What You Need to Know About MMO Coated Metal

225 downloads 0 Views 2MB Size Report
size, life, current rate or manufacturer's name without understanding the characteristics, limitations and evaluation method of the MMO anodes. In addition, there ...
Paper No.

2107

What You Need to Know About MMO Coated Metal Anodes

Miki Funahashi, PE MUI International Co. LLC 831 Marrones Court West Chester, PA 19382 U.S.A. [email protected]

ABSTRACT The excellent stability of mixed metal oxide (MMO) coated titanium anodes have been used in the cathodic protection industry for almost 30 years. MMO coated titanium anodes are used in various environments, including seawater, brackish water, fresh water, carbon backfill, and concrete. However, many structure owners and CP engineers simply specify MMO coated titanium anodes by size, life, current rate or manufacturer’s name without understanding the characteristics, limitations and evaluation method of the MMO anodes. In addition, there is some confusion in the industry regarding the breaking potential of MMO titanium anodes. This paper discusses the background, characteristics, limitations and evaluation methods of MMO coated titanium anodes and the breaking potentials. Key Words: Mixed Metal Oxide, MMO coating, titanium anodes, iridium, ruthenium, breaking potential, titanium conductor bar, cathodic protection, titanium cable, concrete, pH of electrolyte

INTRODUCTION The uses of mixed metal oxide (MMO) catalytic coated titanium anodes have been growing over the last three decades. Many cathodic protection design engineers specify MMO titanium anodes without understanding the detailed characteristics, limitations and evaluation methods and just follow the manufacturers or distributors’ data sheets which simply show the sizes, current ratings, life etc. In many cases, some of the anode data sheets are exactly the same as those of other companies even though the MMO anodes’ manufacturers are different. ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

1

The lack of knowledge on MMO coated titanium anodes by CP design engineers and owners often results in the premature failure of the anodes. In many instances, CP design engineers and owners are also misled by the information supplied by the anode suppliers. Since most anodes are used in soil or concrete for a long time (e.g. 20 to 40 years in soil, 75 to 100 years in concrete), it is difficult to track their long term performance. In addition, there is some confusion in the breaking potential of MMO titanium anodes. Some examples of the “misunderstandings” are:    

The breaking potential of MMO titanium anode is the same as bare titanium. The breaking potential of MMO titanium anode is the same as the voltage between the anode and the cathode structure. The maximum voltage of transformer rectifier should be the same as the breaking potential. The breaking potentials of all MMO titanium anodes are the same for any CP system.

BACKGROUND Based on the good corrosion resistance of titanium, the use of titanium as an electrode was beneficial to the various electrochemical and CP industries. Titanium based impressed current cathodic protection anodes were originally proposed by covering with platinum for used in seawater in 1957, but its use was severely restricted by its high cost.1 Another concern of MMO titanium anode was the breakdown of titanium when it is operated at high DC voltage. Titanium possesses a wide potential window for the passive condition. When the bare titanium polarizes and exceeds the upper limit of the anodic potential, intensive oxygen evolution of the titanium surface occurs, resulting in passive film break-down and dissolution (pitting corrosion) of titanium. The breaking (polarization) potential of titanium oxide formed on the bare titanium surface is about 8 to 9 volts (CSE) in seawater and brackish water.2 In 1958, J.B. Cotton attached platinum foil and wire to a titanium strip by spot welding and tested for the feasibility as an anode in seawater. When it was operated at current of 2.23 amp/m2 (24 amps/ft2) with 15 volts, the titanium conductor was not be broken.3 MMO coatings for cathodic protection are mainly divided into two types, iridium oxide (IrO2-x) and ruthenium oxide (RuO2-x). In some case, a mixture of these is used. In addition, titanium oxide (TiO2-x) and tantalum oxide (Ta2O5) are also used as the bulk materials. Henri Beer developed Ruthenium (Ru) MMO coating in 1960’s and more stable Iridium (Ir) MMO coating in 1972.4 Iridium bases MMO coating is more used in electrolyses to evolve oxygen. MMO coating is applied on titanium substrate as a liquid form of metals salt and then is thermally decomposed to form an adherent layer of mixed oxides at temperatures typically ranging 400 to 600°C. The MMO coatings are applied in many layers and heat treated after each coat. Ruthenium based MMO coating was initially developed to produce hypo chlorite for the chemical industry. Iridium based MMO coating is generally used in highly acidic environments, such as electro galvanizing, anodizing of aluminum, recovery of metals, electro synthesis and membrane processes, precious metal plating, etc. which rapidly fails due to the instability of RuO2. One of the purposes of MMO coating on titanium substrate is to reduce the anodic potential of the titanium, so that the MMO coated titanium anode can be operated at higher voltages without break down. Designing Long Life MMO Titanium Anodes The life of MMO coated titanium anode is largely dependent on the following factors:  

MMO coating thickness Iridium or ruthenium based MMO

©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

2

   

Iridium or ruthenium content Coating uniformity Anode current density Type of electrolyte used

