hlorides and other soluble salts can cause signifi- motic blistering occurs because the salt on the

the substrate in

C cant reductions in coating life and accelerated cor-

rosion of metals if not removed prior to
coating application. In the last several
years, industry has developed new tech-
niques for extracting salts using de- Advances
the presence of moisture forms a highly concentrated solu-
tion. The water on the exterior of the film is at a
much lower concentration. The water is then
drawn through the coating film, which be-
haves like a semi-permeable mem-
vices such as sleeves, magnetic brane. Figure 1 illustrates osmotic
cells, and filter paper. Also avail-
able are improved field methods
for analyzing conductivity,
in Technology blistering.
In addition, salts, particularly
chloride or sulfate, can increase
chlorides, and sulfates. In ad-
dition, organizations such as
SSPC and the International
and Standards the rate of corrosion of metals.
These salts act as catalysts, ac-
celerating the anodic reaction.
Organization for Standard-
ization (ISO) have issued
standard methods to guide
for Mitigating Atmospheric and
Immersion Studies
the specifiers, applicators,
and inspectors in selecting
the Effects Several studies have been un-
dertaken demonstrating the in-
and utilizing the methods. Fol- fluence of soluble salts on the
lowing a review of recent stud-
ies on the effect of soluble salts
of Soluble lifetimes of coatings. Morcillo1
evaluated a series of conventional
on coating performance, this arti-
cle reviews the most commonly used
procedures for extraction, analysis, and
Salts coatings over three to four levels of
sodium chloride and ferrous sulfate.
He exposed the coatings for 4.5 years in
By Bernard R. Appleman, Ph.D.,
removal of salts from steel substrates. In
addition, the article provides guidance for KTA-Tator, Inc.
specifiers to assure that the levels of salt remaining
on the surface will not be detrimental to the coating service
life.

Background
Soluble salts are very widespread on government and in-
dustrial substrates. A major source is sea salt, which affects
ships, offshore structures, waterfront structures, and inland
facilities within several miles of the sea. Soluble salts can
also arise from chemical processes, cooling towers, and
burning of sulfur-containing coal.

Effects of Soluble Salts on Coating Life

General Effects
Soluble salts can affect the ability of coatings to protect steel
in several ways. Salts that remain on the substrate can result
in blistering of the coating. This phenomenon, known as os-
motic blistering, can produce pressures of several thousand Fig. 1: (top) Schematic of
psi, which is enough to cause the coating to disbond. Os- development of osmotic cell.
(left) Osmotic blistering
Photos courtesy of the author
Editor’s Note: This paper was first presented at the U.S. unless otherwise noted.
Navy and Industry Corrosion Technology Information
Exchange, July 16–20, 2001, in Louisville, Kentucky.

42 JPCL May 2002

 C), an epoxy phenolic and an epoxy novolac ex-
Table 1: hibited thresholds of 17.5 µg/cm2 and 7.4
ISO Summary on Acceptable Chloride Levels µg/cm2, respectively. All the other coatings failed
even with no measurable chloride on the surface.
“Safe”
The author concluded that even very small quan-
Coating System Level of Cl* Test Exposure
tities of chloride on the surface reduced the ser-
Epoxy phenolic (1 coat) 1 100% humidity, 40 C
vice life of these coatings.
Epoxy polyamide (3 coats) 5 condensing humidity
Coal tar epoxy (10 mils) 50 500 hrs immersion, deionized water Determining the Maximum Acceptable Levels
Fusion-bonded epoxy <3 48 hrs immersion, 65 C water, A subcommittee from ISO reviewed published
1.5 volts cathodic disbondment data on the performance of various coatings in
Tank lining epoxy 10–20 pressure immersion (90 F, 50 psi)** different exposures when applied over chloride.4
Epoxy mastic (2 coats) 7 pressure immersion (90 F, 50 psi)** Some of the data are shown in Table 1.
* units of micrograms per square centimeters ** 50 psi = 3.4 bar; 90 F = 32 C Most of the above tests were based on apply-
ing the coating over a pre-measured level of
a marine atmosphere and up to 14 years in industrial, rural, chloride or sulfate. This procedure provides the researcher
and urban atmospheres. SSPC2 evaluated typical bridge coat- with a relatively accurate measure of the level of salt on the
ing systems in accelerated laboratory tests and on test surface.
bridges over chloride- and sulfate-contaminated steel. To assess the potential impact of salts deposited on a sur-
Each of these atmospheric studies determined that coating face from the environment, one requires methods for de-
failure was more pronounced at chloride levels starting at tecting the salt after the deposition.
about 10 µg/cm2, but that zinc-rich coatings had a greater Detection is typically a two-step process: first the salt is
tolerance for chloride. Also, coatings had a higher tolerance extracted from the surface. Then, the extracted solution is
for sulfate ions than for chloride ions. analyzed for the type and quantity of salt.
In a more recent study, Mitschke3 studied immersion coat- The next section describes the most common methods
ings. He evaluated a series of nine epoxy, epoxy novolac, used by the coatings industry for extraction and analysis.
and epoxy phenolic tank linings over different levels of chlo-
ride immersed in tap water at different temperatures. He ex- Extracting Salts from Surfaces
amined the panels periodically, defining failure when 20% of The industry has developed several field methods for ex-
the surface was blistered. At 75 F (24 C) after 13 months, the tracting soluble salts from surfaces to be painted. The most
threshold chloride levels (concentrations which could be tol- prominent are swabbing and adhesive patch (Bresle [Fig. 2
erated) ranged from 4 µg/cm2 to 20 µg/cm2. As the temper- and 3]). Others include the adhesive sleeve (Fig. 4) and
ature was raised, the threshold levels decreased. At 190 F (88 wetted filter paper. (See discussion below on conductivity.)

