron) — Clean and Sweet
What it is: A recipe from Canadian barista Kevin Cameron. 62 ppm GH, 23 ppm KH. The 2.7:1 GH:KH ratio keeps alkalinity low enough that acidity stays bright and present.
Mineral profile: Moderate hardness, very low buffer. More extraction power than Melbourne but less acid-buffering than Rao/Perger or SCA.
Flavor profile: Clean, sweet, and clear. Acidity reads as fruit rather than sharpness. Well-defined flavors with good separation between notes. Lacks the heft of higher-TDS recipes.
Best for: Light roast pour-over, especially single-origin Africans (Ethiopian, Kenyan) where you want to showcase bright acidity without it becoming harsh. One of the best light-roast filter recipes available.
4. Barista Hustle #4 — The Balanced Middle
What it is: The fourth (and most popular) recipe from Barista Hustle’s Two Bottles system. 80 ppm GH, 40 ppm KH. A 2:1 ratio that splits the difference between brightness and body.
Mineral profile: Moderate Mg from Epsom salt, moderate NaHCO3 buffer. TDS around 120 — close to SCA target.
Flavor profile: Clean, bright, balanced. More body than Holy Water, more acidity than Rao/Perger. A genuine all-rounder that doesn’t push hard in any direction.
Best for: Light-to-medium roast filter brewing. If you want one recipe for V60, Kalita, AeroPress, and Chemex, this is a strong default.
5. Rao/Perger — The Best All-Rounder
What it is: A recipe popularized by Scott Rao and Matt Perger, widely implemented via the Barista Hustle Two Bottles method at 76 ppm GH, 50 ppm KH. The 1.5:1 GH:KH ratio is the lowest of any recipe on this list, meaning it buffers acidity more aggressively. (Note: the original Rao/Perger formula, as published by Jonathan Gagne, uses five ingredients and targets approximately 69 GH / 40 KH. The 76/50 Two Bottles version is a community approximation that’s close but not identical to the original.)
Mineral profile: Moderate hardness, relatively high alkalinity. This is the “rounder” recipe — less brightness, more sweetness and body.
Flavor profile: Balanced, smooth, forgiving. The higher alkalinity tames aggressive acidity, making this the most versatile recipe across roast levels. Darker roasts taste sweeter. Light roasts lose some of their sharpest fruit notes but gain approachability.
Best for: If you drink a range of coffees — light through medium-dark, filter through immersion — and want a single water that works for all of them, this is it. The coffee community’s closest thing to a universal recommendation.
6. Hendon — The Brightness Maximizer
What it is: Based on the center of the “ideal zone” from Christopher Hendon’s 2014 research paper. 100 ppm GH, 31 ppm KH. The 3.2:1 GH:KH ratio is heavily tilted toward extraction power with minimal buffering.
Mineral profile: High Mg-driven hardness, low alkalinity. Lots of mineral extraction agents, very little acid neutralization.
Flavor profile: Bright, complex, forward. Acidity is prominent and multi-layered. Sweetness develops in the aftertaste. Can be too much for medium or dark roasts — the aggressive extraction paired with low buffering can produce sour or harsh cups.
Best for: Origin-focused pour-over with light roasts. Competition-style brewing where you want maximum flavor development. Ethiopian naturals, Kenyan AA, washed Panamas — coffees with big, complex acid profiles.
7. Third Wave Water Classic — The Convenience Play
What it is: Pre-measured mineral packets. Pour one into a gallon of distilled water, shake, done. No weighing, no concentrates, no math.
Mineral profile: About 150 ppm GH, about 40 ppm KH. Uses magnesium sulfate, calcium citrate, and sodium chloride. The Classic profile hits SCA-range TDS with very high hardness and moderate buffer — a 3.8:1 GH:KH ratio.
Cost: About $1.40-1.50 per packet. Plus $1.00-1.50 for a gallon of distilled water. That’s roughly $2.50-3.00 per gallon, or about $0.65 per liter. Significantly more expensive than DIY concentrates.
Difficulty: Trivial. Rip open a packet, dump it in, shake.
Flavor profile: Intense acidity, high clarity. The very high GH:KH ratio and calcium citrate create a bright, punchy cup. Some people find it aggressive or overdone, especially with medium roasts. Excellent with very light roasts that can handle the extraction power.
Chloride warning: The Classic profile contains sodium chloride. If you’re running it through an espresso machine’s boiler, the chloride can corrode stainless steel and copper over time. TWW makes a separate Espresso profile for this reason.
Best for: People who want good water without buying a scale or making concentrates. Light roast filter drinkers. Travel (packets are tiny and portable). You can find Third Wave Water on Amazon or at their website.
8. Third Wave Water Espresso — Boiler-Safe and Smoother
What it is: TWW’s espresso-specific formula. Same packet convenience, different mineral blend. About 135 ppm GH, about 80 ppm KH. The 1.7:1 ratio doubles the buffer compared to Classic.
