report on corrosion measurements for material

STNP Subcommittee Task-Force on Corrosion Effects on Subsurface Transformers

Report on Corrosion Measurements for Material Compatibility Testing

Prepared by Will Elliott – Prolec GE – Shreveport, LA

Submitted on 10/17/2020 – (Revised 10/20/2020)

Purpose & Test Methods:

A series of (18) corrosion experiments were performed by creating galvanic cells using various

hardware material and copper cathode combinations. These tests were performed as a proof of

concept using materials readily available to demonstrate the concept to the task-force for future

consideration and development. It was assumed that the task-force would specify more precise

methods and other materials of interest for future testing.

Table 1 lists the galvanic cells that were created using different hardware combinations immersed in

a salt-water solution (electrolyte) to evaluate the relative corrosion depth rates between the different

materials. The same hardware combinations were also tested for multiple cathode configurations

(bare copper, coated copper, and no copper present) to evaluate the impact of a (spray-painted)

coated cathode on the corrosion depth rates, and a control with no copper cathode. All tests

durations were for 98 days and at similar temperatures. Figures 1 to 8 show test setup photos.

All hardware tested was bare (uncoated). All weld-nuts were also bare, except the bottom which was

insulated from the electrolyte to provide an electrical connection for galvanic measurements. Nylon

straps were used as provisions to suspend the assembled samples in the electrolyte. When a

cathode was used, it was electrically connected to the hardware external to the electrolyte.

During each test the galvanic potential was measured periodically. The galvanic potential for tests

with a copper cathode were measured between the hardware and the copper (using an insulated

wire). The galvanic potential for the test without a copper cathode was measured using a

submersible copper sulfate reference electrode.

Table 1: Galvanic Cell Materials & Test Durations

Bare Hardware Combinations Bare Copper Cathode Test

Coated Copper Cathode Test

No Copper Cathode Test

Bolt+Washers: 304 Stainless-Steel Weld-Nut: Carbon-Steel

Test#1

2/10/2020 to

5/19/2020

98 days

19.0~22.3°C

Test#2

6/29/2020 to

10/5/2020

98 days

18.6~20.3°C

Test#3

6/29/2020 to

10/5/2020

98 days

18.6~20.3°C

Bolt+Washers: Silicon-Bronze Weld-Nut: Carbon-Steel

Bolt+Washers: Galvanized-Steel Weld-Nut: Carbon-Steel

Bolt+Washers: 304 Stainless-Steel Weld-Nut: 303Se Stainless-Steel

Bolt+Washers: Silicon-Bronze Weld-Nut: 303Se Stainless-Steel

Bolt+Washers: Galvanized-Steel Weld-Nut: 303Se Stainless-Steel

Insulated Supports

for Suspension

Insulated Wire

to Measure

Galvanic

Potential

Electrical Connection Insulated with Silicone Caulk

Figure 2: Photo of Hardware Samples with Insulated Provisions to Suspend & Measure Samples

Figure 1: Photo of Hardware Sample Material Combinations

Figure 3: Test#1 (Bare Copper) Test Setup – Top View

Figure 4: Test#1 (Bare Copper) Test Setup – Side View

Figure 6: Test#2 (Coated Copper) Insulated Copper Cathode

Figure 5: Test#2 (Coated Copper) Test Setup – Top View

Figure 8: Test#3 (No Copper) Hardware Samples & Copper Sulfate Reference Electrode

Figure 7: Test#3 (No Copper) Test Setup – Top View

Materials and Measurement Equipment:

All dimensions listed below are inches (“) and bolts & threads are in customary US nomenclature.

• Electrolyte:

o Solution of 19.2g NaCl (non-iodized table salt) and 600mL tap-water (unknown purity)

• Cathode:

o Copper strip (UNS C11000) measuring 34.5 inches long X 0.75 inches wide X 0.03

inches thick

• Bare Copper

• Coated Copper

• Base-coat of Performix PlastiDip Rubber Coating (to simplify removal)

• Top-coat of Rust-Oleum Universal Gloss White

• Hardware:

o Weld-Nuts (threaded nuts intended for welding to equipment enclosure)

• AISI 303Se stainless-steel (UNS S30323) weld-nut

• 1.0 inch diameter X 0.875 inches tall (½-13 threads)