Fig. 2: Swabbing in progress for extraction of soluble salts from a steel surface

JPCL May 2002 43

 the swabbing method gave extraction efficiencies in the
Table 2: range of 30–40% while the Bresle cell efficiencies were in
Summary of Published Data on Extraction Efficiency the range of 40–60%.5 More recent data and reassessment
of earlier data have shown this to be an over-simplification.
For the swabbing method, extraction efficiencies have
Method Salt/Concentration Efficiency Source
ranged from 20–90% while those for the Bresle cell have
Swabbing Cl: 20, 200, 250 43% to 78% Reference 2
ranged from 20% to over 100%. Measured extraction effi-
& 500 µg/cm2 avg: 56%
ciencies of greater than 100% can arise from inaccuracies
Swabbing SO4: 20 & 200 27% to 42% Reference 2 (e.g., non-uniformity) in the initial deposition of the salt or
µg/cm2 avg: 34% in the extraction or analytical procedures. For the newer
Swabbing NH4: 10 & 100 24% to 86% Reference 2 methods identified above, there is little if any objective
µg/cm2 avg: 55% data available. Table 2 presents a summary of the pub-
Swabbing Cl: 5, 15, 50, 100 82% to 120% Reference 6 lished data on extraction efficiency.
µg/cm2 avg: 100%
Swabbing SO4: 100, 150, 200 80% to 86% Reference 6 Analyzing Soluble Salts
µg/cm2 avg: 83% The most common species analyzed is chloride ion. There
Bresle Cl: 15, 50 133% to 166% Reference 6
µg/cm2 avg: 150%
Bresle SO4: 100, 150 57% to 126% Reference 6
µg/cm2 avg: 92%
Bresle Cl: 10, 25, 50 42% to 90% Reference 7
µg/cm2 avg: 62%
Bresle Cl: 10, 25, 50 17% to 28% Reference 7
µg/cm2 avg: 22%
Bresle Cl: 10, 25, 50 26% to 53% Reference 7
µg/cm2 avg: 35%
Bresle Cl: 1.5, 3.0, 6.0, 18 20% to 80% Reference 8 Fig. 5: (above) Ion
µg/cm2 avg: ~60% detection tubes

Fig. 3: (left) Adhesive

patch (Bresle method)
for extracting soluble
salts

Fig. 6: (left) Paper

chromatography strips
placid in extracted liq-
Fig. 4: (right) Adhesive uid to determine chlo-
sleeve for soluble salt ride concentration
extraction
Courtesy of Chlor*Rid
International