Iridium based MMO coating generally has a longer life than ruthenium type when the same amount of current density is discharge in a same electrolyte. However, the cost of iridium is much higher than that of ruthenium. In the cathodic protection industry, ruthenium MMO coated titanium anodes are typically used in seawater as the evolution of chlorine gas is no concern. In addition, the consumption rate of ruthenium is less in seawater than other electrolytes. However, when ruthenium based MMO is used in coke, freshwater, soil or sand, the consumption rates become higher, so that iridium based MMO is more used. When current discharges to high electrical resistivity electrolyte, such as sand, the consumption rate is much higher than that in wet electrolytes at the same anode current density. In addition, when chlorine gas evolution is a concern or long anode lives, iridium based MMO coating is required. Typically, 20 to 30 years of the anode life is specified in soil and seawater. However, longer life MMO coated titanium anodes can be readily produced by thicker coating with higher content of iridium oxide. Therefore, some projects require longer anode life based on this advantage of MMO titanium anode. However, when the long design life anodes are specified, CP design engineers do not consider the life of cable connecting to the anodes. Typical copper cable insulators used for cathodic protection anodes are HMWPE, RHH-USE, and XLPE/PVC. They are to protect the copper conductor from corrosion. In aggressive environments, PVDF/HMWPE and Kyner/HMWPE insulators are sometimes specified. In general, the cable life expectancy in underground is about 30 years.5, 6, 7,8 However, copper cables prematurely fail due to many factors, including: 1. Existing cracked in the insulator caused by coiling processes 2. Physical damages during the installation 3. Penetration of chlorine gas evolved from the anode surface into the cable 4. Development of micro-channel voids between the insulator and copper conductors8 5. Chemical attack by hydro-chloric acid caused by chlorine evolution by the anodic reactions 6. Improper workmanship on the cable connections 7. Poor durability of the splicing material and method 8. Wrong insulator type for a particular environment 9. Freeze-thaw damage to the insulator located near the ground surface In addition, when the cable is installed in dry soil or concrete, the heat generated by the CP current does not dissipate due to the low thermal conductivity. As a result, the cable temperature increases with time, causing premature failure of the insulator. As shown in Figure 1, once the insulator fails, the copper conductor discharges direct current (DC) and consumes in a short period of time due to the electrolysis corrosion. When chlorine gas evolves at the anode-electrolyte interface, it intrudes under the insulator of the cable from any damaged area or poorly insulated cable splice. Therefore, conventional copper cables have high risk of premature failure. The problem is that when the cables fail deep in soil or concrete, it is not possible to be replaced or repaired them even though the anodes are still intact.

Figure 1: Failures of copper cable by acid/chlorine attack and splice ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

3

In addition, when long life anodes need to be used, any types of copper cables cannot be satisfied the condition. In many cases, 100 years design life for cathodic protection anode systems is common for new concrete structures. This problem is overcome by flexible “titanium cable” as shown in Figure 2. However, the electrical resistivity of titanium is about 30 times higher than copper, so that the voltage drop must be considered.

Figure 2: Titanium cored cable with XLPE/PVC insulator MMO Titanium Anode in Concrete When oxygen evolution is prefer as anodic reaction in high pH concrete electrolyte (pH=13), iridium based MMO coating is used because chlorine evolution is the serious concern for the anodes embedded in concrete. When the anode is operated at a low current density (less than 110 mA/m 2 or 10 mA/ft2), the evolved chlorine gas diffuses away through the surrounding porous concrete before turning it into the acid. When sufficient chlorine gas evolves at the anode-concrete interface, it turns to hypochlorite and hydro-choric acid. As a result, the acid dissolves the cement paste (electrolyte of concrete), leaving only non-conductive aggregates (sand and stones). Acid attack is typically shown by brown stains around the embedded anodes (Figure 3). As a result, the anodeconcrete resistance significantly increases in those areas, and CP current is diminished. This condition is severe for ruthenium based MMO anodes.

Figure 3: Acid generation (brown stains) along the MMO titanium ribbon mesh in slots In addition, many projects require a minimum of 75 years anode life, or most MMO anode data sheets show 75 to 100 years. For MMO anodes to last such a long period of time, iridium based MMO is preferred because ruthenium oxides (RuO2) is more oxidized to RuO4 than iridium oxide by the oxygen evolution reaction.9 In CP anode design, bare titanium conductor bars are used to connect MMO titanium anodes and brought to junction boxes outside the concrete, as shown in Figure 4. This is due to the limited life of copper cables and for easier trouble shootings. However, in some projects, copper cable anode connections in concrete are still specified. Since the cable in concrete cannot be repaired or replaced, the anode system will prematurely fail much earlier than the long anode life. In addition, ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

4

rebar (structure) ground cables are also embedded in concrete, the long life of the ground cables are also as important as the anode system. However, most projects do not address this condition.

Figure 4: Titanium conductor bar for the MMO titanium anode-cable connection MMO Titanium Anode Life Acceleration Tests in Acid Anodes in soil and water For most MMO anode manufacturers, life acceleration tests are used to estimate the expected life of their MMO coatings and determine the quality of their batches of MMO coatings produced for their quality control purpose. The accelerated life testing is intended to provide a measure of the anode’s ability to perform satisfactorily for a specific design requirement in a relatively short period of time. Normally, the design requirements for an anode of a given surface area are expressed as a current rating (amperes) for a system life (years). The acid used for this test is generally sulfuric acid (H2SO4) or sodium sulfate (NaSO4) solution because the acidic reaction on the MMO anode surface is oxygen evolution, so that the heavy current load caused by very high anode current density to the MMO coating can be minimized. Common design requirements for most MMO coated titanium anode in a non-concrete CP system are 100 to 150 A/m2.10 Therefore, the amount of MMO coating and its chemical compositions are determined based on the application and the anode life specified. NACE published TM108-2008 (previous version of TM108-2012) “Testing of Catalyzed Titanium Anodes for Use in Soils or Natural Waters.” This test method accelerated the time-to-failure by operating the anode at higher current density than the application’s design requirements in acid. However, NACE revised TM108 in 2012 as non-mandatory testing.10 One of the reasons is that when MMO coating contains ruthenium, if any, the anode fails in a short period of time because ruthenium in the MMO coating dissolves in the acid solution. Therefore, the life acceleration test in acid solution is only applicable to iridium based MMO coated anodes. In addition, when the test data of one MMO anode is compared with the test data of one to another manufacturer, it may lead to erroneous results. Therefore, new NACE TM108-2012 indicates that this adjustment is required when the different current density is used in design. Therefore, this new standard is the design method of MMO anode life based on the standard MMO manufacturer’s anode data (typically 20 years). In other words, higher anode current density shortens the anode life. The MMO coating life for any types of anodes can be readily changed by:  MMO coating thickness: MMO coating thickness is adjusted by multiple coats, dipping or brushing. For example, a manufacturer uses 6 coats for a sample for testing but only applies only 3 coats for a project.