Mineral profile: Swaps sodium chloride for potassium bicarbonate. More alkalinity means more acid buffering, which also means the water is safe for espresso machine boilers (no corrosive chlorides).
Cost: Same as Classic — about $0.65 per liter all-in.
Flavor profile: Smoother, rounder, less aggressive than Classic. The doubled buffer tames acidity and adds perceived sweetness. Still plenty of extraction power from the high GH.
Best for: Espresso machines, specifically. Also works well for darker roasts in any brew method.
The Practical Science: Why These Recipes Taste Different
Three variables explain nearly all the flavor differences between these recipes.
Magnesium drives brightness. Hendon’s research showed that Mg2+ ions bind selectively to desirable organic acids — citric, malic, lactic — the compounds responsible for fruit, floral, and bright notes. More magnesium means more of those compounds end up in your cup. The sweet spot is roughly 10-30 mg/L of actual Mg (which corresponds to much higher ppm values when expressed as CaCO3 equivalent).
Calcium adds body and sweetness. Ca2+ extracts a broader spectrum of compounds without strong selectivity. It adds mouthfeel and chocolatey/sweet notes but can mute delicate origin flavors. Third Wave Water’s use of calcium citrate is notable — the citrate anion adds some perceived brightness compared to calcium chloride or calcium sulfate.
Bicarbonate controls how much acidity you taste. Every recipe’s KH value determines how much of the coffee’s natural acids survive into the cup. At 12 ppm (Melbourne), almost nothing is buffered — you taste everything the bean produces. At 80 ppm (TWW Espresso), significant acid is neutralized, producing a smoother, sweeter cup.
The GH:KH ratio is the quick heuristic. Higher ratio (3:1 and above) means brighter and more acidic. Lower ratio (1.5:1) means rounder and smoother. The recipes above span from 1.5:1 (Rao/Perger) to 3.8:1 (TWW Classic). RPavlis breaks the pattern entirely — zero hardness, all buffer.
Which Recipe for Which Brew Method
Pour-over (V60, Chemex, Kalita): You want lower alkalinity here. The SCA’s 7.3x rule means alkalinity hits 7.3 times harder in a 15:1 pour-over ratio than in a 2:1 espresso ratio. Recipes with KH under 40 work best. Holy Water for light roasts, Barista Hustle #4 for versatility, Hendon if you want to push brightness.
Immersion (French Press, AeroPress, Clever Dripper): More forgiving because immersion brewing is inherently less sensitive to water variables than percolation. Rao/Perger is the safest bet. Barista Hustle #4 if you want a bit more clarity.
Espresso: Alkalinity tolerance is much higher (up to 150 ppm without muting acidity), but you must avoid chlorides that corrode boilers. TWW Espresso for convenience, RPavlis for simplicity and zero scale risk, Rao/Perger via the Two Bottles method for flavor complexity.
Cold Brew: Longer contact time (12-24 hours) without heat. Cold brew is already low-acid, so you can tolerate higher alkalinity (50-60 ppm) without the cup going flat. Rao/Perger works well. Avoid very low-mineral recipes like Melbourne — the long extraction time can’t compensate for weak extraction agents at cold temperatures.
Cupping: Use SCA Standard targets (68 GH, 40 KH) for evaluation consistency, or Melbourne if you want to taste the bean with minimal interference.
What to Buy and Where
Every Two Bottles recipe (Melbourne, Holy Water, BH #4, Rao/Perger, Hendon) uses the same two ingredients:
Epsom salt (magnesium sulfate heptahydrate, MgSO4·7H2O): Any pharmacy or grocery store. A 4 lb bag is $5-8 and lasts years. Make sure it’s plain — no lavender, no fragrance, no added ingredients. USP grade.
Baking soda (sodium bicarbonate, NaHCO3): Arm & Hammer from any grocery store. $1 per box, lasts forever.
Optional upgrade — Potassium bicarbonate (KHCO3): Use this instead of baking soda if you want to avoid adding sodium. Available on Amazon, $10-15 per pound. If using KHCO3 for Bottle 2, the concentrate is 2.00g per liter instead of 1.68g. The RPavlis recipe uses only this.
Distilled water: $1-1.50 per gallon at any grocery store or pharmacy. This is your base.
A 0.01g jewelry scale: $12-20 on Amazon. Required for making concentrates accurately. A $10 kitchen scale won’t have the precision you need for 1.68g of baking soda.
Storage: Concentrates keep for 3-6 months at room temperature in sealed bottles. If you see cloudiness or sediment, make a fresh batch. The mixed brew water should be used within a week or two — it won’t go bad, but CO2 exchange with air slowly shifts the pH.
Third Wave Water
Available at thirdwavewater.com and Amazon. About $15 for a 12-pack (Classic or Espresso profile). They also make a “Dark Roast” profile with very low hardness and high alkalinity (about 135 ppm KH) designed to tame bitterness in dark roasts. Each packet treats one gallon.