• AISI 1020 carbon-steel (UNS G1020) weld-nut

• 1.0 inch diameter X 0.750 inches tall (½-13 threads)

o Bolts (hex-head bolts)

• AISI 304 stainless-steel (UNS S30400) ½-13x1.75 (inches) full-thread hex bolt

• Silicon-Bronze (UNS C65100) ½-13x2.00" (inches) full-thread hex bolt

• Galvanized-steel ½-13x2.00" (inches) partial-thread hex bolt

o Helical Spring Lock Washers

• AISI 304 stainless-steel (UNS S30400) ½ inch nominal lock washer

• 0.869 inch outer diameter, 0.171 inch width X 0.125 inch thick

• Silicon-bronze (UNS C65100) 5/8 inch nominal lock washer


• Galvanized-steel ½ inch nominal lock washer


o Flat Washers

• AISI 304 stainless-steel (UNS S30400) ½ inch nominal washer

• 0.53 inner diameter X 1.00 inch outer diameter X 0.063 inch thick

• Silicon-bronze (UNS C65100) ½ inch nominal washer

• 0.53 inner diameter X 1.25 inch outer diameter X 0.100 inch thick

• Galvanized-steel ½ inch nominal washer

• 0.53 inch diameter X 1.375 inch outer diameter X 0.100 inch thick

o Hex Nuts

• AISI 304 stainless-steel (UNS S30400) ½-13 (inches) hex nut

• Silicon-bronze (UNS C65100) ½-13 (inches) hex nut

• Galvanized-steel ½-13 (inches) hex nut

NOTE: Identical hardware dimensions were not precisely consistent between the different

hardware materials, as noted above, but consistent hardware dimensions were

used for the same material for the different tests performed.

• Measurement Equipment:

o High-resistance voltmeter

• MC Miller LC4.5 Voltmeter

o Copper Sulfate Reference Electrode

• MC Miller IONX Submersible Copper/Copper Sulfate Electrode

o Mass Scale

• Low cost AWS BT2-201 Jewelry Scale with 0.01g resolution

• 100g calibration weight

o Calipers

• Fowler precision dial calipers with 0.001 inch resolution

Galvanic Cell Circuit Diagram:

The general equivalent circuit diagram of the galvanic cells (with copper cathodes) during the tests is

shown in Figure 9. The resistances shown are not external resistances, rather they are the (very

small) intrinsic resistances of the materials and the resistance of the short (𝑅𝑠ℎ𝑜𝑟𝑡) external to the

electrolyte. The full-cell galvanic potential (𝑉𝐺) that develops in the electrolyte is the difference in

potential between the copper cathode (𝑉𝐶) and more anodic hardware test sample (𝑉𝐴). Because the

circuit is complete a corrosion current (𝑖𝑐𝑜𝑟𝑟) flows in the circuit. The positive and negative circles in

the electrolyte represent ions and electrons in the solution, which complete the circuit in the

electrolyte. The circled e- is of course electrons flowing in the wire.

It is important to note that a galvanic potential also exists between the differing materials in the

hardware test sample, but it is not externally measurable in the test setup. Figure 9 shows the

combination of hardware materials is lumped together; however, Figure 10 shows the galvanic

potential within a generalized hardware sample with a combination of materials.

Figure 9: Circuit Diagram of Galvanic Cell with a Copper Cathode

https://www.mcmiller.com/lc-4-5-voltmeter-5203

https://www.certifiedmtp.com/m-c-miller-ionx-submersible-copper-copper-sulfate-electrode/

Figure 10 provides more detail on the galvanic potential that exists between the metals in a mixed

material hardware sample. The figure generically represents all the combinations tested. In most

cases the bolt & washers were the cathode and the weld-nut was the anode, but in others the

relationship was reversed. The schematic representation in Figure 10 is comparable to Figure 9,

where the galvanic voltage (𝑉𝐺) exists due to electrochemistry (and not an external voltage applied).

It is worth noting that while Figure 10 does provide more detail of the hardware sample sub-circuit

that existed in the tests with a copper cathode (Test#1 & Test#2) it also describes the full galvanic

cell tested in Test#3, which did not use a copper cathode.