©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

5

 Iridium content of MMO coating: If iridium content is reduced, the anode life is reduced. For example, a manufacturer makes MMO coating with 10% iridium for a sample for testing but may be 5% for the anodes or replace with ruthenium for a project.  Change to ruthenium oxide: MMO coating life is reduced when it is used in oxygen evolution condition, such as fresh water, brackish water, and carbon.  Uniformity of MMO coating: When small amount of samples are made, they can apply the MMO coating uniformly. However, when large amount of anodes are produced, the MMO coating application may become sloppy, resulting in less uniformity of the MMO coating. This will cause local anode failure before it reach the design life. The differences of these conditions cannot visually be distinguished. Therefore, the anodes which arrive a project site may not be the same as the data sheet. Anodes in concrete The acceleration life test method for MMO coated titanium anode which is embedded in concrete is described in NACE TM0294, “Testing of Embeddable Impressed Current Anodes for Use in Cathodic Protection of Atmospherically Exposed Steel-Reinforced Concrete.”11 This test is performed using Sodium Chloride (NaCl) Solution, Sodium Hydroxide (NaOH) Solution, and Simulated (Concrete) Pore Water in Sand. The anode is to be operated at a current density of 0.89 milliamp/cm2 (0.827 amp/ft2) for at least 180 days. Such operation results in a total charge of 3,580 amp-hr/ft2. This amount of charge is also equivalent to the amount of charge sustained by the anode when operated at full design current density 110 mA/m2 (10 mA/ft2) for a period of 40 years. However, it should be noted that TM0294 test method would be used to qualify anodes for service because the testing time is too long to be used for quality control. NACE Committee T-3K-6 began its work to develop a Standard Test Method using H2SO4 solution. However, the committee was uncomfortable with the use of H2SO4, as it was so different from the environment in concrete. However, if NACE TM0294 testing is only used to determine the anode manufacturer’s capability, the MMO anodes which are shipped to a job site may not be the same as the MMO coating which was tested a few years ago. Therefore, short-term acid tests are important to evaluate the anodes which are supplied any manufacturer because the MMO coating compositions and the coating thickness can readily be changed for any reasons. If the acid test is used for MMO anode for concrete, it takes only 26 hours equivalent to 40 years anode life when 1450 amp/m2 of anode current density is used. The typical test set-up is shown in Figure 5 (left). An example of the life acceleration test results in acid for MMO titanium anode for concrete is also shown in Figure 5 (right). This test result can also be compared with the sample test result which is provided by the manufacturer with your test results of the anodes on job site.

Figure 5: Typical set-up for life acceleration test for MMO anode in acid and test results ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

6

Breaking Potentials of Bare Titanium Anodes in soil and water When the potential of bare titanium across the titanium oxide (passive) film exceeds a critical value, it breaks down, resulting in pitting corrosion at localized areas of titanium. This is called “Breaking Potential” or “Breakdown Potential.” The breaking potential of bare titanium is reported about 8 to 9 volts (CSE) in seawater and brackish water.2 Another literature shows that the breaking potential of titanium is 12 to 14 volts in chloride solution.12 In high alkaline electrolyte, the breaking potential of titanium increases as high as 15 to 25 volts.13 When titanium is covered with MMO coating, the breaking potential is not the same as that of the bare titanium any more. This misunderstanding statement is found in EN 12474, Annex E, EN 13173 Annex C, EN 13174 Annex C14; “MMO titanium anode should not be operated at a potential above 8V when the anode is not fully platinised (MMO coating). This is the based on the consideration that to avoid the dissolution of titanium at uncoated locations on the anode surface.” Embedded MMO titanium anodes in concrete Many CP specifications for concrete structures using MMO titanium anodes indicate the maximum transformer rectifier voltage is 20 volts. However, this voltage does not have any justification. The breakdown potential of MMO titanium anodes used in concrete quite different from the anodes in water or soil because bare titanium conductor bars to feed the current. Therefore, the breakdown condition of both MMO titanium anode and titanium conductor bars must be considered. Typical anode configuration using MMO titanium anode systems for concrete is shown in Figure 6.

