Petrograpic_Exam_CRT_Report_P-2020_Revision_with_Chem_Analysis.pdf

Concrete Research & Testing, LLC • 400 Frank Road, Columbus, Ohio 43207 • 614.443.0085

CRT Report No. P-2020-2

Petrographic Examination of Three Concrete Cores from

Three Storage Tanks at the Cincinnati/Northern

Kentucky Airport

September 18, 2024

Prepared for: S&ME

Cincinnati, Ohio

Project No. 24-78-0095

CRT Report No. P-2020 www.ConcreteTesting.com

INTRODUCTION

Three concrete cores were received from Mark Westermann with S&ME on July 9, 2024. The cores were removed from the three storage tanks at the Cincinnati/Northern Kentucky International Airport in Hebron, Kentucky. The age and mix design of the concrete is unknown.

The tanks are used to store propylene glycol used for deicing aircraft, and reportedly runoff from the runways also enters the tanks. The interior surfaces of the tanks exhibit shallow deterioration. Site photographs of the tanks are shown in Figure 1 and Figure 2 of the Appendix; the photos were provided by

S&ME.

Concrete Research & Testing, LLC was requested by the S&ME to perform a petrographic examination on the received cores to determine the cause of surface deterioration at the inside of the storage tanks. It was reported that the client is specifically concerned about microbially induced corrosion and sulfuric acid attack.

CORE DESCRIPTIONS

The received cores were labeled “Core #1, Tank 1 NE,” “Core #2, Tank 2 NE,” and “Core #3, Tank 3 NE.”

The cores have a diameter of 3.7 in. and were removed from the exterior of the tank walls through the full thickness. Photographs of the cores in the as-received condition are shown in Figure 3 through Figure 5.

Core #1 has a length of 4.3 to 4.4 in. One end of the core represents the exterior surface of the tank, which is coated with multiple layers of paint. The other end of the core represents the interior of the tank which exhibits deterioration and exposed aggregates; the exposed surface has a dark brown color (Figure 3). The core contains 0.2 in. diameter welded wire steel reinforcement located at depths of 1.5 in., 1.9 in., and

2.1 in. from the exterior surface.

www.ConcreteTesting.com

Core #2 has a length of 4.8 in. One end of the core represents the exterior surface of the tank, which is coated with multiple layers of paint. The other end of the core represents the interior of the tank which exhibits deterioration and exposed aggregates. The core contains 0.2 in. diameter welded wire steel reinforcement located at depths of 1.8, 2.0, 2.3, and 2.4 in. from the exterior surface and a No. 4 rebar located 2.6 in. from the exterior surface.

Core #3 has a length of 2.8 to 3.5 in. One end of the core represents the outer surface which is coated with multiple layers of paint. The other end of the core represents the interior of the tank which exhibits deterioration and exposed aggregates. The core contains 0.2 in. diameter welded wire steel reinforcement located 1.2 in. from the exterior surface and the imprint of a perpendicular rebar located 1.3 in. from the exterior surface.

EXAMINATION AND TEST METHODS

A preliminary stereomicroscopic examination was performed on the cores. Epoxy was applied to the interior surfaces of the cores to protect and preserve features of the near-surface concrete during the sample preparation. The cores were then saw-cut perpendicular to the end surfaces; two parallel saw-cuts were made to obtain a 0.75-in.-thick section from the center of each core.

The depths of carbonation were measured on freshly saw-cut surfaces of the cores following the application of phenolphthalein solution.

www.ConcreteTesting.com

The saw-cut sections were prepared for microscopic examination by lapping on a steel wheel with progressively-finer diamond and silicon carbide abrasives. The lapped sections were examined under a stereomicroscope having a magnification range of 3.5x to 220x. Visual and microscopic examinations were also performed on freshly-fractured surfaces of the concrete. The examinations were performed following the relevant guidelines outlined in ASTM C856, “Standard Practice for Petrographic

Examination of Hardened Concrete.”

The water-cementitious ratio (w/cm) of the concrete was estimated based on observations and qualitative assessments of the cement paste, including hardness, textural features, color, abundance of residual cement particles and rate of water absorption. The hardness is determined by probing the cement paste with a sharpened metal tool.