Galvanic Potential Measurement Procedure:

As already noted, galvanic potential measurements were performed periodically over the duration of

the test. These measurements were done to verify if the potential changed over time. The

measurements were made equivalent to the draft procedure submitted to the task-force previously.3

• Tests with a Copper Cathode (bare or coated):

o The (external) electrical connection between the copper cathode and the hardware

sample was disconnected to perform the measurement.

o The measurement was made with the positive voltmeter lead connected to the copper,

and the negative voltmeter lead connected to the insulated wire connected to the

hardware sample.

o The measurement connection schematic used is shown in Figure 11, using previously

defined symbols and variables.

Figure 10: Circuit Diagram of a Galvanic Cell within Hardware Samples

• Tests without a Copper Cathode:

o The measurement was made with the positive voltmeter lead connected to the

insulated wire connect to the hardware sample (node 𝑉𝑀), and the negative voltmeter

lead connected to the copper sulfate reference electrode (node 𝑉𝐶𝑆𝐸). The electrical

connections were made external to the electrolyte. The measured voltage is labelled

simply as 𝑉 in the highly generalized measurement schematic shown in Figure 12.

• Measurement Duration:

o The galvanic potential was measured until the voltage stabilized. The measurements

for using the reference electrode typically only took a couple of minutes to stabilize, but

the copper cathode measurements generally took 30 minutes to an hour to stabilize due

to the capacitance of the circuit.

o Measurements were considered stable when the voltage did not change by more than 1

mV per minute.3

Figure 11: Schematic for Galvanic Potential Measurements for Tests with a Copper Cathode

Figure 12: Schematic for Potential Measurements for Tests without a Copper Cathode

• General Measurements with the Reference Electrode:

o It is worth noting that the reference electrode was also used to validate the potentials

measured for the copper cathode tests. These measurements were only performed

once to verify theory, but the half-cell potential was measured from the reference

electrode to the copper cathode (𝑉𝐶) as well as from the reference electrode to the

hardware samples (𝑉𝐴). The full-cell galvanic potential in those cells was the difference

in those half-cell potential measurements (𝑉𝐺 = 𝑉𝐶 − 𝑉𝐴), as shown in Figure 11.

o The measurement connection schematic used to measure half-cell potentials is shown

in Figure 13, using previously defined symbols and variables. It is also worth noting

that the measured voltage is always negative when this measurement is performed.

o These measurements were not included in this report, but this information has been

included to explain the general use of reference electrodes in corrosion measurements,

and the procedures listed in this section are consistent with the draft field corrosion

measurement procedure previously submitted to the task-force.3

Galvanic Potential Measurements:

The galvanic potentials were recorded periodically over the course of the tests. There was no

attempt to perform the measurements at precise intervals; they were only performed when it was

practical to do so. Dates are listed as Month/Day (all in 2020). Note the following abbreviations in

the tables:

• “SB HW” refers to silicon-bronze bolt, lock washer & flat washer, as detailed prior.

• “SS” is stainless-steel. “SS HW” is AISI 304 stainless steel bolt, lock washer & flat washer.

“SS Weld-Nut” is the AISI 303Se weld-nut. These materials were defined in detail prior.

• “CS Weld-Nut” refers to the AISI 1020 weld-nut, as detailed prior.

• “GS HW” refers to the galvanized-steel bolt, lock washer & flat washer, as detailed prior.

Figure 13: Schematic for Measuring Half-Cell Potentials with a Reference Electrode

Table 2: Measured Potentials with Bare Copper Cathode (Test#1)

Date Measured 2/11 2/18 2/19 2/25 3/2 3/11 3/24 5/19

Elapsed Days 0 7 8 14 20 29 42 98

SB HW on CS Weld-Nut 0.416 0.474 0.465 0.465 0.477 0.480 0.417 0.461

SB HW on SS Weld-Nut 0.015 0.141 0.138 0.142 0.165 0.164 0.148 0.168

SS HW on CS Weld-Nut 0.427 0.470 0.473 0.473 0.470 0.469 0.422 0.470

SS HW on SS Weld-Nut 0.240 0.223 0.223 0.222 0.226 0.222 0.194 0.165

GS HW on CS Weld-Nut 0.795 0.798 0.785 0.771 0.762 0.694 0.595 0.477

GS HW on SS Weld-Nut 0.817 0.806 0.793 0.778 0.724 0.665 0.571 0.432

Temperature [°C] 18.7 19.1 18.9 19.0 22.3 21.5 21.0 22.0

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

0 10 20 30 40 50 60 70 80 90 100

Po

ten

tial

[V

]

