Corrosion of water pipes: out of sight, out of mind?
Pipe corrosion is a major cause of water quality complaints by customers, but how often is the water itself responsible for causing it? MWH's Lisa Barrott explores the tests and conditioning for ensuring drinking water is non-corrosive
In a review of six large water companies in the UK, MWH asked how final pH targets are derived for a site, which water conditioning indices are used, how the targets are applied and what chemical methods of corrosion control are employed.
Two water companies had specific corrosion control policies; the others did not. Four out of the six had final water requirements for LSI, the other two did not include a corrosion control index in their final water quality requirements. One water company varied its pH target seasonally in response to the change in the alkalinity of its final water and to achieve a positive LSI.
What is the LSI?
The Langelier Saturation Index (LSI) is a calculated number used to predict the stability of water with respect to calcium carbonate. It indicates whether the water has potential to precipitate, dissolve, or be in equilibrium with calcium carbonate. The LSI is expressed as the difference between the actual system pH and the saturation pH (pHs). It is calculated from the pH, alkalinity, calcium ions, temperature and dissolved solids.
Corrosion control parameters
Three important parameters for corrosion control are alkalinity, buffering capacity and dissolved organic carbon. Alkalinity is a measure of the ability of a water to resist a pH change - poorly buffered waters are those with alkalinity levels less than 50 mgCaCO3/l. Buffering capacity is the ability of a water to resist a change in pH, at a given pH. Finally, dissolved inorganic carbon is the sum of carbonate, bicarbonate and carbon dioxide. All three parameters are linked and overlap and all three must be considered in an effective corrosion control strategy.
by Lisa Barrott, Senior Technical Specialist, MWH
Water companies now place their customers’ experience of their water supplies at the very heart of their business. The use of Outcome Delivery Incentives (ODIs) has incentivised improvements to water quality both at the treatment stage and at the customers’ taps. Bluntly put, it costs money if water quality degrades in the network.
The UK water industry has resolved to minimise water quality complaints, particularly the ‘red-water’ complaints linked to corrosion of pipes. Water companies routinely assess the impacts of flow and pressure changes in the networks on water quality, and invest large sums to reduce these impacts. But what about the effect of the treated water quality on the pipes that make up the networks and cause the discolouration in the first place? Could optimising final water chemistry help reduce complaints more efficiently through maintaining pipe integrity, reducing troublesome blockages and preventing expensive pipe failures? For years the Langelier Saturation Index (LSI) has been used widely to determine whether a water is corrosive. But just how useful is it?
Turbidity, the scattering of light caused by large numbers of individual particles often invisible to the naked eye, is a key indicator of water quality. Historically, water companies have carried out final water conditioning using lime-based dosing systems. More recently, a focus on disinfection and stringent final water quality targets have driven a shift towards using caustic or pure limewater. Water companies which dosed lime slurry struggled to achieve tight turbidity targets due to the introduction of particulates from lime slurry, and final water pH would not be optimised for corrosion control purposes. Conditioning of final water was effectively sacrificed for final turbidity compliance. This was perhaps understandable given the importance of turbidity as a surrogate for monitoring Cryptosporidium – something which is still very important today. The impact of corrosive water on the distribution side of the supply chain was to all intents and purposes ‘unseen’.
What do water companies do now?
In 2016 MWH reviewed the conditioning policies and strategies for conditioning of six large UK water companies (see box). All the water companies had a strategy to reduce lead, and compliance with the demanding lead standard of 10 µg/l has improved dramatically. There was a diversity of approaches with respect to other parameters, partly reflecting the quality of the source waters treated by the companies, but in some cases taking quite fundamentally different approaches to corrosion indices and pH. In one case final pH targets are changed seasonally, but for some companies there is no target pH. However of particular interest was the inconsistent use of the well-known corrosion index, the Langelier Saturation Index (LSI).
Is the Langelier Saturation Index still useful?
The use of the well-known LSI is very widespread, but has limitations, and some water companies are moving away from setting final water quality targets solely based on this index. The fundamental question has to be asked: why use an index based on calcium carbonate chemistry to control the corrosion of iron pipes? For many years it was thought that the precipitation of a thin layer of calcium carbonate would form a protective layer on the surface of the pipes and prevent corrosion of the pipes and the production of discoloured water. However, it is now well-established that the layer of calcium carbonate does not prevent corrosion. Where the LSI is undoubtedly effective is to predict the ability of the water to corrode structures such as cement-lined pipes, which contain calcium as an essential part of their structure and to prevent blockages in small pipes and mixers from calcium carbonate precipitation. Buffering of the water, as measured through the buffering intensity, and other mitigation measures to reduce dissolution of inner pipe surfaces and scale also need to be considered. The buffering capacity can be increased by raising the dissolved inorganic carbon through the addition of alkalinity in the form of bicarbonate/carbonate.