Figure 6: Typical anode layouts for ribbon mesh (left) and discrete anodes (right) in concrete

EXPERIMENTAL SECTION MMO Coating Analyses MMO coatings on titanium anodes were analyzed by Energy Dispersive X-ray Fluorescence (EDXRF) and Electron Probe Micro-Analysis (EPMA) methods. EDXRF spectroscopy can be used to determine the chemical and/or elemental composition. EPMA is a fully qualitative and quantitative method of non-destructive elemental analysis of micron-sized volumes at the surface of materials, with sensitivity at the level of ppm. EPMA method provides more accurate results than EDXRF. The samples tested were one MMO titanium rod anode for coke backfill (Sample C), one MMO wire anode for fresh water (Sample CD), and three ribbon mesh anodes for concrete (Samples E, M and D). All samples are commercially available and used worldwide. The chemical compositions of MMO coatings were analyzed for iridium and ruthenium. The samples are shown in Figures 7 and 8. ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

7

Figure 7: MMO titanium wire anode for fresh water (left) and MMO titanium rod anode for coke backfill (right)

Sample E

Sample M

Sample D

Figure 8: MMO titanium ribbon mesh anode samples for concrete Life Acceleration Tests Using Acid Solution The life acceleration tests were conducted for the MMO titanium ribbon mesh (Samples E, M and D). Since the thickness of the MMO coating for the anodes used in soil or water generally is much greater than those of the ribbon mesh anodes, the test period is much longer. Therefore, Samples C and CD were not tested at this time. Each ribbon mesh anode with a titanium cathode were immersed in 2.0 litters of 150 g/l of sulfuric acid (H2SO4) at ambient solution temperatures (25 – 35°C) in a 2.5 litters beaker. The sample sizes are 12.7 mm (W) x 152 mm (L) for Samples E and M and 20 mm (W) x 100 mm (L) for Sample D to get the anode surface area of 0.005 m2. In Sample E, two samples were cut from various portions of one roll (76 m) of ribbon mesh based on the color darkness. One sample for Sample D and Sample E was randomly cut and tested because the color was more uniform. Titanium bar (12.7 mm W x 0.9 mm thick) was used as the cathode. 5.9 amperes of the current was applied to each sample at the constant current mode of a DC power supply. The cell voltage between the anode and the cathode was monitored every 10 minutes by a data logger. The solution was not agitated during operation. The test was stopped when the cell voltage showed a sharp increase because it is indicative of the failure. Breaking Potentials of Bare Titanium and MMO Coated Titanium Anodes Bare Titanium and MMO coated titanium breaking potentials were investigated in fresh water, NaCl solution and saturated calcium hydroxide (Ca(OH)2) solution. 2,000 ml of each solution was used for all tests. The pH of the fresh water and the calcium hydroxide solution was about 7.6 and 12.7, respectively. NaCl salt was added in the solutions to produce various chloride concentrations. The bare titanium is ASTM B-265 Grade 1, and the sizes are 150 mm long by 12.7 mm wide by 0.9 mm thick. MMO anodes were produced using the same size of the bare conductor bar and the 70 mm long portion of each conductor bar was coated with Iridium based MMO coating. The chemical ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

8

composition is the same as Sample M. The 50 mm long portion of the MMO coated area was immersed. The test set-up is shown in Figure 9.

Figure 9: Test set-ups for breaking potential studies of bare titanium and MMO titanium (left) and bare titanium connecting with the same size of MMO titanium anode (right) The following tests were conducted to determine the breaking potentials for bare titanium and MMO coated titanium anode: Bare Titanium 1. Polarization of bare titanium: The titanium specimens were immersed in the following types of solutions:       

Fresh water (pH=8) 1% NaCl solution (pH=8) 3% NaCl solution (pH=8) Saturated calcium hydroxide solution (pH=12.6) Saturated calcium hydroxide solution including 0.05% NaCl salt (pH=12.6) Saturated calcium hydroxide solution including 0.25% NaCl salt (pH=12.6) Saturated calcium hydroxide solution including 0.5% NaCl salt (pH=12.6)

The voltage was increased in one volt increments at every 5-minute intervals, and the current and potential of the titanium were monitored. A silver-silver chloride reference electrode (3.5 mol, KCl) was used for potential measurements. The tip of the reference electrode was contacted on the specimen’s surface, as shown in Figure 10 to eliminate the voltage-drop.

Figure 10: Potential measurement of the titanium specimen using Ag-AgCl reference electrode 2. Visual observation of the titanium surface condition under constant voltage: The titanium specimen was immersed in each type of solution and maintained a certain voltage of the power supply until the anode showed any sign of pitting corrosion. In addition, the stability of the potentials was also monitored. ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

9

MMO Coated Titanium 1. Polarization of MMO coated titanium anode: The voltage was increased in one volt increments at every 5-minute intervals, and the current and potential of the anode were monitored. 2. Visual observation of non-coated area of MMO titanium anode: A small portion of MMO coating was damaged and exposed bare titanium substrate, as shown in Figure 11. The MMO anode was immersed in each type of solution and maintained a certain voltage of the power supply. The specimen was visually inspected for any sign of pitting corrosion on MMO coating and at the damaged area.

Figure 11: Damaged MMO coating on titanium anode Field Measurement of MMO Coated Titanium in Seawater To verify the laboratory test results, the MMO titanium anode which has been operated in seawater to protect pipelines for over one year was evaluated. The voltage and current of the operating transformer-rectifier were 15.5 volts and 15 amps. Two anodes are connected by individual copper cable (25 mm2, about 10 m long each) to the rectifier. The ground negative cable is the same size and about 100 m long. The potential of the anode was measured with a seawater Ag-AgCl reference electrode.

RESULTS MMO Coating Analysis MMO Rod and Wire EDXRF test results of Samples C (rod in coke backfill) and CN (wire in water) are shown in Figures 12 and 13, respectively. Since the anode surface of both samples had too small-radius curve surface, EPMA could not be used. Sample C is the anode supplied to use in coke backfill material and Sample CN is in fresh water. In general, MMO coating used in fresh water and carbon backfill is designed similarly due to the similar anode current density. However, the chemical compositions are quite different by the different manufacturers.