Time [days]

Test#1 - Measured Potentials with Bare Copper Cathode over Time

Series1 Series2 Series3 Series4 Series5 Series6

Figure 14: Graph of Measured Potentials with Bare Copper Cathode (Test#1)

Table 3: Measured Potentials with Coated Copper Cathode (Test#2)

Date Measured 6/29 7/6 9/16 9/28 10/5

Elapsed Days 0 7 79 91 98

SB HW on CS Weld-Nut

0.155 0.240 0.223 0.226

SB HW on SS Weld-Nut

0.441 0.250 0.251 0.251

SS HW on CS Weld-Nut

0.477 0.475 0.472 0.470

SS HW on SS Weld-Nut

0.215 0.255 0.266 0.269

GS HW on CS Weld-Nut

0.787 0.503 0.463 0.461

GS HW on SS Weld-Nut

0.781 0.761 0.746 0.744

Temperature [°C]

20.3 19.2 18.8 18.6

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

0 10 20 30 40 50 60 70 80 90 100

Po

ten

tial

[V

]

Time [days]

Test#2 - Measured Potentials with Coated Copper Cathode over Time


Figure 15: Graph of Measured Potentials with Coated Copper Cathode (Test#2)

Table 4: Measured Potentials with No Copper Cathode (Test#3)

Date Measured 6/29 7/6 7/20 8/26 9/21 9/30 10/5

Elapsed Days 0 7 21 58 84 93 98

SB HW on CS Weld-Nut

-0.736 -0.747 -0.727 -0.735 -0.731 -0.732

SB HW on SS Weld-Nut

-0.319 -0.312 -0.426 -0.471 -0.471 -0.474

SS HW on CS Weld-Nut

-0.757 -0.736 -0.742 -0.753 -0.749 -0.748

SS HW on SS Weld-Nut

-0.436 -0.464 -0.468 -0.481 -0.481 -0.481

GS HW on CS Weld-Nut

-1.089 -1.080 -1.059 -1.065 -1.068 -1.064

GS HW on SS Weld-Nut

-1.095 -1.080 -1.049 -1.060 -1.067 -1.054

Temperature [°C]

20.3 20.1 20.2 19.1 17.8 18.6

-1.200

-1.000

-0.800

-0.600

-0.400

-0.200

0.000

0 10 20 30 40 50 60 70 80 90 100

Po

ten

tial

[V

]

Time [days]

Test#3 - Measured Potentials with No Copper Cathode over Time


Figure 16: Graph of Measured Potentials with No Copper Cathode (Test#3)

Physical Corrosion Measurement Procedure:

The mass change due to corrosion was measured by weighing the materials in each galvanic cell

before and after all tests. Additionally, when pitting corrosion was identified, the depth of the pit was

measured because mass change alone does not quantify the risk presented by pitting corrosion.

After testing, all materials were cleaned using tap water and a thick bristle nylon brush (a basic tile

and grout brush), as well as 91% isopropyl alcohol, prior to reweighing. The coated copper was also

cleaned with a power washer (with tap water) to completely remove the paint prior to reweighing.

The coated copper was also weighed after painting, but it was not practical to weigh the straps with

paint after testing due to salt ingress and delamination of the paint. Cleaning corrosion products

from the painted surfaces would have removed paint and rendered the measurement meaningless.

The insulated parts used for galvanic potential measurements were weighed before and after the

tests but are not reported in the tables with mass loss measurements because no mass loss occurred.

Figures 1 and 2 show these parts (bottom flat washer, bottom hex nut, and insulated wire). The

silicone caulk and insulated wires were not measured.

Pit depth was measured by inserting a thin copper wire into the pit and then using calipers to

approximately measure the resulting depth that the wire was inserted. The calipers that were used

measured inch units but were converted to millimeters by calculation.

