Why Kratky setups fail often gets decided before planting, when you fill the tank with tap water. Bernard Kratky of the University of Hawaii gives this example (Acta Horticulturae 843, 2009): water with an EC of 0.5 mS/cm from its own impurities reaches 2.0 once only a quarter of the solution is left. The volume dropped fourfold, but the salts stayed. By the end of the cycle, the lettuce is sitting in a solution strong enough to cause tipburn.
That’s the first of six Kratky method problems, and each one is described in Kratky’s own papers from 2004, 2005, and 2009. Below are all six, from salts in the water to mosquitoes in the tank, with numbers from his trials and a way around each one. How the method itself works, and why oxygen roots matter, is covered in the wiki entry Kratky method.
Failure 1: salts from the water concentrate by the end of the cycle
In a Kratky system, salts from the water don’t go anywhere: the plant drinks the water, and the impurities stay in the tank. In his 2004 paper, Kratky works through a harsher case: water with an EC of 0.2 mS/cm reaches 2.0 when 10% of the solution is left. In regular hydroponics, draining or flushing takes care of this, but here the solution isn’t changed until harvest.
What Kratky himself recommends:
- Where water quality is poor, use rainwater.
- If rainwater isn’t an option, use a deeper tank: in a larger volume, salts concentrate less.
- With poor water, drain the tank after every crop.
Before you fill the tank, measure the EC of the water itself. If it’s already 0.5 or higher, account for that in your recipe and in how far down you let the tank run. Why a water test comes first in any system is explained in the wiki entry water in growing.
Kratky is more relaxed about pH. With commercial hydroponic fertilizers and Hawaii’s water, crops usually finished without any pH adjustment. If the solution does turn too acidic, he suggests hanging dolomite in a nylon stocking in the tank until pH recovers. The stocking keeps undissolved residue off the bottom. The pH range to aim for is in the wiki entry pH in hydroponics.
Failure 2: topping off mid-cycle
In the Kratky method, the solution level can stay put or drop, but it can’t go up. Following the Japanese researcher Imai, Kratky calls the roots above the solution oxygen roots: they breathe the moist air under the lid. Topping off puts them underwater, and the plant, in his words, “drowns.” His 2004 paper gives a specific threshold: once roots have formed, raising the level by more than 2 cm (about 3/4 inch), whether by topping off or by rain, can injure the plant.
Between crops, the rule changes. Kratky notes that some growers drain the tank after harvest, while others top off the remaining solution with fresh solution and replant right away. His general rule: drain and refill the tank at least every three crops. With poor water, as in failure 1, drain it after every crop.
Failure 3: rain raises the level and dilutes the solution
For an outdoor Kratky setup, rain does what topping off does, and it dilutes the solution too. The clearest example is his 2005 tomato trial, where 62 inches (158 cm) of rain fell during the crop. A tank under a simple polyethylene rain shelter produced 3.9 lb (1.75 kg) of marketable fruit per plant; the same tank with no shelter produced 1.7 lb (0.78 kg), 55% less. Unsheltered 2-gallon (7.6 L) pots produced no marketable fruit at all: each pot caught about 15 gallons (57 L) of rain, and the solution was diluted.
So keep an outdoor Kratky setup under a rain cover or in a greenhouse. Lettuce grown this way is normally raised in screened, plastic-covered rain shelters or greenhouses, Kratky writes.
Failure 4: the solution runs out before harvest
All the solution for the crop goes in up front, so volume decides which crops the method can carry at all. For lettuce, Kratky gives 3-6 L (about 0.8-1.6 gallons) of consumption per plant, and a 7-ounce (200 g) head in Hilo drank 3-4 L. That’s why a 4 L (1-gallon) bottle only suits crops that need less than 4 L over their whole life. The crop is finished when less than 10% of the solution is left in the tank.
Tomatoes are a different story. A pound of fruit takes 3-4.8 gallons of water (25-40 L per kg). In a single-fill trial, it came to 3.7 gallons of solution per pound (30.5 L per kg), and Kratky called this version an academic curiosity for tomatoes: a commercial tank would have to be orders of magnitude larger.
He did grow tomatoes without a pump, but with topping off. The level was maintained: 3 inches (7.5 cm) at transplanting, then under 2 inches (5 cm) until the end of the trial. In one setup, a float valve fed from a separate tank did this. ‘Big Beef’ in 12-ounce (0.35 L) beverage cans produced 5.9 lb (2.68 kg) of marketable fruit per plant over a 72-day harvest. A 4-inch (10 cm) pot of perlite sitting on an upside-down 2-inch (5 cm) nursery tray outyielded a can on the tank floor: 7.8 lb vs 5.5 lb (3.55 vs 2.49 kg) for ‘Healani’. On the tray, more of the roots end up in moist air. Strictly speaking, this is no longer “fill it and forget it,” but it doesn’t need electricity either.
For choosing a system by crop, there’s a separate article: NFT vs DWC vs Kratky vs drip.
Failure 5: the lid rests on the solution
A lid floating directly on the solution takes the air gap away from the roots. In trials where the lid floated without support for the whole crop, lettuce yields were up to 23% lower than with a lid resting on the tank frame. The gap has to grow as the level drops, so the lid either rests on the frame or settles onto supports over time.
The material matters too. Kratky advises against expanded (bead) polystyrene for a floating board: it has occasional air pockets, gets waterlogged, and grows algae. Extruded polystyrene has a consistent structure and doesn’t get waterlogged. And in the tomato trial, wind moved unsecured boards together with the plants, which put the roots at risk. Outdoors, fasten the board down.
Failure 6: light and standing water
Light on the solution means algae, so Kratky recommends covering or darkening the bottle, and in his trials the tanks were lined with black polyethylene. If algae are already there, the wiki entry algae in hydroponics covers how to get rid of them.
Standing, unaerated water attracts mosquitoes. At 4,300 feet (1,300 m) in Hawaii it was cool, and they were barely a problem. At lower, warmer elevations, the tanks became breeding grounds. Kratky’s working fix: stretch window screen above the solution. At transplanting, the level sits above the screen so the growing medium gets wet, then it drops below it. Newly hatched mosquitoes are trapped under the screen and die. Other options from his paper: salt-tolerant fish, or Bti (Bacillus thuringiensis israelensis) products. Cool weather and dry wind also hold mosquitoes back. A tight lid with holes sized to the pots keeps them out.
Warm standing water carries one more risk: root rot. That’s a question of temperature rather than of the method, so it’s covered in the wiki entry on Pythium.
Why your Kratky setup failed: a quick check
| What you see | Likely cause | What to do |
|---|---|---|
| Leaf edges burn near the end of the cycle | Salts from the water concentrated | Measure your water’s EC, use rainwater, drain the tank after every crop |
| Plant struggles after topping off or heavy rain | Oxygen roots underwater | Don’t top off mid-cycle, add a rain cover |
| Solution ran out before harvest | Too little volume for the crop | Plan 3-6 L (0.8-1.6 gal) per lettuce; for tomatoes, a float valve |
| Lower yield, lid resting on the water | No air gap | Lid on the frame or on supports |
| Green film in the tank | Light reaching the solution | Opaque container, black liner |
| Mosquito larvae in the solution | Open standing water | Screen above the solution, tight lid |
Get these right, and the method delivers what it promises: in Kratky’s trials, lettuce commonly took less than 2.4 gallons of water per pound (20 L per kg), and the best result was 1.3 gallons per pound (11 L per kg).