Lotus Water Drops
Four separate mineral concentrates (Ca, Mg, Na, K) for fully independent control over every parameter. More expensive and more complex than TWW, but more versatile. Endorsed by Scott Rao, who found non-linear flavor effects — cups at 10 KH and 70 KH both “popped” more than moderate 30-50 KH values. Available at lotuscoffeeproducts.com.
The Lazy Alternative: Bottled Water
If mixing concentrates sounds like too much work, certain bottled waters land close to coffee-friendly targets:
| Bottled Water | TDS (ppm) | Notes |
|---|---|---|
| Crystal Geyser | 40-100 | Varies by source. The Alpine Spring source (California) runs about 40 ppm. Low, but usable. |
| Volvic | about 130 | French volcanic source. Moderate mineral content. Widely recommended in the coffee community. |
| Trader Joe’s Mountain Spring | 60-80 | Decent option at low cost. |
Avoid: Evian (about 309 ppm), Fiji (about 210 ppm), San Pellegrino (about 850 ppm, carbonated). All over-mineralized for coffee.
The catch with bottled water: inconsistency. TDS can shift between production runs, and you can’t control the mineral ratios. A gallon of distilled water plus a TWW packet gives you the same result every time. Bottled water is the backup plan, not the system.
Starting Recommendations
Brand new to water chemistry? Start with RPavlis. One ingredient, zero risk of scale, genuine improvement over tap. Live with it for two weeks, then try Rao/Perger to hear what hardness minerals add to the cup.
Filter brewer who drinks light roasts? Holy Water or Hendon. The low alkalinity and moderate-to-high Mg will bring out everything those beans have. Pair with good grind size technique for maximum results.
Espresso machine owner? TWW Espresso for convenience, or RPavlis for simplicity and guaranteed boiler safety.
Want one recipe for everything? Rao/Perger. It’s not the most exciting water at any extreme, but it’s the most consistently good across methods and roast levels.
Don’t want to make anything? Volvic from the grocery store. It’s not perfect, but it’s better than most tap water and it’s consistent.
The beauty of the Two Bottles method is that switching between recipes costs nothing — you already have both concentrates. Pour different amounts, get different water. Try Hendon for your weekend pour-over and Rao/Perger for your weekday AeroPress. The ingredients are pennies. The experiments are free.
Some links above are affiliate links. If you purchase through them, we earn a small commission at no extra cost to you.
Frequently Asked Questions
- How long do DIY coffee water concentrates last?
- Concentrate bottles (Epsom salt solution and baking soda solution) keep 3-6 months at room temperature in sealed containers. If you see cloudiness or sediment forming, make a fresh batch — it takes 60 seconds. Mixed brew water is best used within one to two weeks. It won't spoil, but CO2 exchange with air gradually shifts the pH and can subtly change extraction behavior.
- Is Third Wave Water worth the price compared to making your own?
- Third Wave Water costs roughly $0.65 per liter (packet plus distilled water). DIY concentrates using the Barista Hustle Two Bottles method cost about $0.27 per liter (almost entirely the distilled water). The mineral cost itself is under a penny. TWW's advantage is pure convenience — no scale, no concentrates, no math. If you brew a liter a day, the price difference is about $140 per year. For most home brewers, TWW is worth it for the first month while you decide whether water chemistry matters enough to invest in a jewelry scale and two bottles.
- Can I use the same water recipe for both espresso and pour-over?
- You can, but one method will be suboptimal. Alkalinity hits 7.3 times harder in pour-over than espresso due to the water-to-coffee ratio difference. A recipe with 50 ppm KH works beautifully in espresso but may mute acidity noticeably in a V60. If you brew both methods, Rao/Perger (76 GH, 50 KH) is the best single compromise. If you want to optimize each, use a lower-alkalinity recipe (Holy Water or Hendon) for pour-over and RPavlis or TWW Espresso for your machine.
- What's the difference between Third Wave Water Classic and Espresso profiles?
- The Classic profile uses sodium chloride (table salt) as part of its mineral blend and has low alkalinity (about 40 ppm KH). The Espresso profile replaces the sodium chloride with potassium bicarbonate, doubling the buffer to about 80 ppm KH. Two practical differences: First, the Espresso profile is boiler-safe because it contains no chlorides, which corrode stainless steel and copper under heat. Second, the higher buffer produces a smoother, sweeter cup with less aggressive acidity. Use Classic for filter brewing with light roasts, Espresso for machines and darker roasts.
- Do I really need a jewelry scale, or can I use a kitchen scale?
- You need a scale that reads to 0.01g. The Barista Hustle hardness concentrate calls for 2.45g of Epsom salt — a standard kitchen scale that reads to 1g increments can't hit that accurately. A 0.01g jewelry scale costs $12-20 on Amazon and doubles as a coffee dose scale. You only weigh the minerals once when making concentrates (which last months), so this isn't a daily inconvenience. The RPavlis recipe (0.38g potassium bicarbonate) is even more scale-dependent — 0.38g is invisible to a kitchen scale.