Formulas for Calculating Mass Loss & Corrosion Depth Rates:

Equations for calculating the mass loss from corrosion:

𝑚∆ = 𝑚𝑖 − 𝑚𝑓 (1)

𝑚∆% =𝑚𝑖 − 𝑚𝑓

𝑚𝑖 (2)

Where:

𝑚∆ is the mass loss (grams)

𝑚∆% is the percent mass loss (percentage) relative to the initial mass

𝑚𝑖 is the initial mass before testing (grams)

𝑚𝑓 is the final mass after testing (grams)

Equations for calculating mass loss rates assume a linear corrosion depth rate based on test duration:

𝑚∆𝑦 =𝑚∆

𝑡𝑡𝑒𝑠𝑡 (3)

𝑚∆%𝑦 =𝑚∆%


Where:

𝑚∆𝑦 is the mass loss per year (grams/year)

𝑚∆%𝑦 is the percent mass loss per year (percentage/year) relative to the initial mass

𝑡𝑡𝑒𝑠𝑡 is the duration of the test (years), where a year is 465.25 days

Equations for calculating average depth of uniform surface corrosion and corrosion depth rate,

assuming a linear corrosion depth rate based on the test duration:

𝑑∆𝑠 =1

2(𝐷𝑖 − 𝐷𝑓) (5)

𝑑∆𝑠𝑦 =𝑑∆𝑠


Where:

𝐷𝑖 is the initial diameter before testing (mm)

𝐷𝑓 is the final diameter after testing (mm)

𝑑∆𝑠 is the average uniform surface corrosion depth (mm)

𝑑∆𝑠𝑦 is the corrosion depth rate per year (mm/year)

Equations for calculating corrosion depth rates assume a linear corrosion depth rate based on test

duration:

𝑑∆𝑦 =𝑑∆𝑝


Where:

𝑑∆𝑝 is the pitting corrosion depth (mm)

𝑑∆𝑝𝑦 is the pitting corrosion depth rate per year (mm/year)

NOTE: It is important to realize that corrosion depth rates are not necessarily linear over the service

life of equipment, but this is a typical method to quantify corrosion in an intuitive way.

Corrosion Measurements - Mass Loss & Rates:

Tables 5 through 15 list the mass loss and corrosion measurements. To make the tables easier to

read, parts which recorded zero mass loss have their mass loss values in gray.

The duration of all tests was 98 days, 0.268 years, or approximately 2,350 hours.

The AISI 1020 carbon-steel weld-nuts exhibited uniform surface corrosion, whereas 303Se stainless-

steel weld-nuts exhibited pitting corrosion in the crevice where they mated with the flat washer.

Photos of these weld-nuts after testing are shown in Figures 20 through 24.

The galvanized-steel bolt threads were very difficult to use after testing, especially with the 303Se

weld-nut. All other bolt materials could be removed easily after testing. This parameter is not

quantifiable, but hardware functionality over the life of the equipment is important if maintenance is

required at some point.

Table 5: Mass Loss Measurements for Silicon-Bronze Hardware with Carbon-Steel Weld-Nut

Test#1 Bare Copper Test

Initial Mass 𝑚𝑖 [g]

Final Mass 𝑚𝑓 [g]

𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]

Silicon-Bronze Bolt 63.88 63.88 0.00 0.0% 0.00 0.0%

Silicon-Bronze Lock Washer 11.74 11.74 0.00 0.0% 0.00 0.0%

Silicon-Bronze Flat Washer 10.19 10.19 0.00 0.0% 0.00 0.0%

Carbon-Steel Weld-Nut 58.94 56.52 2.42 4.1% 9.02 15.3%

Bare Copper Strap 111.26 111.24 0.02 0.0% 0.07 0.1%

Test#2 Coated Copper Test



𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]





Coated Copper Strap (Bare) 111.58 111.54 0.04 0.0% 0.15 0.1%

Test#3 No Copper Test



𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]





The carbon-steel weld-nut exhibited uniform surface corrosion on the outer surface only. Table 3

lists the measured material depth lost due to corrosion:

Table 6: Corrosion Measurements for Silicon-Bronze Hardware with Carbon-Steel Weld-Nut

Silicon-Bronze Hardware Carbon-Steel Weld-Nut

Corrosion Depth 𝑑∆𝑠 [mm]

Corrosion Depth Rate 𝑑∆𝑠𝑦 [mm/year]