Polished samples were prepared from the near-surface concrete for chemical analysis using a Scanning

Electron Microscope (SEM) with Energy Dispersive X-Ray Spectroscopy (EDS) capability. The EDS analyzer provides a semi-quantitative elemental chemical analysis of the material examined. The elemental data are converted to oxides by assuming stoichiometric cation/oxygen ratios. The SEM/EDS analyses were performed following the relevant guidelines of ASTM C1723 “Standard Guide for Examination of

Hardened Concrete Using Scanning Electron Microscopy.”

www.ConcreteTesting.com

EXAMINATION RESULTS

The concrete represented by the cores is judged to be similar, and the examination results will be presented together. Photographs of the lapped sections prepared from the cores are shown in Figure 6 through

Figure 8.

Coating Materials – The exterior of the tanks is coated with a light brown elastomeric material (see

Figures 3 through 5). In cross section, the tanks have several layers of elastomeric materials over a skim coat of mortar.

In Core #1, the mortar has a thickness of 0.5 mm and contains a quartz sand with a maximum size of

0.3 mm; the mortar is low to poor quality. The mortar is covered with a gray, elastomeric sanded material having a thickness of 0.3 to 1.0 mm. The light brown elastomeric material at the surface does not contain sand and has a typical thickness of 0.2 mm. The coating materials exhibit small areas of bond failure within the skim coat mortar.

In Core #2, the mortar has a thickness of 0.3 to 0.6 mm and contains a quartz sand with a maximum size of

0.3 mm; the mortar is low to poor quality. The mortar is covered with a gray, elastomeric sanded material having a thickness of 0.3 to 1.0 mm and a typical thickness of 0.5 mm. The light brown elastomeric material at the surface does not contain sand and has a thickness of 0.2 to 0.4 mm. The coating materials exhibit moderate areas of bond failure within the skim coat mortar.

In Core #3, the mortar has a thickness of 0.5 to 1.3 mm and contains a quartz sand with a maximum size of

0.3 mm; the mortar is low to poor quality. The mortar is covered with a thin layer of tan elastomeric material having a thickness of 0.1 to 0.2 mm, and a gray, elastomeric sanded material having a thickness of

0.1 to 0.5 mm. The light brown elastomeric material at the surface does not contain sand and has a thickness of 0.1 to 0.5 mm. The coating materials exhibit small to moderate areas of bond failure within the skim coat mortar.

Aggregates – The coarse aggregate is a 0.4 in. maximum size natural gravel. The gravel is dominated by carbonate rock with moderate amounts of chert and sandstone, small amounts of igneous/metamorphic rock and quartzite, and minor amounts of siltstone and shale.

The fine aggregate is a natural sand likely produced from the same source as the gravel. The sand is dominated by quartz particles with small to moderate amounts of carbonate rock particles. The sand also contains small amounts of chert, sandstone, igneous/metamorphic rock, feldspar, siltstone, shale, and quartzite particles; minor amounts of magnetite, pyroxene/amphibole, limonite, and mica particles; and trace amounts of garnet and coal particles.

The fine and coarse aggregates are judged to be good quality and are uniformly distributed throughout the examined concrete.

www.ConcreteTesting.com

Cement Paste – The cement paste in all three cores contains portland cement and fly ash as the cementitious constituents.

The cement paste in the bulk of the cores is judged to be good quality with estimated water-cementitious* ratios (w/cm’s) in the range of 0.42 to 0.47. The cement paste in Core #2 is variable on a microscopic scale with small, localized areas having an elevated w/cm.

*Portland cement + Fly Ash

Air Void Systems – The concretes represented by the cores is air entrained.

Core #1 has an estimated total air void content in the range of 3 to 4 percent. The concrete is generally well consolidated, although a concentration of irregular and elongate entrapped air voids having sizes of 1 to

12 mm are present within 1 to 2 in. of the exterior surface. Although the total air void content is low, the air void system is judged to be adequate quality with regard to freeze/thaw durability.

Core #2 has an estimated total air void content in the range of 7 to 8 percent. The concrete is not well consolidated with abundant irregular entrapped air voids in the size range of 1 to 13 mm. The air void system is judged to be adequate quality with regards to freeze/thaw durability.

Core #3 has an estimated total air void content in the range of 4 to 5 percent. The concrete is well consolidated. The air void system is judged to be good quality with regards to freeze/thaw durability.

Reinforcing Steel – The cores contain No. 4 rebar and 0.2 in. diameter welded wire steel reinforcement.

The steel is not corroded.

Interior Surface Deterioration – The concrete at the interior end surface of the cores exhibits shallow deterioration exposing the near-surface coarse aggregate particles. The cement paste in the near-surface concrete is very absorptive and very weak with little to no inherent strength. Carbonate rock aggregate particles at the surface exhibit shallower degradation. Photomicrographs showing the interior surfaces of the cores in cross-section are shown in Figure 9 through Figure 11.

