Aqua Focus 

Water treatment supplies, services and consultancy

 

Ion Exchange Softened Water and Domestic Central Heating Systems

Summary

There is ongoing conflict about the suitability of filling the domestic central heating system with ion exchange softened water. Historically, this emanates from the BSi code of practice BS7593, which originated in 1992 (revised in 2006), stating that naturally soft and base-exchange softened water have an increased potential for corrosion and the primary circuit should not be filled with softened water unless a corrosion inhibitor specifically formulated for the purpose is added. And that softened water should not under any circumstances be used where a boiler contains aluminium.

This misrepresentation of softened water being more corrosive due to the softening process originated because it is incorrectly associated with naturally soft water. Hard water which has been softened, and naturally soft water, are very different. 

This has led to confusion and conflict for more than 20 years about the suitability of base exchange softened water for filling the domestic central heating system, alleging that softened water could result in corrosion and failure of the heating system – particularly for aluminium boilers.

The allegation is not supported by scientific consideration of the water quality parameters, field evidence or laboratory testing but is based on commercial pressure. 

British Standard BS 7593

BSi Committee Cii/62, Treatment of Water for Boilers, is currently revising BS7593:2006 - Code of Practice for the preparation, commissioning and maintenance of domestic central heating and cooling water systems. The current revised draft states that, if a water softener is installed in the building, the heating system should be filled with unsoftened water. It is therefore more restrictive than the current version.

The proposed revision conflicts with current market practice in that:

  • some boiler manufacturers state that the water should besoftened above a certain hardness level
  • some inhibitor manufacturers state that their products are suitable for softened water
  • government guidance for compliance with Building Regulations on domestic central heating systems states that the water should be treated if the hardness is above 200 mg/l and lists softeners as an option
  • British Gas advertise a boiler for use in homes with a water softener.
  • a major UK water softener manufacturer offers a guarantee against boiler corrosion with their softeners.

Inclusion of this requirement is not only an unjustified negative for the softener industry, but also contrary to the BSi principles (BS O) in that it is prescriptive by excluding products, that are currently validly on the market, ignores regulatory guidance and inhibits innovation.

It was not raised by any European delegates of CEN/TC164/WG13 when the EN14743 for Water Softeners (requirements, for performance, safety and testing) was drafted in 2003 (or its subsequent revisions).

In the current BS 7593 drafting committee, a sustained objection has been overridden by the majority. This means that there is not consensus within the committee which is a prerequisite and foundation of the standards drafting process before release forDPC (Draft for Public Comment). But is has now been released for public consultation with 3 proposed options (see http://standardsdevelopment.bsigroup.com/projects/2017-03900).

 

Softened water and naturally soft water.

 The main constituents of hard water are calcium and magnesium together with bicarbonate, sulphate, chloride, nitrate. When the water is passed through an ion exchange water softener, the calcium and magnesium ions are replaced by sodium ions; the anions (bicarbonate, chloride, sulphate, etc), and hence the alkalinity, remain unchanged. 

Naturally soft water, on the other hand, contains very little dissolved solids; its pH is often low and buffering capacity is negligible. Hence, naturally soft water is often very corrosive to metals.

So ion exchange softened water has not undergone any change which significantly affects the parameters which might increase corrosivity[1], such as dissolved oxygen, pH, temperature, conductivity, chloride level, sulphate level, etc.  

Historical Evidence.

Water softeners have been used for many years. The technology dates back to 1905 when Robert Gans first patented the process using naturally occurring zeolites. Softened water was necessary to protect steam engines against scale deposition in the engine boiler tubes. Similarly, water softeners have been used domestically since 1916. Polystyrene/DVB resins were developed in the mid 1940s. For over 100 years, no specific mention, concern or requirements relating to an increase in corrosivity have been found necessary for domestic or industrial water softening applications.  

The water softener industry, through trade associations such as the EWTA, UAE, Aqua Belgica, UKWTA and WQA, is alert to any operational issues which water softening may incur. Throughout the world, there is not one situation, of which these associations are aware, where water softening has been the proven cause of a corrosion failure. It should be borne in mind that the WQA (Water Quality Association, USA), with more than 2,400 members, represents about 50% of the world water softener market. 

The Misconceptions.

 Some boiler manufacturers are under the misapprehension that softened water is more corrosive. Situations have been encountered where a heating system has failed and the boiler manufacturer has negated its warranty because a water softener was fitted to the premises. In all of these situations that have been investigated, the failure was eventually attributed to a cause other than softened water. Harvey Water Softeners Ltd has cooperated with Worcester Bosch to investigate 18 instances of boiler failure on premises where a water softener was installed and in all 18 cases the cause of failure was not attributable to softened water. 