Test#1 (Bare Copper) 0.318 1.18

Test#2 (Coated Copper) 0.191 0.71

Test#3 (No Copper) 0.102 0.38

Table 7: Mass Loss Measurements for Silicon-Bronze Hardware with Stainless-Steel Weld-Nut




𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]




303Se Stainless-Steel Weld-Nut 67.19 66.15 1.04 1.5% 3.88 5.8%





𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]









𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]





Pitting corrosion on the surface of weld-nut was found in the crevice where it mated to the flat

washer. Table 5 lists corrosion depth measured:

Table 8: Corrosion Measurements for Silicon-Bronze Hardware with Stainless-Steel Weld-Nut

Silicon-Bronze Hardware Stainless-Steel Weld-Nut

Corrosion Depth 𝑑∆𝑝 [mm]

Corrosion Depth Rate 𝑑∆𝑝𝑦 [mm/year]




Table 9: Mass Loss Measurements for Stainless-Steel Hardware with Carbon-Steel Weld-Nut




𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]

304 Stainless-Steel Bolt 52.49 52.49 0.00 0.0% 0.00 0.0%

304 Stainless-Steel Lock Washer 5.13 5.13 0.00 0.0% 0.00 0.0%

304 Stainless-Steel Flat Washer 6.39 6.39 0.00 0.0% 0.00 0.0%






𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]









𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]







Table 10: Corrosion Measurements for Stainless-Steel Hardware with Carbon-Steel Weld-Nut

Stainless-Steel Hardware Carbon-Steel Weld-Nut






Table 11: Mass Loss Measurements for Stainless-Steel Hardware with Stainless-Steel Weld-Nut




𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]









𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]









𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]





Pitting corrosion on the surfaces of both the flat washer and weld-nut was found where they mated.

Table 9 lists corrosion depths measured:

Table 12: Corrosion Measurements for Stainless-Steel Hardware with Stainless-Steel Weld-Nut

Stainless-Steel Hardware Stainless-Steel Weld-Nut






Stainless-Steel Hardware Stainless-Steel Flat-Washer





Test#3 (No Copper) 0.000* 0.00*

*some corrosion occurred, but it was less than the calipers could measure

Table 13: Mass Loss Measurements for Galvanized-Steel Hardware with Carbon-Steel Weld-Nut




𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]

Galvanized-Steel Bolt 60.79 60.60 0.19 0.3% 0.71 1.2%

Galvanized-Steel Lock Washer 6.28 6.19 0.09 1.4% 0.34 5.3%

Galvanized-Steel Flat Washer 1 17.52 17.28 0.24 1.4% 0.89 5.1%









𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]












𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]










Table 14: Corrosion Measurements for Galvanized-Steel Hardware with Carbon-Steel Weld-Nut

Galvanized-Steel Hardware Carbon-Steel Weld-Nut





Test#3 (No Copper) 0.000* 0.00*


Table 15: Mass Loss Measurements for Galvanized-Steel Hardware with Stainless-Steel Weld-Nut




𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]












𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]








Coated Copper Strap 111.69 111.66 0.03 0.0% 0.11 0.1%




𝑚∆

[g]

𝑚∆%

[%]

𝑚∆𝑦

[g/year]

𝑚∆%𝑦

[%/year]








Pitting corrosion was not observed on the stainless-steel weld-nut for these tests; however, the bolts

were no longer functional and could not be easily removed.

Photos of Hardware Samples After Testing:

The jars contained a substantial amount of corrosion product by the end of each test. Photos of the

jars containing the galvanic cells after testing are shown in Figures 17 & 18 for Test#1 (bare copper

cathode) and Figure 19 for Test #2 (coated copper cathode); however, a photo of the jars for Test#3

(no copper cathode) was not taken.

Figures 20 through 22 show the hardware samples from all tests side-by-side.

Figure 23 shows the pitting corrosion the stainless-steel weld-nuts experienced.

Figure 24 shows the surface corrosion the carbon-steel weld-nuts experienced. Note that the wavy

pattern at the bottom of each weld-nut corresponds to the irregular silicone caulking that was applied

to insulate the bottom of the weld-nuts (for galvanic potential measurements).