Figure 12: Chemical compositions of MMO coating of Sample C by EDXRF ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

10

Figure 13: Chemical compositions of MMO coating of Sample CN by EDXRF MMO Ribbon Mesh for Concrete EPMA test results are shown in Figures 14 to 16 for Samples E, M and D, respectively. Samples E and M showed that the MMO coating used was iridium base. The average intensities of iridium in Samples E and M were 25 and 27, respectively. Sample D contains large amount of ruthenium. The intensity of ruthenium and iridium is 6 and 1, respectively. In the manufacturer’s data sheet for Sample D indicates that the catalysis (MMO) is oxygen evolution even though the MMO coating includes ruthenium oxide. In addition, the anode life is 100 years even though the iridium and ruthenium contents are much lower than Samples E and M, which show 75 years life in their data sheets. As indicated before, ruthenium in MMO coating increases the risk of acid attack to the concrete (dissolution of cement paste) due to the higher chlorine gas evolution and possible shorter anode life.

Mapping of Iridium

Mapping of Ruthenium

Figure 14: Mapping of iridium and ruthenium of MMO coating in Sample E (Greater intensity: blue greenyelloworangered)

Mapping of Iridium

Mapping of Ruthenium

Figure 15: Mapping of iridium and ruthenium of MMO coating in Sample M (Greater intensity: bluegreenyelloworangered) ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

11

Mapping of Iridium

Mapping of Ruthenium

Figure 16: Mapping of iridium and ruthenium of MMO coating in Sample D (Greater intensity: bluegreenyelloworangered) Life Acceleration Tests for MMO Titanium Anodes Figure 17 shows the results of the life acceleration tests on Samples E, M and D in the acid solution. The cell voltage of the darker color portion of Sample E increased about 1 volt in 49 hours which was the equivalent amp-hours for 75 years at 110 mA/m2; however; the cell voltage for the lighter color portion immediately increased and failed shortly. The increase of the cell voltage of Sample M was less than 1 volt in about 49 hours, which was equivalent to 75 years’ life. As found in the EPMA test result, the MMO coating contains high amount of iridium. Sample D failed when 4 years equivalent of current discharged. As indicated in the EPMA, this is a mixture of ruthenium-iridium based MMO coated anode. Even though the data sheet indicates 100 years anode life, MMO coating contained smaller amount of ruthenium and iridium than Samples E and M, which show 75 years life. Even though ruthenium dissolved in the acid, the remaining iridium should handle this test, but the low content of iridium appears to shorten the life.

Figure 17: The life acceleration test results of Samples E obtained from the same roll (left), Sample M (center) and Sample D (right) Breaking Potentials of Bare Titanium and MMO Titanium Anodes Bare Titanium The comparisons between the anode “Potential” and the anode-cathode “Voltage” of bare titanium are shown in Figure 18 in fresh water (pH=7.7) and salt water (pH=7.7, 1% and 3% NaCl) and in saturated calcium hydroxide solutions (pH=12.7, no chloride, 0.05%, 0.25% and 0.5% NaCl).

©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

12

In the neutral pH solutions, the potentials of the bare titanium were typically 20 to 30% lower than the anode-cathode voltages until it reached the breaking potential. When the anode was operated at 15 volts, the anode potential was about 10 to 13 volts depending on the salt concentrations. The anode potential became steady after the titanium exceeded the breaking potential which was about 12 to 13 volts even though the anode-cathode voltage was further increased. In the high pH solutions, similar behaviors were observed. However, the following two conditions were different from those in the neutral pH solution. 

The breaking potential was not observed even though the bare titanium anode was operated at 40 volts which polarized to 35 volts in potential.



The voltage to reach the breaking potential (about 12 to 13 volts) was about 15 volts in the NaCl mixed solutions.

pH = 12.7

pH = 7.6

Figure 18: The comparisons between the anode “Potential” and the anode-cathode “Voltage” of bare titanium in fresh water and salt water (left), in saturated calcium hydroxide solutions (right) MMO coated titanium As expected, the anode potentials were much lower than the anode-cathode voltage regardless of the chloride concentrations in the neutral pH solutions, as shown in Figure 19 (left). When the anode voltage was as high as 40 volts, the anode potential was only about 5 volts in the fresh water and 3% NaCl solution. Even though 60 volts of the anode-cathode voltage was applied in fresh water, the potential of the MMO anode was only 7 volts. When the MMO anode was in the high pH solution, 60 volts of the anode-cathode voltage only could polarize the anode to less than 10 volts in potential regardless of the chloride concentration, as shown in Figure 19 (right).

©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

13

pH = 12.7

pH = 7.6

Figure 19: The comparisons between the anode “Potential” and the anode-cathode “Voltage” of MMO titanium anode in fresh water and neutral pH 3% NaCl solution (left) and in saturated Ca(OH)2 solutions in various salt concentrations (right) This indicates that MMO coating significantly reduces the anodic potentials on the titanium substrate compared with the bare titanium in neutral and high pH solutions. Using a 60 volts DC power supply, the MMO titanium anode could not reach to the breaking potential in this experiment. In most countries, the maximum DC voltage of transformer-rectifiers is 50 or 60 volts due to the safety reason. Therefore, normal CP systems cannot polarize the MMO anodes to the breaking potential. Figure 20 shows the bare titanium behavior in the 3% NaCl solution when the same size of MMO titanium coated anode was electrically connected. The polarization behavior of the bare titanium shifted lower as indicated in Figure 28(left). As discussed above, to polarize the bare titanium to about 8 volt in potential, the anode voltage required was about 11 volts. However, by connecting the MMO anode, the bare titanium required about 24 volts of the voltage to reach to 8.5 volts in potential. Therefore, when large bare titanium surface exists in the MMO coated titanium anode system in the same electrolyte, the bare titanium polarizes less at the same voltage. This was also analyzed based on the current discharging from the bare titanium and the MMO titanium anode, as shown in Figure 28(right). The bare titanium connecting with the MMO anode discharged little current until the bare titanium polarized to 8.5 volts (breaking potential).