The recent catastrophic case of Flint Water in Michigan showed the importance of examining other corrosion indices. The water that caused the health problems in Flint had a reasonable LSI (based on the published data) but the big difference was in the chloride and sulphate levels. Consideration of the Larson ratio, based on chloride, sulphate and alkalinity, would have highlighted a potential water quality issue. The sole use of the LSI is not sufficient to prevent water quality problems, and ensuring a positive LSI does not always prevent corrosion.
Balancing water quality needs
Balancing all the water quality needs in hard, well-buffered waters tends not to be problematic. Many water companies dose ortho-phosphate for lead minimisation and in hard waters lead pipes become covered in lead carbonate and lead phosphate from phosphate dosing. The LSI tends to be positive as the pHs is lower than in poorly buffered waters.
It is expensive to make changes to pH of well-buffered waters and the pH requirements for chlorination ensure that the pH of the treated water at the point of disinfection is closer to what is needed to maintain a positive LSI.
With poorly buffered waters it is different. Any chemical addition significantly impacts the pH and the alkalinity, changing the characteristics of the water throughout the works. The final water may require a high pH to achieve a positive LSI with very low alkalinity levels, a pH which is difficult to achieve, maintain and which has the potential to change rapidly in the network. This conflicts with the immediate upstream needs for a relatively low pH for chlorine to achieve effective disinfection. Waters dosed with ortho-phosphate have their own pH requirements and chloraminated waters also need to be within a suitable pH range.
Developing a conditioning strategy – it’s complicated!
Developing a robust conditioning strategy is undoubtedly complex. It requires detailed understanding of water chemistry, pipe materials, potential for corrosion in the network, sampling and assessment of treated and network water quality and the balancing of conflicting water quality needs. Public health remains the primary basis for developing a conditioning policy and rightly so. But sometimes the only way of meeting all the water quality needs and to replace waters where needed is to change the treated water quality more fundamentally than just adjusting the pH.
Mindful of the need to manage water quality contacts in a region with both soft poorly buffered waters and hard groundwaters, Wessex Water has reviewed the corrosion indices for a range of its sources to inform the development of its future water quality strategy. Training seminars have been run and a design standard developed to complement its existing corrosion control policy. For one of its works it is developing plans for the installation of a carbon dioxide/lime dosing plant at a water treatment works treating soft, poorly buffered water to reduce the on-going corrosion in the networks and to use the water interchangeably in its supply zones. It will unquestionably cost a significant amount to install and operate. But replacing corroded mains is expensive too, as is the cost of customer complaints of discoloured water and burst pipes. Improving the water chemistry is an excellent step in the right direction.
-About the author: Lisa Barrott is a Technical Specialist at MWH (now part of Stantec), and was recently elected Chair of the CIWEM Water Supply and Quality Panel. The author would like to thank Paul Williams, Principal Process Engineer, MWH and Julian Welbank and Martin Gans of Wessex Water who funded the research behind this article.
- Keep on tracking The Norfolk Rivers Trust has installed a remote river monitoring station that has been tracking water quality and flow... Read More >
- UV delivered to four Wessex sites Wessex Water's plans for AMP5 included refurbishment of four groundwater sourced drinking water treatment plants. Natasha... Read More >
- Scottish Water completes 1.2km tunnel One of Scotland's largest projects in AMP5 has been completed in Airdrie. Scottish Water’s William Ancell explains how it... Read More >
- Getting to Grips with… highway pollution Pollution from urban highways, trunk roads and motorways poses significant risks to the environment from toxic metals and... Read More >
- Comment: Moving sensors from the lab to the real world Innovative sensor technologies of various materials are out there - the key now is to apply them effectively and to make... Read More >
- Corrosion of water pipes: out of sight, out of mind? Pipe corrosion is a major cause of water quality complaints by customers, but how often is the water itself responsible... Read More >
- Top tips for... analyser maintenance Keeping continuous water analysers well maintained is vital to ensure you can rely on the information they provide, writes... Read More >
- Data: Predictive analysis pays off for Welsh Water Data scientists at Welsh Water built an analysis tool that can predict when service reservoirs are at risk of bacterial... Read More >