12 Litres a Week Saved by a Blumat Drip System on a Raised Herb Bed

I set up two raised beds side by side, each 1.2 by 2.4 metres, and filled both from the same batch of peat-reduced potting mix cut with 30 percent horticultural grit. One bed ran off a Blumat Classic gravity system fed by a 30-litre food-grade barrel raised 60 cm on breeze blocks. The other I watered from a can filled at that same barrel, applied by eye whenever the top few centimetres looked dry. Across eight weeks in the hottest part of the season, the barrel readings held steady, and the hand-watered bed used 12 to 14 litres more per week.

The reason was simple enough once I watched it happen. Hand-watering overshot the bed’s holding capacity. A can lays water down in short, heavy bursts, and the grit-heavy mix could not take it up fast enough, so part of every pour ran straight through the profile and out the base. The Blumat line, by contrast, released water only as the ceramic cone dried against the surrounding soil. The supply tracked soil tension instead of what the surface happened to look like.

Placement mattered more than the kit

The Blumat Classic works through soil tension. Each ceramic carrot gets filled with water, capped, and pressed into the soil near a plant’s root zone. As the soil around it dries, it pulls water through the porous ceramic, the pressure inside the sensor drops, and a small valve on the drip line opens. When the soil wets back up, the valve closes. The whole thing runs on physics with nothing to plug in.

Most of the gap between a steady bed and a wasteful one came down to where I put the cones. A carrot set too shallow reacts to drying that the deeper roots have not felt yet, so the line waters too often. A carrot set too deep in a bed of shallow-rooted basil and thyme stays wet long after the top 5 cm has baked hard.

On the test bed the cones sat at 8 to 10 cm deep, angled slightly, and roughly 15 cm from each plant stem. Two sensors covered the 2.4-metre run because the soil tension moved sideways through the uniform mix.

The adjustment screw on the sensor head sets the trigger point. Fully open, the valve flows at the slightest dryness. In the fast-draining grit mix I left the screw about two-thirds open. A denser loam bed would have wanted something closer to a quarter open, otherwise the system keeps chasing surface evaporation it should ignore.

Grit changed the result before the water reached the roots

The 30 percent grit was the condition that made the bed behave. Plain multipurpose compost holds water like a full reservoir and can stay saturated for days, which starves roots of oxygen and rots the base of Mediterranean herbs. Grit opened air channels and let surplus water clear the root zone within minutes.

That same drainage is what exposed the watering can. Delivered in one pour, much of the water sheeted through the coarse structure before the finer compost fraction could wick it sideways. The slow drip gave the compost time to pull water horizontally out from each emitter, wetting a wider column with far less loss.

Terracotta pots around the same area behaved in the opposite direction. Terracotta breathes and draws moisture out through the clay wall, so a pot with a very free-draining mix dries much faster. In pots I held back a little on the grit, added a crock or fibreglass mesh disc over the drainage hole, and let the clay do most of the aeration work. A 25 cm terracotta pot with one central drainage hole and a grit-light mix held its moisture roughly a day longer than the same pot filled with a free-draining seed-and-cutting blend.

The self-watering planter told a different story

A self-watering planter with a built-in reservoir and wicking base gave me a useful point of comparison. The Lechuza-style insert held 4 litres, needed topping up every four to five days at peak, and gave me no real control over the moisture level.

It wicked until the reservoir ran dry, which suited thirsty tomatoes far better than drainage-loving herbs. Sage and rosemary sitting in a constantly damp wicking mix picked up the same kind of root risk the grit was there to prevent.

Feeding stayed off the drip line

I kept feeding separate from the water supply on purpose. The bed and containers ran on Osmocote Pro, a coated slow-release granule rated for five to six months. Worked into the top 5 cm at spring planting, it released nitrogen, phosphorus, and potassium across the season, with faster release in warm soil when the plants wanted more.

Herbs are light feeders, and a full-strength vegetable dose can scorch them, so the bed got half the label rate for containers, about 2 grams per litre of mix. Coated granules also sidestepped a problem that liquid feed would have caused on the drip system. Liquid fertiliser run through Blumat lines can leave mineral deposits inside the ceramic cones and the fine emitters. With the granules already in the compost, the supply line carried nothing but water.

Where the 12 litres went

The Blumat setup won on moisture control because soil tension governed every drop it delivered. Herbs that prefer to dry out between waterings still got their drying cycle, which the can could never guarantee. The self-watering planter was easier to live with, since it only wanted a weekly top-up and had no narrow line to clog, but it offered nothing like the same discrimination.

The drip system did ask for attention at the close of the season. One ceramic cone partially clogged with fine sediment, and I had to soak and back-flush it before the flow came back.

The saving really belonged to a specific set of conditions coming together: a fast-draining mix, sensors tuned to that drainage rate, and herbs that genuinely want a dry interval. Swap in a water-loving crop or a heavier soil and the water use shifts, maybe enough to erase the advantage. What I still cannot say is how the same two sensors would have performed in a dense loam bed, since I never ran that comparison, and the trigger point that suited my grit almost certainly would not.