Figure 20: Bare titanium behavior in the 3% NaCl solution when the same size of MMO titanium anode was electrically connected (left), the discharging current from MMO titanium and bare titanium when they were electrically connected together (right) ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

14

The same analysis was made for the bare titanium in the high pH solution containing 0.5% NaCl. The results are shown in Figure 21. Similar behaviors were also observed. However, 32 volts of the anode-cathode voltage was required to polarize the bare titanium to 12 volts with the MMO anode, comparing with 13 volts by itself. When the MMO anode discharged about 1.4 amps at 40 volts, the bare titanium discharges little current.

Figure 21: Bare titanium behavior in the saturated Ca(OH)2 solution containing 0.5% NaCl when the same size of MMO titanium anode was electrically connected (right), the discharging current from MMO titanium and bare titanium when they were electrically connected together (right) Anodic Polarization Figure 22 shows the anodic polarizations of bare titanium in the neutral and high pH solutions containing various chloride concentrations. Bare titanium in neutral pH solution When chlorides did not exist in the solution, the titanium did not show the passive film break-down behavior even though its potential reached to 30 volts. On the other hand, the bare titanium showed the passive-film breakdown when the potential exceeded about 12 volts in the solution containing 1% of NaCl and about 8 volts in 3%. Therefore, the breaking potential of titanium appears to decreases with increasing the chloride concentration. Bare titanium in high pH solution When the chloride does not exist in the high pH solution, the bare titanium does not show the passive film breakdown behavior even though the potential reached to 53 volts by the anodecathode voltage of 60 volts.

©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

15

Figure 22: Anodic polarizations of bare titanium in the neutral solution (left) and saturated Ca(OH)2 solution (right) with various chloride concentrations MMO Coated Titanium in neutral and high pH solutions Figure 23 shows the anodic polarizations of MMO titanium anodes in the neutral and high pH solutions. None of the MMO titanium anodes show the breakdown potentials regardless of the chloride concentration. In high pH solution, the maximum 60 volts of the anode-cathode voltage could not polarize the MMO anode more than 11 volts in 0.25% concentration of NaCl and 7 volts in 0.5% solution.

Figure 23: MMO titanium anode-anodic polarization in the neutral (left) and high pH solution (right) Comparisons of Bare Titanium and MMO Titanium Anodic Polarization Figure 24 shows the comparisons of the polarization behaviors of bare and MMO titanium anodes in neutral pH solution containing no chloride and 3% of NaCl. As indicated, the MMO coated titanium anode polarized much less than the bare titanium by the same current density.

©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

16

Figure 24: The comparisons of the polarization behaviors of bare and MMO titanium anodes in fresh water (left) and 3% NaCl solution (right) Anode-to-Electrolyte Resistance of Bare Titanium and MMO Titanium Anode Neutral pH Solution The anode-to-electrolyte resistances of the bare titanium and the MMO titanium anode in the neutral solution were calculated using the anode-cathode voltages and the amounts of the applied current. Since the exposed surface area and the cathode are the same in all experiments, the calculated resistances can be simply compared. Figure 25 shows the calculated resistances in various voltages in fresh water and in 3% NaCl solution. In the fresh water, the resistance of the MMO titanium was about 1/2 to 1/3 lower than that of the bare titanium. However, in 3% NaCl solution, the MMO titanium anode resistance was much lower than those of the bare titanium above 10 volts, which was 70 amp/m2 of anode current density. In general, the maximum MMO anode current density in seawater is 100 to 150 amp/m 2. This range is also shown Figure 25 (right). Therefore, when the MMO titanium anode is normally operated in seawater, the MMO coating provides much lower resistance than any exposed bare titanium area (if exists), so that the bare titanium discharges little current.

Figure 25: Calculated electrode-to-electrolyte resistance in various anode-cathode voltages in fresh water (left) and in 3% NaCl solution (right)

©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

17

High pH solution Figures 26 and 27 show the calculated resistances in the high pH solutions, containing 0.25% and 0.5% of NaCl, respectively. The resistance trends are quite different from those in the neutral pH solutions. In the solution containing 0.25% of NaCl, the resistance of the MMO anode was much less than those of bare titanium alone below 20 volts of anode-cathode voltage. In the higher voltages, the resistance of the bare titanium was getting similar to those of the MMO titanium anode with increasing the voltage due to the passive film break-down. The similar behaviors were observed in the 0.5% NaCl solution. However, the resistance of the bare titanium alone was much greater in the 0.5% solution than those in 0.25%. The resistance of the bare titanium became closer to that of the MMO anode above 15 volts after the passive film breakdown.

Figure 26: Calculated electrode-to-electrolyte resistance in saturated Ca(OH)2 containing 0.25% of NaCl

Expansion View

Figure 27: Calculated electrode-to-electrolyte resistance in saturated Ca(OH)2 containing 0.5% of NaCl Visual Observations Bare Titanium in neutral pH solutions containing no chloride and 3% NaCl Figure 28 shows the potential behaviors of bare titanium with time in fresh water when the anodecathode voltage was maintained at constant voltages of 15, 17 and 20 volts. Pitting corrosion occurred at the end of the titanium electrode at 17 volts in a day, as shown in the right side. The titanium potentials were about 12.5 volts by 17 volts of anode-cathode voltage when the pitting corrosion developed. When the potential was below 12 volts, pitting corrosion did not occur during the 2 days test period and the potential was stable. The anode-cathode voltage was 15 volts. This condition was not observed in the 5-minute interval polarization tests. It seems that it takes some time for bare titanium to start breaking in fresh water. ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

18

When the solution contained 3% of NaCl, the bare titanium failed at 14 volts of the anode-cathode voltage, which was also about 12 volts in potential, as shown in Figure 29. The failure condition is also shown in the right side of the figure. An interesting result was obtained when the bare titanium was connected to the MMO titanium anode. As discussed above, the potential of the bare titanium decreased significantly with the MMO titanium anode together, so that 20 volts of the anodecathode voltage was required to break the bare titanium, instead of 14 volts. However, it is noted that the breaking potential of the bare titanium decreased to 8 volts.

