37

Port of Rotterdam - Results Directaapa.files.cms-plus.com/SeminarPresentations/2015Seminars/2015Fac... · accelerated low water corrosion occurring at ... Unexpected corrosion 11

  • Upload
    leanh

  • View
    228

  • Download
    4

Embed Size (px)

Citation preview

3

Port of Rotterdam

More than 50 kilometres !

44

1400 - 1800

1800 - 1900

1929 – 1949

1948 - 1957

1960 - 1970

1970 – 2008

2008 - 2030

Development of the port

1934 - 1946

1906 - 1922

5

Salinity in the port

6

Port in figures

Total port area 26,000 acres (net 12,500 acres)

Total employment 350,000 people (140,000 regionally)

Throughput 430 million tons, 11,1 million TEU (2010)

Depth up to 75 ft (= 24 m)

Port is growing westwards from the old town with fresh water of river Rhine into the salt Northsea

Tidal zone ca. 2 meter in the whole port

7

Quay walls: total length 70 km

Replacement value ~ 1.5 billion €

About 40 km with steel substructures, like combi

walls, sheet piling, etc.

Port infrastructure in Rotterdam

Asset Management on Quay Walls

A quay wall’s remaining lifetime and system integrity is

mainly determined by the quality of the sub and

superstructure.

When the quay wall’s integrity is in danger, it’s often

due to:

concrete deterioration in the superstructure

accelerated low water corrosion occurring at

the substructure

9

Corrosion rates according to EAU-Curves

No steel but concrete in the splash-zone

Extra thickness added to steel parts

Inspection of a few quay walls every 5 year

Corrosion strategy (up to 1999)

What was discovered ?

Before cleaning with high pressure After cleaning with high pressure

11

Unexpected corrosion

11

Holes in tubular piles

Hole, followed by loss of soil

Original

thickness

including

original coating

Probable causes (long list)

More salty environment

Galvanic corrosion of combiwalls

pitting influences on the strength

Aeration due to extra oxygen by (bow) thrusters or cooling

Fabrication process of steel parts

Water flow

Cathodic protection to the ship

Vertical layering of water (temperature, oxygen, salinity)

MIC

Etc...

13

Action plan for investigation

Stress corrosion test

Coupon ladder

Coupons (galvanic cells)

Micro biological influence

Weakening due to pitting

Thickness measurement

14

Stress corrosion test

Coupons for galvanic corrosion

test

Coupon ladder test

17

Results of coupon ladder

Coupons corrosieladder Amazonehaven

-18,0

-16,0

-14,0

-12,0

-10,0

-8,0

-6,0

-4,0

-2,0

0,0

0,000 0,050 0,100 0,150 0,200 0,250 0,300 0,350

Corrosie in milimeters per jaar

Die

pte

s in

met

ers

t.o

.v. N

.A.P

.

mm/jr 2003 mm/jr 2005 mm/jr 2008 Diepte (m)

18

Research influence MIC

Sampling

19

Weakening due to pitting

Kracht- rek diagram 78547-8 Westerkade

-50000

0

50000

100000

150000

200000

250000

300000

0

0,0

1

0,0

2

0,0

3

0,0

5

0,0

8

0,1

0,2

4

0,3

9

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

0,6

7

rek

kra

ch

t N

coupon 1

coupon 2

coupon 3

coupon 4

20

Ultrasonic measurements

Method: Ultrasonic measurement

Inspection locations: horizontal & vertical

Amount of inspections: one million points

Inspection results processed statistically

Representative values: (incl. pitting!)

21

Corrosion rate in specific

conditions

What is the salinity ( fresh, 10 ppm < brackish <25 ppm, salt)?

Steel quality and electro-chemical corrosion?

Biological activity (general in salt water, sometimes in fresh)?

Galvanic cells (e.g. combi-walls, repair, welding joints)?

Fabrication influence (e.g. purity mix, cold rolled)?

Extra high oxygen level (e.g. propellers, cooling system)?

High flow velocities because of geometry ?

Strong vertical differences in fresh and salt layers?

Cathodic protection on board of moored ships?