Figure 18: Test#1 Carbon-Steel Weld-Nut Galvanic Cells after Testing

Figure 17: Test#1 Stainless-Steel Weld-Nut Galvanic Cells after Testing

Figure 19: Test#2 Galvanic Cells after Testing Stainless-Steel Weld-Nut (Left) & Carbon-Steel Weld-Nut (Right)

Figure 20: Test#1 (Bare Copper Cathode) Hardware Samples after Testing

Figure 21: Test#2 (Coated Copper Cathode) Hardware Samples after Testing

Figure 22: Test#3 (No Copper Cathode) Hardware Samples after Testing

Figure 23: Stainless-Steel Weld-Nut Pitting Corrosion Test#1, Test#2 & Test#3 (Left to Right)

Silicon-Bronze Bolts & Washers Top Row, Stainless-Steel Bolts & Washers Bottom Row

Figure 24: Carbon-Steel Weld-Nut Surface Corrosion Test#1, Test#2 & Test#3 (Left to Right)

Silicon-Bronze Bolts & Washers Top Row, Stainless-Steel Bolts & Washers Bottom Row

Opportunities for Improvement in Test Methods:

These tests were meant to explore potential test methods, and they could be improved.

Improvements could establish more rigorous test methodology that better represents underground

service environments for submerged equipment with different levels of corrosive chemistry in the

electrolyte. An incomplete list of improvements and expansions of the test methods presented in this

report include:

• Test enclosure materials that are assembled and welded as they are actually used on

equipment.

• Selection of more representative materials, such as “copper-bearing” carbon-steel (low-alloy

carbon steel with copper content ≥0.2%), AISI 409L, AISI 304L, AISI 316L, etc.

• A more representative ratio of surface areas between anode (steel enclosure) and cathode

(copper) materials.

• Better control of electrolyte chemistry: For instance, de-ionized water could be used instead

of tap water to control and test different electrolyte chemistry. Alternately, naturally occurring

water from equipment vaults could be used (with appropriate chemical analysis), or these

experiments could be done in the vaults themselves.

• More precise measurement devices: Obviously a more precise mass scale could be used. A

more precise method of measuring pitting corrosion should be established also.

• Tests to quantify the value of sacrificial anodes or other methods of cathodic protection could

be established.

• Corrosion experiments in jars are not perfect representations of service environments. The

best place to perform these experiments is in actual service locations; however, that may not

be a realistic proposal for many reasons. The closer a test can be to real-world applications,

the more useful the resulting data will become.

• Coated copper cathodes (or any cathode material) in future testing should use a different

coating method. The coating base-coat used in this test was used specifically for easy

removal to measure mass loss after testing, which likely resulted in more paint delamination

than would normally be seen. Based on the negligible corrosion on the coated copper it is

believed that weighing a coated cathode after the test does not provide much value and could

be omitted in future testing.

Comments on Galvanized-Steel Testing:

Galvanized-steel was tested to see how it would perform out of curiosity to validate assumptions. No

subsurface equipment standards currently permit galvanized hardware, and it was expected to

perform badly. As seen in the galvanic potential measurements, as well as the photos, the zinc

coating was rapidly consumed due to corrosion. If left in saltwater for years, galvanized hardware

would perform no better than carbon-steel (which is a problem because hardware often cannot be

painted to the same high quality as a plate of carbon-steel on an equipment enclosure). Since this

testing was done as a curiosity, further comments about this material are not included in the

conclusion.

Conclusions:

The test methods evaluated in this report were substantially different than the enclosure material

testing required in IEEE C57.12.32-2019.1 Preliminary testing based on IEEE C57.12.32-2019

indicated that AISI 304L (and 316L) experienced negligible corrosion.2 The same materials and

geometry were not specifically tested is these series of tests, but the 300-series stainless-steel grades

that were tested did experience appreciable corrosion in the presence of a bare copper cathode when

submerged in an electrolyte. As already stated, these tests are not equivalent, but the 303Se and

304 stainless-steel grades that were tested exceeded the 2.5% mass loss evaluation criteria

established in C57.12.32-2019.