Figure 28: The potential behaviors of bare titanium with time in fresh water when the anodecathode voltage was maintained at constant voltage of 15, 17 and 20 volts, Pitting corrosion occurred at the end of the titanium electrode at 17 volts in 24 hours

Figure 29: The potential behaviors of bare titanium with time in 3% NaCl solution with and

without MMO titanium anode (top), no failure at anode-cathode voltage of 13 volts and14 volts without MMO anode Bare Titanium in high pH solutions containing no chloride free and 3% NaCl When “bare” titanium was impressed at 15 volts in the high pH solution containing 0.5% of NaCl, the potential exceeded the 12.5 volts and failed, as shown in Figure 30. This result consented with the other test results discussed above. However, when the voltage decreased to 14 volts and the potential was just below 12.5 volts, the titanium did not fail for over 60 hours. When the MMO titanium anode was connected to the bare titanium, it did not fail until the anodecathode voltage increased to 36 volts in saturated CaOH2 solution containing 0.5% of NaCl, increasing the potential to about 12.5 volts. When the high pH solution did not contain chlorides, the bare titanium did not fail even though the anode voltage was operated at 50 volts, as shown in Figure 31. ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

19

Figure 30: Bare titanium with time in saturated Ca(OH)2 solution containing 0.5% NaCl with and without MMO titanium anode (left), no failure at anode-cathode voltage of 13 volts (left), 15 volts (center) and 36 volts with MMO anode (right)

Figure 31: No failure of bare titanium in saturated Ca(OH)2 solution containing no chlorides MMO Titanium Anode As long as the titanium anode is 100% covered with MMO coating, it requires a very high voltage to reach the breaking potential of the anode. This could not be achieved in this test because it required more than 60 volt of a DC power supplier with very high current capacity. However, this is not the realistic situation because the maximum voltage of cathodic protection power supply (rectifier) is generally restricted below 60 volts. In practice, during the anode installation, MMO coatings are often damaged and exposing titanium substrate. To find out the condition of bare titanium at the MMO coating holiday, a small portion of MMO coating was damaged to expose bare titanium substrate. These damaged MMO anodes were immersed in 3% NaCl solution and high pH solution containing 0.5% of NaCl. The MMO anodes were impressed at 40 volts (this is the maximum current capacity of this power supply) in the salt water and 60 volts in the high pH solution. As indicated in Figure 32, the exposed bare titanium substrate in the MMO titanium anode at 40 volts in 3% NaCl solution and at 60 volts saturated Ca(OH)2 solution containing 0.5% NaCl after 72 hours of the exposure, respectively. The bare exposed titanium substrate of either anode did not show any sign of failure. By considering all test results obtained above, the bare titanium surface area is a small portion of the MMO coated area, so that the resistance of the bare titanium to the electrolyte is significantly higher than that of the MMO anode area.

©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

20

Figure 32: The condition of exposed bate titanium substrate in MMO coated anode in 3% NaCl solution (top) and in saturated Ca(OH)2 solution containing 0.5% NaCl (bottom) Field Measurement The MMO titanium anode potential which has been operating in seawater by the rectifier (15.5 volts) was only 5.8 volts to a seawater Ag-AgCl reference electrode. Since the CP current output from the rectifier was 15 amps, the calculated voltage drop in the cables is about 1 volt in the 110 m long copper cable (25 mm2 in cross-section area). Therefore, the anodes have never reached to the breaking potential by this set up of the transformer rectifier. This finding was consent with the experience results.

CONCLUSIONS MMO Titanium Anode Life 1. Chemical compositions of MMO coatings are different from manufacturer to manufacturer. The coating thickness and compositions of the MMO coating influence the life of the anode. They are readily adjustable without changing their visual appearance and may be different from their data sheets. Therefore, the anodes which are sent to a project site should be certified by independently testing to confirm that they meet the project specification. 2. To determine the quality of the MMO titanium anodes, the life acceleration test using acid is one of the powerful methods to determine their quality. To conduct the life acceleration test, MMO coating must be iridium oxide base and not contain ruthenium oxide. Because the test set-up is simple, it can be readily done by an independent laboratory or the third party without high cost. 3. When a long life MMO titanium anode is designed, the limited life of copper cable must be considered to meet the anode life. Breaking Potentials of Bare Titanium and MMO Titanium Anode 1. The breaking “Potentials” of the MMO titanium anodes are not the same as the anode-cathode “voltage” or “Bare Titanium Braking Potential.” 2. When the breaking potential of “MMO titanium anode” is a consideration in the design of the transformer rectifier maximum voltage, the following factors need to be considered:  

Voltage drop in both anode and cathode cables Electrolyte, particularly pH and chlorides

