Measurement of chloride ion concentration of reinforced concrete

Measurement of chloride ion concentration of reinforced concrete

Shaw, John

Highly alkaline conditions at the reinforcement/concrete interface lead to the formation of a passive film. Free chloride ions have the ability to locally break down this passive film, and support intensive reinforcement corrosion activity, even in alkaline concrete.

Chloride ion concentration

It is important to try to assess the source of chlorides. There are two possibilities; either they entered the mix before casting, or after the concrete hardened. These conditions are normally referred to as cast-in chlorides and ingressed chlorides. The former may be from the mixing water, inadequately washed marine aggregate, or through the addition of chloride-based accelerator compounds, as was common practice in the 1960s/1970s. The latter may be from a marine or industrial environment, or deicing salts. With age and exposure, ingressed chlorides increase in quantity and depth into the concrete.

Generally, cast-in chlorides are substantially chemically bound, while ingressed ones are substantially free – and only free chloride ions take part in chloride attack of reinforcement. The interpretation is somewhat clouded by the fact that bound chlorides are substantially released by the carbonation of concrete, and hence care in assessment is needed.

Chloride contents are normally expressed as a percentage by mass of concrete. The content by mass of cement may be derived, using either an assumed value, or a value determined by chemical analysis. The location of the test and the chloride levels at different depths should be recorded.

Sampling concrete

It is important that samples are large enough to be fully representative: a minimum of 25g of concrete dust should be taken for each gradient. All the dust resulting from drilling must be collected and care taken that wind does not blow the finer dust away. Studies have shown that more chloride is contained in the finer component of the drill dust and so a sample may be biased if it is not collected carefully. A simple dust collector can help (see Figure 1).

Samples from depth increments of 15-25mm are often used (for example 5-25, 25-50, 50-75mm, etc). The first 5mm drillings are normally discarded as non-representative. In practice, it has been found that dry drilling two nominal 20mm-diameter holes with a rotary percussive drill approximately 30-40mm apart and collecting the dust at each depth gives a good representative sample.

Determining chloride content

For accurate chloride determination, samples must be prepared for chemical analysis and analysed, using either Volhard’s method or potentiometric titration.

The analytical results, even when carried out by experienced laboratories, have historically shown some alarmingly wide variations. It is believed that this is due to a lack of care in the use of control samples when checking the accuracy of a batch of tests. Control samples must be actual dust samples of known chloride ion concentration and they must be run through the full preparation and test procedure. In that way, the accuracy of the whole test can be achieved and not just the performance of the testing machine(1,2).

It is important to ascertain that laboratory work is carried out to the prevailing national standards (e.g. BS 1881-124: 1988 Testing concrete. Methods for analysis of hardened concrete(3) and that the laboratory is a recognised establishment, preferably having accreditation such as the United Kingdom Accreditation Service (UKAS). The British Standard test for chloride content will be superseded by BS EN 14629: Products and systems for the protection and repair of concrete structures. Test methods. Determination of chloride content in hardened concrete(4), which is expected to be published in 2004.

In practice, given the competitive cost of accurate laboratory analyses, industry uses UKAS laboratories almost exclusively for accurate chloride analysis with site operations limited to sampling only, for speed and accuracy during concrete operations.

Threshold levels

To establish whether a chloride content above the threshold value constitutes a risk of reinf orcement corrosion, the content within the concrete cover and especially at the surface of the reinforcement must be determined.

The subject of ‘safe’ levels of chloride in the concrete cover of reinforced concrete is somewhat controversial. One view is that at 0.4% chloride by mass of cement in the cover there is a 50% probability of corrosion initiation, and at 1.0% chloride in the cover this may increase to 95%. Another fairly traditional view is that below 0.4% chloride by mass of cement represents a low corrosion risk, 0.4-1 % a medium risk, and above 1 % a high risk.

These are all now felt to be rather simplistic models, as they do not necessarily take into account the source of the chloride (cast-in or ingressed), an estimate of the likelihood of large amounts of free chloride ion to promote chloride attack, or indeed the profile of penetration into the concrete. The best interpretation is found in BRE Digest 444: Part 2(5). This considers all of the above aspects, plus the cover, carbonation depth, and age of the concrete. It then gives a much more elaborate consideration of the risk of reinforcement corrosion, with models for both cast-in and ingressed chlorides. In the assessment of the latter, it is important to be aware of the source and likelihood of everincreasing levels of chlorides with the further passage of time.

Additional tests

In any assessment of the deterioration of reinforced concrete as a result of corrosion, it is important to measure not only the level of chloride ion contamination but also the depth of carbonation and the concrete cover. Other tests such as measuring half-cell or corrosion potential, linear polarisation for corrosion rate, and concrete resistivity may also be used to identify ‘hot spots’ where corrosion and hence concrete deterioration are more likely, and therefore where some corrosion prevention or repair techniques should be used(6-8).


1. GRANTHAM, M. and VANES, R. Admitting that chlorides are variable, Construction Repair, November/December 1995.

2. NUSTAD, G. Experience on accuracy of chloride and sodium analysis of hardened concrete (in) proceedings of Corrosion of reinforcement in concrete – monitoring, pretention and rehabilitation, European Federation of Corrosion, Publication No, 25, 1997, pp. 150-157.

3. BRITISH STANDARDS INSTITUTION. BS 1881-124:1988 Testing concrete. Methods for analysis of hardened concrete, London, 24pp.

4. BRITISH STANDARDS INSTITUTION, pr EN 14629. Products and systems for the protection and repair of concrete structures. Test methods. Determination of chloride content in hardened concrete, 11pp.

5. BUILDING RESEARCH ESTABLISHMENT. Corrosion of steel in concrete: investigation and assessment, Digest 444, Part 2, Building Research Establishment, Garston, 2000.

6. TECHNICAL LIAISON COMMITTEE OF THE INSTITUTE OF CORROSION and THE CONCRETE SOCIETY. Current Practice Sheet 120: Half-cell potential surveys of reinforced concrete structures, CONCRETE, Vol. 34, No. 7, July/August 2000, pp.43-45.

7. BROOMFIELD, J. and MILLARD, S. Current Practice Sheet 128: Measuring concrete resistivity to assess corrosion rates, CONCRETE, Vol. 36, No. 2, February 2002, pp.37-39.

8. BROOMFIELD, J. and MILLARD, S. Current Practice Sheet 132: Measuring the corrosion rate of reinforced concrete using linear polarisation resistance, CONCKETE.Vol. 37,No. 3, March 2003, pp.36-38.

John Shaw, Consultant and Roel van Es, Makers UK Ltd

Copyright The Concrete Society Sep 2003

Provided by ProQuest Information and Learning Company. All rights Reserved