Contribution of influences

Causes / influences Fresh Brackish Salt

1 Electrochem. Corrosion 5% 5% 9%

2 Biological activities 5% 9% 18%

3 Galvanic influence 0% 9% 27%

4 Fabrication 0% 5% 5%

5 Aeration 5% 5% 9%

6 Local high waterflow 0% 5% 9%

7 Vertical salt-fresh layers 0% 5% 0%

8 Cath. prot. on board of ship 5% 5% 5%

9 Splash-zone 18% 18% 18%

Relative Contribution 38% 66% 100%

Avg. corrosion [mm/50yr] 3,4 5,9 9,0

Max. corrosion [mm/50yr] 10,3 17,8 27,0

Subsets of uniform

corrosion conditions and

uniformized

measurements

time [yr]

co

rro

sio

n[m

m]

50

9

8

7

6

5

4

3

2

1

10

CP + m

(+c)

+d + m

+d + m

+cp (+c)

9 corrosion-scales

Measures:

d = extra thickness

m = monitoring

cp = start with cath. protection

CP= Cathodic protection

(c)= coating if needed

27

Factor 2.5

9

avg

max.

Corrosion scales and measures

25

Overview

23. ...........

3. ...................2. Biological activity1. Galvanic working

0-hypothesis( 9 main causes )

Probable causesExperts:

in lab in situMeasurements on

specific aspects:

Tuning in practice:

= uniform group

Then from scale to degradation

Uniform group of measurements are placed on the

corrosion scales (example: average 1,9 mm after 16

years)

26

Scale defenition

0,000

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54

Inspection year

co

rro

sio

n s

ca

lel

schaal 1

schaal 2

schaal 3

schaal 4

schaal 5

schaal 6

schaal 7

schaal 8

schaal 9

wanddikte

Then from scale to degradation

Then from scale to degradation

Uniform group of measurements are placed on the

corrosion scales

Extrapolation of corrosion within scales (example

between scale 3 and 4)

28

Standard Corrosion Scales

0,000

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54

Year

To

tal a

ve

rag

e c

orr

os

ion

[m

m]

standard scale 1 Input new Quaywall

9

8

7

6

5

4

2

3

1

Then from scale to degradation

Then from scale to degradation

Uniform group of measurements are placed on the

corrosion scales

Extrapolation of corrosion within scales

Translation of corrosion to degradation (example

remaining safety factor 1,30)

Prediction of remaining lifetime

30

Then from scale to degradation

Degradatio nline

1,00

1,10

1,20

1,30

1,40

1,50

1,60

1,70

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54

Year

Safe

ty f

acto

r

Degradationline

beheerfase 1

Beheerfase 2

Beheerfase 3

maintenance phase 1

maintenance phase 2

maintenance phase 3

Input for KMS

KMS: Quay wall modelling system

Port of Rotterdam’s next step in asset management

Supports the asset manager

Combines technical risk and business value in order to prioritise maintenance measure

Degradation model is part of KMS

Degradation model in KMS

Ongoing research to improve the

steel model

Research in 2013 A + B + C + D

Coupon frame Measurement waterconditions Coupontest Galvanostat/Potentostat-measurement

H1 hypothesis Divice: Hach Lange

Z1AL L3 Locations Standard coupons

Houston O2

Boston pH exhibition coupons in lab

Corpus Christy Ec

Sohar T under variation of:

P.O.L.B Bart's (Microbiology ) pH

Rotterdam (reference) T

Upper layer O2

Mid. Layer

S235JO Sheetpiles

S465 Piles

"Bart's test":

Research on presence on

Bottom layer sulfate reducer

Coupontest on site

USA Sohar and the

Netherlans (for the

reference)

Water measurements:

O2 - pH - Redox

EC - temp Lab test

coupons

Electrical chemical

test for

determination of

accelerated

corrosion rates

Presentation corrosion research

potentiostat

device

coupon weighting and measuring of

water conditions, is always in

combination

Risks on CP systems

Unsufficient protection

Caused by anodes that don’t work properly

Incomplete coverage of structure

Caused by mechanical damage

Decrease of the anode lifetime

Caused by higher load on anode than designed due to

changing conditions

| 37

Monitoring of CP systems

Visual inspections with divers

mV measurement

Weighing of anode to measure

consumption in time

US thickness measurement

Condition monitoring

| 38

Monitoring water conditions

Water conditions determine how

efficient the system will work

Monitoring of:

Conductivity / Salanity

Temperture of the water

pH

Dissolved oxygen

39

Results of monitoring @PoR

mV is more than sufficient

Brackish/ fresh water :

anode consumption is lower

than in salt conditions

In salt water, the

consumption is less at end

of life of the anode

Development of anode

model

40

Thanks for your attention !