These are described in SSPC-TU 4, “Field Methods for Re- are three common field methods for analyzing chloride.
trieval and Analysis of Soluble Salts on Substrates,” and These include ion detection tubes (Fig. 5), paper chro-
other publications. matography strips (Fig. 6), and titration. Each of these is
Unfortunately, the efficiency of extraction varies quite sufficiently accurate and precise for determining the con-
significantly among these methods and within a given centration of chloride in the extracted liquid. The ion de-
method. Earlier data published by SSPC had indicated that tection tube is the most sensitive. (Of the three methods, it
44 JPCL May 2002
 Fig. 7: Pocket chloride and conductivity.
conductivity meter Methods for Removing Soluble Salts
Traditionally, the most productive and effective method for
preparing steel for application of a coating is dry abrasive
blasting. Dry blasting, however, is most suitable for mechan-
ically breaking up layers of rust, mill scale, and coating, and
for eroding the steel to produce a profile. It is not intended to
remove water-soluble salt or grease and oil. For more effec-
tive removal of salts, some form of water in the surface
preparation is desirable.
Waterjetting at high-and ultra-high-pressures of 10,000 to
35,000 psi (666 to 2,333 bar) is generally effective at dissolv-
ing and removing any salts the waterjet can reach. However,
Fig. 8: (below) Meter mea- these jets by themselves (i.e., without abrasive) are relatively
sures conductivity of pre-
wetted filter paper placed on ineffective at removing tight rust or mill scale. In addition,
substrate water alone cannot produce a surface profile normally rec-
Courtesy of Elcometer Ltd. ommended for high-performance coatings. These systems
are, however, suitable for maintenance painting where the
surface previously had a profile that can be restored.
Another approach is wet abrasive blasting, in which water
and abrasive are utilized. One version is to inject water, typ-
ically at 1,000–3,000 psi (66 to 200 bar), into a conventional
air abrasive blasting nozzle. Alternatively, one can inject
abrasive into a waterjetting system. These techniques are de-
scribed in SSPC-TR 2/NACE 6G198, Wet Abrasive Blast
Cleaning. Some recent data on the effectiveness of wet and
dry methods in removing chlorides are given in Table 3.9–12

Table 3:
Comparison of Salts from Wet and Dry Cleaning Methods

Remaining Salt
Method (µg/cm2) % Extracted Source
Wet blasting 0–3.2 avg*: 96.2% Reference 9a
can measure the lowest level of chloride.) Waterjetting (35 ksi) 0–2.4 avg*: 95.9% Reference 9
Conductivity measures the ability of the extracted liquid Hand tool (SP 2) 160–288 avg*: 43.8% Reference 9
to carry an electric current. It is a measure of the total dis- Power tool (SP 3) 212–296 avg*: 35.4% Reference 9
solved salts, but does not provide direct information on the Blast (SP 6) 44–68 avg*: 83.0% Reference 9
specific chemical ions. There are several types of field con- UHP waterjet 1.6–1.8 avg: 93.5% Reference 10b
ductivity meters available. One is the pocket conductivity Blast (SP 10) 3.3 84% Reference 10
meter shown in Fig. 7. A relatively new device is a meter Needle gun (SP 3) 11.4 3% Reference 10
that measures the conductivity of pre-wetted filter paper 15.2
Wire brush (SP 2) 9% Reference 10
placed on the substrate (Fig. 8). The units are furnished
Blast (SP 5) <3.2–3.4 avg: 90.2% Reference 11c
with calibration solutions. They are sensitive to conductiv-
Power tool (SP 3) 16.2–24.1 avg: 43.5% Reference 11
ities of 2–3 microsiemens/cm. (This concentration is ap-
SP 3 + steam 8.6–12.9 avg: 69.9% Reference 11
proximately equivalent to 6–10 parts per million [ppm]
Power tool (SP 11) 7.0–13.9 avg: 72.1% Reference 11
chloride if one assumes that chloride is the only soluble
SP 11 + steam 3.9–7.7 avg: 84.5% Reference 11
ion.) These instruments are also sufficiently accurate and
Power tool (SP 3) 22–97 avg: 45.4% Reference 12d
precise for measuring the soluble salts at levels that can af-
fect coating performance. Power tool (SP 11) 41–124 avg: 17.2% Reference 12
Field test kits are also available for measuring the levels a Salts extracted by Bresle and analyzed by conductivity
b Methods not included in paper
of soluble sulfate and ferrous ions as well as nitrate ions.
c Salts extracted by boiling and analyzed by selective ion electrode
These salts are much less frequently specified than are d Salts extracted by swabbing and analyzed by conductivity