Methods for measuring pit-depth should be improved, but the pitting and crevice corrosion found on

stainless-steel was substantial. In fact, the pitting and crevice corrosion depth measured on the

stainless-steel samples was 7 to 12 times deeper (depending on bolt material) versus AISI 1020

carbon-steel (which does not have a copper content ≥0.2%, i.e. “copper-bearing”). Pitting and

crevice corrosion of 300-series stainless-steel alloys in salt-water is well established in technical

literature.4,5

It is also noteworthy that coating the copper cathode resulted in at least a 50% reduction in

corrosion for carbon-steel. A higher quality paint system (than was tested) could improve that

further. In theory, it appears the upper limit is a 67% to 78% reduction in corrosion, which is what

was measured when there was no copper present. This data supports that painting other exposed

metals in an equipment vault could be a very cost-effective method of reducing corrosion that has

not been addressed in any IEEE standards.

Table 16 summarizes and compares the percent mass loss and annual corrosion depth rates for the

various materials and tests performed.

The methods and data presented in this report are not meant as a conclusion. Rather this report is

meant to show that further work is required to establish test methods that will more realistically

represent actual service environments that equipment may experience in underground vaults. A list

of potential improvements on test methods and other ideas are provided. Further, this report seeks

to quantify that coating (painting) other bare metals in an underground equipment vault can

substantially reduce corrosion of equipment enclosures carbon-steel alloys at dramatically lower costs

than changing to the equipment enclosure material to a stainless-steel alloy. Additionally, this report

also reinforces that stainless-steel enclosures are susceptible to pitting corrosion, which can be

deeper than uniform surface corrosion (found on carbon-steel), unless other mitigation methods are

employed.

Table 16: Summary and Comparison of Mass Loss & Corrosion Depth Rates for Materials & Tests

Bolt & Washer

Hardware Corroded Part Test

%Mass Loss 𝑚∆%

[%]

Corrosion Depth Rate 𝑑∆𝑠𝑦 or 𝑑∆𝑝𝑦

[mm/year]

Mass Loss Reduction vs Bare Copper

Corrosion Depth Rate

Reduction vs Bare Copper

UNS C65100 Silicon-Bronze

AISI 1020 Carbon-Steel Weld-Nut

Bare Copper (Test#1)

15.3% 1.18

Coated Copper (Test#2)

7.7% 0.71 50.0% 40.0%

No Copper (Test#3)

4.9% 0.38 67.7% 68.0%

UNS C65100 Silicon-Bronze

AISI 303Se Stainless-Steel Weld-Nut


5.8% 15.05


2.1% 11.27 63.3% 25.2%

No Copper (Test#3)

1.1% 6.06 80.7% 59.7%

AISI 304 Stainless-Steel

AISI 1020 Carbon-Steel Weld-Nut


14.6% 1.37


6.7% 0.80 54.0% 41.4%

No Copper (Test#3)

3.2% 0.33 78.3% 75.9%


AISI 303Se Stainless-Steel Weld-Nut


3.4% 9.66


2.6% 7.01 22.6% 27.5%

No Copper (Test#3)

0.3% 5.11 90.1% 47.1%


AISI 304 Stainless-Steel Flat Washer


9.5% 0.19


4.9% 0.09 48.0% 50.0%

No Copper (Test#3)

0.0%* 0.00* ≈100.0%* ≈100.0%*


References:

1. IEEE C57.12.32-2019, IEEE Standard for Submersible Equipment—Enclosure Integrity

• See §4 for substrate test requirements

• See §4.1 for control test panel materials

• See §4.3.2 test requirements and evaluation criteria

2. Elliott, Will & Dauzat, Tom, SNTP Task-Force Report on Salt-Spray Testing based on

C57.12.32-2019, 2020

3. Elliott, Will, SNTP Task-Force (Draft) Field Corrosion Measurement Procedures, 2020

4. Fontana, Mars G., Corrosion Engineering, 1987

• Galvanic Corrosion: pages 41-51

• Grade 304 & 316 pitting susceptibility in seawater: page 73

5. Roberge, Pierre R., Handbook of Corrosion Engineering, 2000

• Austenitic pitting susceptibility: pages 364-365

• Copper-bearing carbon steel: pages 737, 746-747

REVISION#1: The term “Corrosion Rate” was replaced with the term “Corrosion Depth Rate” per

the feedback from the Task-Force in the Fall 2020 meeting to clarify that this value

is independent of mass loss.

report on corrosion measurements for material

Documents