In Core #1, the cement paste is poor quality to depths of 0.5 to 1.0 mm (see Figure 9). Areas of the underlying cement paste are weakened to a depth of 2 mm and is judged to be fair quality.

In Core #2, the cement paste is very poor quality and has an orange color to depths of 0.5 to 1.5 mm (see

Figure 10). The transition to good quality cement paste is abrupt.

In Core #3, the cement paste is poor quality to depths of 0.5 to 1.5 mm (see Figure 11). The transition to good quality cement paste is abrupt.

Secondary Deposits – White, crystalline secondary deposits judged to be ettringite are present in air voids within 1 to 2 in. of the interior surfaces of the cores. The voids are typically lined with minor amounts of ettringite, although the deposits are more abundant with proximity to the interior surface. In some areas of

Core #2, voids up to 0.2 mm are full of the deposits to a depth of 4 mm (see Figure 10).

Core #2 contains clear gypsum deposits within air voids in the chemically attacked, near-surface cement paste (see Figure 12). The gypsum is present to a typical depth of 1 mm and a maximum depth of 1.5 mm.

www.ConcreteTesting.com

Cracking Distress – Core #1 has a tight shrinkage crack present at and oriented perpendicular to the exterior surface. The crack extends to a depth of 17 mm and has a width of ≤0.01 mm.

Core #2 exhibits three tight cracks extending from the No. 4 rebar located 2.6 in. from the exterior surface.

Two of the cracks terminate 13 and 15 mm from the rebar, and the third crack extends 38 mm where it exits the side of the core. The cracks have widths of ≤0.01 mm. The cracks appear to be structural in nature.

Core #3 has two tight shrinkage cracks present at and oriented perpendicular to the exterior surface. The cracks extend to depths of 4.5 and 7 mm and have widths of ≤0.005 mm.

Carbonation – The cement paste in the near surface concrete is carbonated to the depths shown below:

Depth of Carbonation, mm

Core ID Exterior Interior

#1 4.0 – 8.0 0.5 – 2.0

#2 1.0 1.0 – 3.0

#3 0.5 – 1.0 1.0 – 2.0

SEM/EDS EXAMINATION RESULTS

Samples from the outer 0.7 in. of each core were polished for SEM/EDS analysis of the cement paste. For each sample, three analyses were performed on the weak, deteriorated surface cement paste, and two analyses were performed on the sound concrete at roughly 0.7 in. depth. Representative analyses are shown in Figure 13 through Figure 18, and the full data from the analyses are shown in Table 1 of

Appendix B.

The analyses of the sound cement paste are roughly consistent with portland cement paste with fly ash.

The attacked surface cement paste is heavily depleted in calcium, which is expected for any type of acid attack. The following table shows the calcium oxide levels in the very near-surface, chemically attacked cement paste relative to the levels in the underlying sound concrete.

Core ID CaO Content

0.7 in. depth Surface

#1 50.9 – 51.5 10.1 – 16.4

#2 47.2 – 51.6 8.2 – 13.6

#3 49.8 – 52.9 10.5 – 12.0

Secondary deposits within air voids in the upper 1 to 1.5 mm of Core #2 were confirmed by SEM/EDS analysis to be gypsum. A representative analysis is shown in Figure 19.

www.ConcreteTesting.com

SUMMARY AND CONCLUSIONS

A petrographic examination was performed on three concrete cores from the storage tanks used for deicing chemicals (propylene glycol) and runway runoff at the Cincinnati/Northern Kentucky International Airport in Hebron, Kentucky. The examination was prompted by deterioration of the interior surfaces of the tanks.

General Quality of the As-Placed Concrete

The coarse and fine aggregates are judged to be good quality and are well-distributed throughout the examined cores. The cement paste in the bulk of the cores is judged to be good quality having an estimated w/cm in the range of 0.42 to 0.47. The concrete is air-entrained with estimated total air void contents in the range of 3 to 8 percent. The air void systems are judged to be good quality with regard to freeze/thaw durability.

Condition of the Concrete

The concrete in the bulk of the examined cores is judged to be in good condition with no significant distress. The inner surface of each core is deteriorated, exposing the coarse aggregate particles. The near-surface cement paste is significantly weakened to typical depths of 1 to 1.5 mm and a maximum depth of

4 mm in Core #1 (see Figure 9 through Figure 11). The near-surface portion of exposed carbonate aggregate particles are also weakened.