JPCL May 2002 45

 • sampling (where to measure the surface),
Table 4: • extraction,
Selected Contents of SSPC-TU 4 • analysis, and
• acceptance level.
Section Description
3.3 Adhesively Bonded Cell (Bresle Cell)
Sampling
3.4 Swabbing or Washing Method
Presently, there is little consensus within the industry re-
3.5 Total Extraction Method
garding the number or location of samples. The U.S. Naval
4.2 Field Measurement of Conductivity (Total Soluble Salts)
Facilities Engineering Command (NAVFAC) guide specifi-
4.3 Field Detection of Chloride Ion by Kitigawa Tube cation, Interior Coating of Welded Steel Petroleum Fuel
4.4 Field Detection of Chloride Ion by Quantab Method Tanks, stipulates three tests for the first 100 m2 (1,100 ft2)
4.5/4.6 Field Detection of Chloride Ion by Titration Methods plus one additional test for each additional 200 m2 (2,200
1 and 2 ft2). The specification also instructs the inspector to con-
4.7 Laboratory Reference Method for Detection of Chloride centrate testing at areas of coating failure, pitting, and
Ion by Titration welds. Others have also suggested selecting the areas with
4.8 Qualitative Field Detection of Ferrous Ion the greatest likelihood of high salt levels (e.g., under
4.9 Quantitative Field Detection of Ferrous Ion bridge expansion joints or at the base of ballast tanks).
4.10 Field Detection of Sulfate Ion SSPC and NACE are drafting procedures for sampling.
4.11 General Method for Estimating Surface
Concentration of Salts Extraction and Analysis
Appendix C Estimating Surface Salt Concentration SSPC-TU 4 includes specific methods for extraction and
analysis for most of the techniques described above. (A re-
Other surface preparation methods have also been eval- vised edition is expected in 2002 that will include more de-
uated. Power and hand tool cleaning methods are very in- finitive descriptions of methods.) The relevant contents of
effective in removing salts. Several proprietary treatments TU 4 are shown in Table 4.
have been developed. These entail spraying a water solu- The ISO subcommittee identified above also contacted
tion of a proprietary chemical to the surface. The concen- coating manufacturers from around the world for recom-
trations and the pressure vary, depending on the particular mendations on the acceptance level for chloride.4 Their
product. Several owners and specifiers have rated these recommendations along with those of the U.S. Navy and a
products very highly. However, there is little published Norwegian classification society13 are shown in Table 5.
data to corroborate these claims. For each source, the original report provides recommenda-
tions for the specific methods of extraction and analysis.
Specifying Soluble Salt Removal These requirements are primarily based on marine expo-
This section provides guidance to specifiers and owners in sures such as ocean-going ships and offshore structures.
preparing and implementing a specification to assure ade- Studies by Morcillo, Appleman, and others have demon-
quate salt removal. It is essential that the specification in- strated that the susceptibility of coatings to early degrada-
clude clear language for tion from salts is strongly dependent on the type of coat-

List of ISO Standards on Soluble Salts

ISO 8502-2:1992 Preparation of Steel Substrates before Application of Paints
and Related Products—Tests for the Assessment of Surface Cleanliness
Part 1: Field Test for Soluble Iron Corrosion Products (ISO/TR 8502-1:1991)
Part 2: Laboratory Determination of Chloride on Cleaned Surfaces ( ISO 8502-2:1992)
Part 5: Measurement of Chloride on Steel Surfaces Prepared for Painting (Ion Detection Tube Method) (ISO 8502-5:1998)
Part 6: Extraction of Soluble Contaminants for Analysis—The Bresle Method (ISO 8502-6:1995)
Part 9: Field Method for the Conductometric Determination of Water-Soluble Salts (ISO 8502-9:1998)
Part 10: Field Method for the Titrimetric Determination of Water-Soluble Chloride (ISO 8502-10:1999)
Part 11: Field Method for the Turbidimetric Determination of Water-Soluble Sulfate (ISO/AWI 8502-11)
Part 12: Field Method for the Titrimetric Determination of Water-Soluble Ferrous Ions (ISO/DIS 8502-12)
Part 13: Field Method for the Determination of Soluble Salts by Conductometric Measurement (ISO/WD 8502-13)

