Water softeners: which one lives up to its certificate?

The figures in this matrix rest on manufacturer datasheets and retailer listings; each value carries where it comes from and when we retrieved it. A “DVGW ✓” appears only where the manufacturer documents evidence — otherwise “in Prüfung” (under review). The sort order answers your input, not our affiliate links.

How to read this matrix

This comparison puts 18 systems for single- and multi-family homes side by side along six dimensions that decide the purchase: DVGW evidence, capacity, salt consumption, control, warranty and price. Set the filters to your situation — household size, form factor, DVGW — and the field shrinks to the devices left for you. Three things to know before you judge any number:

Capacity beats headcount. “For 4 people” is a rule of thumb; the reliable measure is the capacity in m³×°dH. It states how much water of a given hardness the system can soften before it must regenerate. Only the Grünbeck and BWT systems state this value to the standard (per DIN EN 14743); on several LEYCO and SYR models it is missing — there the entry reads “k. A.” (no data) instead of an estimated number.

The DVGW check mark has tiers. A test mark with a verifiable registration number carries more weight than the bare claim “DVGW-certified” without a number. Where a manufacturer does not document device-level evidence, we show “under review” — in this round that also applies to individual BWT models where only the supplied connection set is verifiably tested.

Salt consumption is a running cost. The value per regeneration co-determines the operating costs. It spans widely across the field — from a few grams to around 2.3 kilograms per cycle — and repays a close look, because it adds up over the years.

Do you need a water softener at all?

This question deliberately comes before choosing a device, because the honest answer is often “no”. Stiftung Warentest considers filter and softening systems for drinking water in many cases “Meist überflüssig” (“mostly unnecessary”) — the very title of its article (test.de, retrieved 13.07.2026). The reason: hard, calcareous water is considered harmless to health — the German Environment Agency sees no health drawback in high water hardness (umweltbundesamt.de, retrieved 13.07.2026); whether the limescale noticeably strains your appliances depends on the local hardness band.

Your water utility knows that band and must tell you: under Section 9 of the German Detergents Act (WRMG), hardness is divided into three bands — “soft” up to 8.4 °dH, “medium” from 8.4 to 14 °dH and “hard” above 14 °dH; the utility publishes the applicable band at least once a year. As a rough rule: in the soft and medium bands, limescale rarely does enough damage for the purchase, salt and servicing to pay off. Only in the hard band — and the higher the hardness, the more clearly — does the calculation shift: less scale in the boiler, washing machine and fittings, less detergent, less wear.

Honestly, it remains a trade-off, not a buying command: whoever fits out a new house with expensive building services at 25 °dH thinks differently from a tenant at 9 °dH. This page helps you find the right device once you have decided on a system — it does not push you into a decision your hardness band may not even support.

From water hardness to the system: the calculation chain

Whoever plans a system goes through four steps at the core — and it is exactly this chain that hardly any of the usual guides calculates through completely:

1. Determine the hardness. Your utility’s °dH value is the starting point. It determines everything else.
2. Dimension the capacity. Hardness times daily consumption tells you how quickly the system uses up its capacity (in m³×°dH) and must regenerate. A system chosen too small regenerates constantly with hard water; a generously sized one runs calmer and saves salt per litre.
3. Estimate salt and water demand. More frequent regeneration means more salt and more rinse water. The salt consumption per regeneration is in the matrix — multiplied by the regeneration frequency, it yields the running costs.
4. Check the sodium limit. In ion exchange, limescale leaves the water and sodium enters it. The drinking-water limit for sodium is 200 mg/l, and per the figawa technical bulletin (dated 05/2020), every degree of hardness removed raises the sodium content by around 8.2 mg/l. Softening from 20 °dH down to 0 °dH would therefore mean roughly +164 mg/l of sodium — close to the limit before the source water even counts.

That is why a correctly configured system never targets 0 °dH but blends: it mixes hard raw water back in until a target hardness (often around 8 °dH) is reached. This spares not only the sodium budget but also salt and resin. Blending is thus the point where technology, costs and drinking-water law converge — and the real answer to the question “how far may I soften?”.

DVGW or only “UBA-compliant”: the trap in the model name

A test mark is not marketing garnish but a statement of fact — and not all systems carry it through equally cleanly. Nowhere is that clearer than within a single brand: LEYCO offers both DVGW-certified and merely UBA-compliant models, and the difference hides in the name suffix.

The LEYCO Soft 15 and LEYCO Soft Pro 15 models are listed by the retailer with a DVGW certificate (Klimaworld product page, retrieved 13.07.2026). The visually almost identical LEYCO Soft One 9 and Soft One 15, by contrast, carry per the retailer only the note “UBA-compliant” — the product slug even literally reads “uba-konform” — and no DVGW evidence. For the buyer this means: two systems in the same housing, in a similar price region, but with different depths of testing. Anyone who takes the “One” suffix for a mere trim level may overlook that the certification tier changes here.

This is not a LEYCO peculiarity but the pattern: on the five Grünbeck models, the DVGW registration number is printed in the datasheet and thus verifiable — the most robust evidence in the field. On several BWT systems (Perla one, AQA life S, AQA Perla Duplex), by contrast, we found no device registration number in the manufacturer’s datasheet; there, only the supplied connection set is verifiably DVGW-tested, which is why the device check mark stays at “under review” until the manufacturer provides the evidence. And the model positioned as the budget/entry device in the field, the Water2Buy W2B800, carries no DVGW evidence at all — a point that puts any price advantage into perspective.

What buyers report

Unlike with everyday products, the review situation here is thin — and saying so honestly is part of the comparison. Water softeners are mostly sold through specialist trades and niche retailers, not through review-heavy marketplaces. For only two of the 18 systems is there any product-level review with a traceable source at all (as of 13.07.2026):

• The JUDO i-soft SAFE+ stands at 5.0 out of 5 stars at Selfio — however, from just two reviews. The direction is positive; for a verdict the base is too narrow, and we state the thin foundation explicitly.
• The Water2Buy W2B800 reaches around 4.4 out of 5 stars on Amazon.de. The exact count could not be cleanly determined because Amazon blocks automated retrieval — the order of magnitude is consistent across several search results, the exact number of reviews remains open. This good product score, however, faces a weak vendor rating: the manufacturer Water2buy scores only 2.2 out of 5 stars on Trustpilot (387 reviews, rated “Poor”, as of 13.07.2026). That rates the shop and the brand, not the individual device — anyone considering the budget/entry device in the field should weigh any price advantage against the weak vendor reputation.

Two verified product ratings do not make a ranking, and we do not construct stars from a thin base that do not exist. For the purchase route, the retailer reputation is more tangible: the shop behind most purchase links (Klimaworld) stands at around 2.5 out of 5 on Trustpilot (120 reviews), the alternative retailer Selfio at around 4.0 out of 5 (606 reviews), each retrieved 13.07.2026. These are shop ratings, not product verdicts — but a consideration when you choose between two sources for the same system. The technical values in the matrix remain unaffected; they come from the datasheets, not from the shop.

Purchase links are affiliate links and have no influence on the measurement. We link the source with the documented price; where several retailers carry the same device, the matrix names the price recorded in each case, with its as-of date.