Cause of the Surface Deterioration

The deterioration of the concrete surfaces at the interior of the tanks is related to acid attack. The exposure of concrete to glycol solutions is not directly deleterious. However, propylene glycol can degrade in the presence of oxygen to form various acid products including glycolic acid [C2H4O3], formic acid [CH2O2], acetic acid [C2H4O2], and oxalic acid [C2H2O4].

www.ConcreteTesting.com

The near surface concrete of Core #2 contains deposits of gypsum (CaSO4 ∙ 2H2O). The gypsum deposits could be an indication that the chemical attack in the tank represented by Core #2 is related to sulfuric acid

(H2SO4).

Acid Attack Mechanism

Acids attack the concrete by leaching the calcium from the calcium hydroxide [Ca(OH)2] and calcium silicate hydrate [Ca1.5SiO3.5•nH2O] components of the hydrated cement paste. These chemical reactions destroy the cementing qualities of the hydrated portland cement paste, resulting is a cementitious phase that is extremely soft and highly porous relative to the unaltered cement paste.

The acid dissociates as the reaction occurs. Concrete that exhibits acid attack generally shows an abrupt transition from the weak, chemically-attacked cement paste to good quality, unaltered cement paste.

Biogenic Sulfuric Acid Attack

Propylene glycol is typically used as a solution mixed with water. The amount of propylene glycol should be kept above 20 percent to inhibit the potential of bacterial growth. It was reported that the tanks receive runoff from runways, which would dilute the concentration of the glycol solution.

Wastewater treatment systems are well known to be susceptible to severe chemical attack by sulfuric acid.

Sulfuric acid is a waste product of microorganisms common in wastewater treatment systems. In such systems, sulfur compounds from the waste are reduced by anaerobic bacteria to hydrogen sulfide (H2S).

The hydrogen sulfide gas is absorbed into the near-surface concrete. The hydrogen sulfide is then converted to elemental sulfur. Sulfur is a substrate for many thiobacilli. These aerobic bacteria metabolize the sulfur into sulfuric acid, which chemically attacks the concrete. This type of acid attack is known as

Biogenic Sulfuric Acid Attack1 and is often also referred to as Microbially Induced Corrosion (MIC).

www.ConcreteTesting.com

The leached calcium and the sulfate released from the acid attack produces gypsum (CaSO4•2H2O). The formation and presence of gypsum can also cause sulfate attack of the concrete.

For biogenic sulfuric acid attack to occur within the glycol storage tanks, an external source of sulfur would be required.

Acid Attack of the Examined Cores

The chemical analyses showing the severe depletion of calcium within the near-surface concrete confirms acid attack as the mechanism of distress. If the acid attack was related to sulfuric acid, it would be expected that the sulfate content would be significantly elevated within the chemically attacked cement paste. With the depletion of calcium oxide from the cement paste, an increase of the other compounds in the cement paste are expected to show a proportional increase (see Table 1). The near-surface concrete of the analyzed samples does not show a significant increase of the sulfate content beyond the expected values. It also should be noted that although Core #2 contains gypsum (calcium sulfate) deposits within the near-surface air voids, the sulfate content of near-surface cement paste of this core is not elevated relative to the other cores.

The formation of the gypsum secondary deposits within the near-surface air voids of Core #2 indicates that the tank is being exposed to an external source of sulfur. However, it does not appear that the acid attack is related to sulfuric acid.

Trisha Piercey, Petrographer Nick Scaglione, President & Petrographer

Concrete Research & Testing, LLC Concrete Research & Testing, LLC

• All concrete samples will be discarded 1 month after the report issue date unless otherwise requested.

www.ConcreteTesting.com

REFERENCE

1. De Belie, N., Monteny, J., Beeldens, A., Vincke, E., Van Gemert, D., Verstraete, W., 2004, Experimental research and prediction of the effect of chemical and biogenic sulfuric acid on different types of commercially produced concrete sewer pipes, Cement and Concrete Research 34, 2223-2236.

Page A1 www.ConcreteTesting.com

APPENDIX A

Figure 1. Site photographs showing the storage tanks at the Cincinnati/Northern Kentucky airport.

The photos were provided by S&ME.

Page A2 www.ConcreteTesting.com

APPENDIX A

Figure 2. Site photographs showing the deteriorated, interior surface of storage tanks at the

Cincinnati/Northern Kentucky airport. The photos were provided by S&ME.

Page A3 www.ConcreteTesting.com

APPENDIX A

Figure 3. Photographs of Core #1 in the as-received condition. The core was taken from the

Northeast quadrant of Tank #1. The surface representing the interior of the tank is to the right in the photos.