46 JPCL May 2002

 Soltz, “Effect of Surface Contaminants on Coating
Table 5: ISO Survey of Manufacturers’ Life,” Report No. FHWA-RD-91-101, Federal Highway
Recommended Chloride Limits Administration (June 1991).
3. Howard Mitschke, “Effects of Chloride Contamination
Source Level for Immersion Level for Atmospheric on Performance of Tank and Vessel Linings,” JPCL
Coating suppliers 3–10 5–25 (March 2001), pp. 49–56.
U.S. Navy 3 5 4. ISO/PDTR 15235 “Preparation of Steel Substrates before
Det Norske Veritas10 2 — Application of Paint and Related Products—Collected
units = µg/cm2
Information on the Effect of Levels of Water Soluble
Salt Contamination before Application of Paints and
ing, its thickness, and the service environment. Related Products,” prepared by ISO TC35/SC12 Work-
Specification Language ing Group 5, International Organization for Standard-
Listed below are several examples of clauses that could be ization (November 18, 1999).
incorporated into procurement documents. 5. B.R. Appleman, J.A. Bruno, Jr., and R.E.F. Weaver,
• Extraction by Patch Cell—“Extract salts from the surface “Maintenance Coating of Weathering Steel: Interim Re-
in a minimum of 3 test areas within a 100 sq ft (9 sq m) port,” Publication No. FHWA-RD-91-087 (Washington,
section at representative units of the structure or vessel in DC: Federal Highway Administration, January 1992),
accordance with Section 3.3 (Patch Cell) from SSPC-TU 4.” SSPC Publication 91-04.
• Chloride Analysis by Ion Tube—“Test each portion of liq- 6. S. Flores, J. Simancas, and M. Morcillo, “Methods for
uid extracted for soluble chloride ion in accordance with Sampling and Analyzing Soluble Salts on Steel Sur-
Section 4.3 of SSPC-TU 4 (ion detection tubes); record the faces: A Comparative Study,” JPCL (March 1994), pp.
ppm.” 76–83.
7. Simon K. Boocock, “SSPC Research and Performance
Conclusion Testing of Abrasives and Salt Retrieval Techniques,”
Soluble salts, particularly chlorides, are widely prevalent JPCL (March 1994), p. 28.
in many industrial exposures where coatings are applied. 8. Unni Steinsmo and Sten B. Axelsen, “Assessment of Salt
If not removed prior to application of the coatings, the sol- Contamination and Determination of Its Effect on
uble salts can adversely affect coating lifetime, resulting in Coating Performance,” Proceedings of the PCE 98 Con-
early degradation and failure. The most successful means ference and Exhibition, The Hague, The Netherlands,
of removing the salts is a combination of water and abra- April 1–3, 1998 (Pittsburgh, PA: Technology Publishing
sive, but even this technique will not be 100% effective. Co., 1998), pp. 71–85.
Detection of salt on the surface entails extraction, then 9. A. Forsgren and C. Appelgren, “Comparison of Chloride
analysis for conductivity or for a specific salt such as chlo- Levels Remaining on the Steel Surface After Various
ride. Standard extraction methods are available, but they Pretreatments,” Proceedings of PCE 2000, Genoa, Italy,
give variable results which depend on the operator, the March 8–10, 2000 (Pittsburgh, PA: Technology Publish-
concentration of salt, the roughness of the steel, and other ing Co., 2000).
factors. Standard field analytical methods are available for 10. Bill Allen, “Evaluating UHP Waterjetting for Ballast
analyzing chloride and conductivity. These methods pro- Tank Coating Systems,” PCE (October 1997), p. 38.
vide consistent, accurate results. 11. Kenneth A. Trimber, “An Investigation into the Re-
The industry is moving toward consensus on acceptable moval of Soluble Salts Using Power Tools and Steam
levels of salt. There is, however, relatively little solid data Cleaning,” Proceedings of the 7th SSPC Technical Sym-
to support these levels. Guidelines and consensus stan- posium, SSPC Publication 88-03 (1988), pp. 56–67.
dards are available to assist specifiers in mitigating the ef- 12. John W. Peart, “The Effectiveness of Power Tool Clean-
fects of soluble salts. ing as an Alternative to Abrasive Blasting,” Report No.
NSRP 0447, National Shipbuilding Research Program
References (June 1995).
1. Manuel Morcillo and Joaquín Simancas, “Water-Soluble 13. E. Askheim, “Ballast Tank & Cargo Holds in DNV’s
Contaminants at the Steel/Paint Interface: Their Effect Guidelines for Corrosion Protection of Ships,” PCE
in Atmospheric and Marine Services,” Paper 12 of the Guide to Marine Coatings (Pittsburgh, PA: Technology
Proceedings of PCE 97, The Hague, The Netherlands, Publishing Co., 2000), pp. 332–341.
March 17–20, 1997 (Pittsburgh, PA: Technology Pub-
lishing Co., 1997).
2. B.R. Appleman, S.K. Boocock, R.E.F. Weaver, and G.C.

JPCL May 2002 47