©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

21



Length of bare titanium conductor bar if used for the anode connection

3. When “bare titanium” is in chloride free solution, the breaking potential was not observed below 30 volts (Ag-AgCl, 3.5 mol KCl) in neutral pH and 53 volts in saturated Ca(OH)2 solution. 4. When the “MMO titanium anode” is connected to a larger surface area of “bare titanium”, such as conductor bars, the increase of the bare titanium potential occurs. In this case, high voltage between the anode and the cathode may reach the breaking potential of the titanium conductor bar. Breaking Potentials of Bare Titanium and MMO Titanium Anode in Neutral Solutions 1. The breaking potential of “bare titanium” was about 12.5 volts (3.5 mol KCl, Ag-AgCl) in fresh water and 3% NaCl solution. 2. When “bare titanium” was electrically connected to the “MMO titanium anode” which had the same surface area, the breaking potential of the “bare” titanium decreased to 8 volts in neutral pH solution containing 3% of NaCl. 3. “MMO titanium anode” with a small MMO coating holiday exposing the bare titanium substrate did not show the breaking potential condition in 3% NaCl solution even though 40 volts of the anode-cathode voltage was applied 72 hours. 4. The statement in EN 12474, Annex E, EN 13173 Annex C, EN 13174 Annex C; indicating that MMO titanium anode should not be operated at a potential above 8 volts is misleading. 5. When the MMO titanium anode was electrically connected to the bare titanium which had the same surface area, 20 volts of anode-cathode voltage could polarize the bare titanium to the breaking potential (8.5 volts). However, when the solution did not contain chlorides, the bare titanium did not fail even though the anode-cathode voltage was operated at 50 volts. Breaking Potentials of Bare Titanium and MMO Titanium Anode in High pH Solutions 1. The breaking potential of “bare titanium” in saturated calcium hydroxide solution (pH=12.7) containing 0.5% of NaCl was about 12.5 volts. On the other hand, when chlorides do not exist, the breaking potential was not observed below 53 volts. 2. When the “MMO titanium anode” was electrically connected with the “bare titanium” which has the same surface area in saturated calcium hydroxide solution (pH=12.7) containing 0.5% of NaCl, the breaking potential of the bare titanium was not observed even though the anode potential reached to 12 volts with 40 volts of the anode-cathode voltage. 3. “MMO titanium anode” with a MMO coating holiday exposing the bare titanium substrate did not show the breaking potential condition in the solution containing 0.5% NaCl even though it was operated at 60 volts of anode-cathode voltage over 72 hours. 4. When the MMO titanium anode was electrically connected to the bare titanium which has the same surface area, 36 volts of anode-cathode voltage could polarize the bare titanium to the breaking potential (12.6 volts). However, when the solution did not contain chlorides, the bare titanium did not fail even though the anode-cathode voltage was operated at 50 volts.

Recommendations 1. The life acceleration test using acid should be used as QC Test for the MMO titanium anodes provided in a job site if the anode design life is critical to the structure even though the acid test result is conservative. ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

22

2. Titanium cables should be used for anodes which cannot be replaced after the installation. When a long anode cable is required, the voltage-drop needs to be considered. The voltagedrop can be adjusted by using larger titanium cables. 3. When long titanium conductor bars are used in MMO titanium anode systems in concrete containing high chlorides, thin coated MMO conductor bars should be used and embedded in non-conductive material, such as polymer concrete or epoxy. 4. When the failure of MMO titanium anodes by the over-breaking potential is of concerned, measurement of the anode potential with a standard reference electrode should be conducted. REFERENCES

1. J.B. Cotton and B.P. Downing, “Corrosion Resistance of Titanium to Sea-water,” Trans. Inst. Marine Engineering, 69, (8), pp. 311-319, 1957

2. Almar-Naess and J. M. Drugli, “ Prevention of Corrosion in Paper Making Machine,” Platinum Metals Rev., 10, (2), pp. 48-51, 1966

3. J.B. Cotton, “Platinum-faced Titanium for Electrochemical Anodes,” Platinum Metals, Rev., 2, (2), pp. 45-47, 1958

4. MAGNETO special anodes B.V.'s website, http://www.magneto.nl 5. Oklahoma Corporation Commission, Public Utility Division Staff, “Oklahoma Corporation Commission’s Inquiry into Underground Electric Facilities in the State of Oklahoma,” June 2008.

6. Commonwealth of Virginia,” Placement of Utility Distribution Lines Underground,” House Document No. 30, 2005

7. North Carolina Natural Disaster Preparedness Task Force, “The Feasibility of Placing Electric Distribution Facilities Underground,” Nov. 2003

8. University of California, CIEE, “Project Summary: Fault Analysis in Underground Cables,” 2007 9. Marilyn J. Niksa, President Water Star Inc., Dr. Eric J. Rudd, Electrochemical Consulting Inc., “Superior Anodes”, available from http://www.waterstarinc.com/files/Resources/White_paper.pdf

10. NACE TM108-2012, “Testing of Catalyzed Titanium Anodes for Use in Soils or Natural Waters “ Appendix A TM108-2012) “Testing of Catalyzed Titanium Anodes for Use in Soils or Natural Waters.”

11. NACE TM0294, “Testing of Embeddable Impressed Current Anodes for Use in Cathodic Protection of Atmospherically Exposed Steel-Reinforced Concrete.”

12. L. l. Shreir, “Tantalum-Platinum and Titanium-Platinum Bi-Electrodes” Platinum Metals Rev., 4, (1), pp. 15-17, 1960

13. Prusi, L.J. Arsov, B. Haran and B. N. Popov, “Anodic Behavior of Ti in KOH Solutions,” Journal of the Electrochemical Society, 149 (11), B491-B498, 2002

14. BS EN 13174 2001 “Cathodic Protection for Harbor Installations” English Version, January 2001 ©2013 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole, must be in writing to NACE International, Publications Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association.

23