Page A4 www.ConcreteTesting.com

APPENDIX A

Figure 4. Photographs of Core #2 in the as-received condition. The core was taken through the

Northeast quadrant of Tank #2. The surface representing the interior of the tank is to the

Page A5 www.ConcreteTesting.com

APPENDIX A

Figure 5. Photographs of Core #3 in the as-received condition. The core was taken through the

Northeast quadrant of Tank #3. The surface representing the interior of the tank is to the

Page A6 www.ConcreteTesting.com

APPENDIX A

Figure 6. Photograph of the lapped section prepared from Core #1. The surface representing the interior of the tank is to the left in the photos.

Page A7 www.ConcreteTesting.com

APPENDIX A

Figure 7. Photograph of the lapped section prepared from Core #2. The surface representing the

Page A8 www.ConcreteTesting.com

APPENDIX A

Figure 8. Photograph of the lapped section prepared from Core #3. The surface representing the

Page A9 www.ConcreteTesting.com

APPENDIX A

Figure 9. Photomicrographs of the near-surface concrete in Core #1 at 10x magnification. The cement paste at the immediate surface is poor quality. Portions of this poor quality cement paste were lost during specimen preparation (arrows). The lighter-colored underlying cement paste is judged to be fair quality. The photos also show the shallow deterioration of carbonate aggregate particles.

Epoxy applied by CRT

Page A10 www.ConcreteTesting.com

APPENDIX A

Figure 10. Photomicrographs of the near-surface concrete in Core #2 at 10x magnification. The cement paste at the immediate surface is poor quality. The transition to good-quality cement paste is abrupt. The photos also show deteriorated carbonate aggregate particles.

Page A11 www.ConcreteTesting.com

APPENDIX A

Figure 11. Photomicrographs of the near-surface concrete in Core #3 at 10x magnification. The cement paste at the immediate surface is poor quality. The transition to good-quality cement paste is abrupt. The upper photo also shows a deteriorated carbonate aggregate particle.

Page A12 www.ConcreteTesting.com

APPENDIX A

Figure 12. Photomicrographs showing clear gypsum deposits in the near-surface concrete of Core #2 at 50x magnification. The deposits are present within air voids and are marked with arrows.

Page A13 www.ConcreteTesting.com

APPENDIX A

Element Weight, % Compound Compound

Weight, %

Ca 36.4 CaO 50.9

Si 14.6 SiO2 31.2

S 1.1 SO3 2.7

Mg 1.7 MgO 2.9

Al 3.9 Al2O3 7.3

Fe 2.7 Fe2O3 3.9

Na 0.3 Na2O 0.4

K 0.1 K2O 0.1

Mn 0.1 MnO 0.2

Cl- 0.3 Cl- 0.3

O 38.7 — —

Total 100.0 100.0

Figure 13. Representative SEM/EDS analysis of the sound cement paste at 0.7 in. depth in Core #1.

The analysis shows the typical chemistry of the unattacked cement paste containing fly ash. The polygon in the upper image represents the area analyzed.

Page A14

Weight, %

Ca 9.7 CaO 13.6

Si 24.2 SiO2 51.7

S 0.6 SO3 1.6

Mg 2.8 MgO 4.6

Al 8.1 Al2O3 15.3

Fe 6.6 Fe2O3 9.5

Na 0.4 Na2O 0.6

K 1.3 K2O 1.5

Ti 0.6 TiO2 1.1

Cl- 0.6 Cl- 0.6

O 45.1 — —

Figure 14. Representative SEM/EDS analysis of the cement paste near the exposed surface of

Core #1. The analysis shows severe decalcification of the cement paste. The polygon in the upper image represents the area analyzed.

Page A15

Weight, %

Ca 36.9 CaO 51.6

Si 14.7 SiO2 31.4

S 1.2 SO3 2.9

Mg 1.3 MgO 2.1

Al 4.1 Al2O3 7.7

Fe 2.1 Fe2O3 3.0

Na 0.3 Na2O 0.4

K 0.1 K2O 0.1

Ti 0.3 TiO2 0.5

Cl- 0.4 Cl- 0.4

O 38.8 — —

Figure 15. Representative SEM/EDS analysis of the sound cement paste at 0.7 in. depth in Core #2.