In addition, indiscriminate application of somewhat ageing prediction tools such as the Langelier Saturation Index, without taking account of more recent advances in the understanding of the factors that influence corrosion, can lead to misdiagnosis of corrosive conditions. The Langelier Index was developed to determine whether a water source is potentially scale-forming or scale-dissolving (aggressive). Although it was not intended to be applied to softened or naturally soft water, its hypothesis was that a thin, uniform layer of scale deposition can, under carefully controlled conditions, contribute to a protective barrier.

However, in normal household plumbing applications, widely variable physical and chemical conditions render a deposit that is very variable in thickness, porosity, morphology, location, etc, such that reliable prediction of a beneficial effect is not possible. Deposition is usually restricted to the hottest parts of the boiler. But, the scale can in fact exacerbate the corrosion problem in that formation of crevices, and/or bacteriological activity, can cause accelerated localised corrosion with consequent failure and property damage. 

 Laboratory Evidence.

Tests have been conducted which have compared the corrosivity of hard water before and after softening. A study by the USEPA (United States Environmental Protection Agency) was undertaken between 1994 and 1996 comparing the corrosion and corrosive properties of hard water with that after softening by ion exchange[2]. The results showed no pattern of higher leaching from lead, copper, brass and galvanised materials nor significant deterioration in the water quality parameters that influence corrosion rate. Similar results were obtained from tests in Belgium (click here for results of Belgian study http://www.aquabelgica.be/files/ABE-Pos_Paper-ENG-Copper_corrosion-Rev20070606.pdf).

In 2012, the UKWTA commissioned BSi to carry out a corrosion test on a simulated central heating system which concluded – “over the life time of a modern central heating system, filling with base exchange softened hard water mains would appear to present no significantly greater risk of corrosion of system metals than filling with hard mains water” (click here for a copy of the report http://www.ukwta.org/assets/NewFolder/BSi-Corrosion-Report-12.12.12.pdf).

Expert opinionhas been sought internationally. NSF (National Sanitation Foundation, USA) provides an international testing and certification service for water treatment products and food. When questioned by WRc-NSF, its joint venture company in the UK, as to their awareness or possible concern about potential for increased corrosivity of softened water, its response was that, in their international experience, they are not aware of any such issue and it would be addressed in product standards if they were necessary.  

In the UK, corrosion companies and experts who have investigated boiler failures state that from their experience they have “never seen boilers fail due to the use of softened water and that this is supported by theory”.

 Chemical Inhibitors.

 Water treatment with a corrosion inhibitor is often advisable or necessary depending upon water quality parameters which are known to affect corrosion. In such circumstances, an appropriate inhibitor should be selected. Some inhibitors may increase the corrosivity of softened water and so selection should be based, not only on whether an inhibitor is required but whether the inhibitor is suitable for the quality of the water being used. 

 In the UK, chemical inhibitors are tested and certified by the BuildCert scheme which is operated by WRcNSF (link to their website). The test process involves aeration procedures which result in displacement of carbon dioxide from the water and consequent rise in pH. This is significant with respect to aluminium as most aluminium components in a heating system will begin to corrode at a pH above 9. This evolution of carbon dioxide will not occur in a closed loop heating system because the carbon dioxide gas cannot escape (even in a vented tank system, the hot water is under the pressure head below the tank and the carbon dioxide is held in solution). Arguably, this challenges the representativeness of the test, but, even if it is over-excessive, an inhibitor which passes is more than capable of meeting the real-life conditions. 

 Definitive Experience.

The predominant cation and anion in ion exchange softened water are usually sodium and bicarbonate, respectively, and sodium bicarbonate is used in closed circuit systems to assist in corrosion control by buffering (stabilising) the pH between 8.0 and 8.5.

 Moreover, pitting corrosion of copper in hot water systems can be caused by localised deposits of iron, manganese or aluminium impurities in the water[3]. Ion exchange water softening will effectively remove these impurities at point of entry.  

The positions of European trade associations (UAE, Aqua Belgica, UKWTA) and international (WQA) are clear and based on extensive experience and expertise.

The German VDI 2035 Part 1 – Prevention of damage in water heating installations – scale formation in drinking water and water heating installations, states in clause “4.4.2 Water Heating Systems – Softening – Softening is the preferred method of avoidance of scaling, since it results in long term removal of alkaline earths (calcium and magnesium ions) from the system.”

CONCLUSION

Water softening by base exchange does not increase corrosivity and it is suitable for filling the primary circuit of a central heating system. Any inhibitor used should be suitable for softened water.

 

[1]Note: corrosivity and aggressivity are often confused; they are not the same. A corrosive water will attack metals and cause damage to plumbing systems. An aggressive water tends to dissolve scale (calcium carbonate) but is not necessarily corrosive.

[2]Sorg T. J., Schock M.R. and Lytle D.A., (April 1998). Leaching of Metals from Household Plumbing Materials: Impact of Home Water Softeners, USEPA.

[3]Harfst F.H., (1998). Corrosion in copper pipes can be prevented., Water Technology, Vol. 22 No. 3. 

Back to top