Page A16

Weight, %

Ca 7.8 CaO 10.9

Si 23.4 SiO2 50.0

S 2.0 SO3 5.0

Mg 2.0 MgO 3.3

Al 6.4 Al2O3 12.2

Fe 10.5 Fe2O3 15.0

Na 0.6 Na2O 0.8

K 0.9 K2O 1.1

Mn 0.1 MnO 0.2

Ti 0.4 TiO2 0.6

P 0.2 P2O5 0.4

Cl- 0.5 Cl- 0.5

O 45.2 — —

Figure 16. Representative SEM/EDS analysis of the cement paste near the exposed surface of

Core #2. The analysis shows severe decalcification of the cement paste. The polygon in

Page A17

Weight, %

Ca 37.8 CaO 52.9

Si 14.5 SiO2 31.0

S 1.1 SO3 2.8

Mg 1.4 MgO 2.4

Al 3.6 Al2O3 6.8

Fe 1.6 Fe2O3 2.4

Na 0.2 Na2O 0.3

K 0.1 K2O 0.1

Ti 0.3 TiO2 0.5

P 0.2 P2O5 0.4

Cl- 0.3 Cl- 0.3

O 38.7 — —

Figure 17. Representative SEM/EDS analysis of the sound cement paste at 0.7 in. depth in Core #3.

Page A18

Weight, %

Ca 7.5 CaO 10.5

Si 25.7 SiO2 55.0

S 2.1 SO3 5.3

Mg 3.7 MgO 6.1

Al 7.2 Al2O3 13.7

Fe 4.3 Fe2O3 6.1

Na 0.5 Na2O 0.6

K 1.0 K2O 1.2

Ti 0.6 TiO2 0.9

Cl- 0.5 Cl- 0.5

O 46.9 — —

Figure 18. Representative SEM/EDS analysis of the cement paste near the exposed surface of

Core #3. The analysis shows severe decalcification of the cement paste. The polygon in

Page A19

Weight, %

Ca 28.8 CaO 40.3

S 23.9 SO3 59.7

O 47.3 — —

Figure 19. SEM/EDS analysis of clear, crystalline secondary deposits within air voids in the weak, deteriorated cement paste of Core #2 (see Figure 12). The analysis confirmed the material as gypsum [CaSO4∙2H2O]. The upper images show the deposits at 402kx (left) and 40x (right). The circles in the images represents the area analyzed.

Epoxy applied by CRT Epoxy applied by CRT

Page B1 www.ConcreteTesting.com

APPENDIX B

Table 1. Summary of the SEM/EDS analyses performed on the cement paste in the cores.

Core ID Depth CaO SiO2 SO3 MgO Al2O3 Fe2O3 Na2O K2O MnO TiO2 P2O5 Cl-

#1

Surface 1 16.4 47.6 6.8 4.9 12.3 8.3 0.7 1.5 0.2 0.8 — 0.7

Surface 2 13.6 51.7 1.6 4.6 15.3 9.5 0.6 1.5 — 1.1 — 0.6

Surface 3 10.1 56.9 2.6 3.1 13.6 9.2 0.7 2.3 — 0.8 — 0.5

Sound 1 50.9 31.2 2.7 2.9 7.3 3.9 0.4 0.1 0.2 — — 0.3

Sound 2 51.5 30.0 3.1 3.6 7.6 3.0 0.4 0.3 — — — 0.4

#2

Surface 1 8.2 53.3 6.9 1.1 18.6 8.2 0.7 1.5 — 0.7 0.4 0.5

Surface 2 13.6 52.8 5.0 3.5 13.2 8.1 0.6 1.4 — 0.9 0.3 0.6

Surface 3 10.9 50.0 5.0 3.3 12.2 15.0 0.8 1.1 0.2 0.6 0.4 0.5

Sound 1 51.6 31.4 2.9 2.1 7.7 3.0 0.4 0.1 — 0.5 — 0.4

Sound 2 47.2 34.1 2.4 2.3 9.1 2.7 0.6 0.9 — 0.5 — 0.2

#3

Surface 1 12.0 53.8 4.9 2.4 17.4 5.6 0.7 1.4 — 0.8 0.4 0.5

Surface 2 11.2 53.7 8.3 1.6 14.8 5.9 0.8 1.7 — 1.0 0.4 0.6

Surface 3 10.5 55.0 5.3 6.1 13.7 6.1 0.6 1.2 — 0.9 — 0.5

Sound 1 49.8 31.5 2.8 2.4 8.4 3.3 0.3 0.2 — 0.7 0.3 0.3

Sound 2 52.9 31.0 2.8 2.4 6.8 2.4 0.3 0.1 — 0.5 0.4 0.3