Technical article

Retrofit strategy using aluminum anodes for the internal sections of wind turbine monopiles

by Alex Delwiche, Isaac Tavares

Abstract

This paper discusses the corrosion mitigation strategy for protecting the submerged sections of a monopile wind turbine structure off the east coast of England in UK waters. The choice of the CP system was based on a remotely monitored trial, to minimise the retrofit installation works and thereby keeping costs and offshore work activity down, minimising the safety risk, yet provide an effective working system. The details and results of the initial trial are presented, as is the planning and preparation details to install aluminum anode strings inside of the supposed sealed internal sections of the wind turbine monopiles.

Introduction

Lynn and Inner Dowsing (LID) is a WindFarm located off the coast off the East Lincolnshire coastline comprising of 54 No. 3.6 MW turbines, each supported by steel monopile foundations. The monopiles (MP) and their steel transition pieces (TP) were installed in 2007 and the turbines (WTG) in 2008.

The monopiles are driven through the seawater into the seabed and remain flooded. The TP sits on top of the monopile and was originally grouted in place. The lower deck of the TP sits inside the top of the monopile whilst above this is a sealed platform creating an airtight seal into the monopile. One to three J-tubes come through the bottom of the monopile into the TP and are “sealed” into the base of the monopile. This is best shown in the figures below.

Figure 1 - Monopile structure construction

According to early versions of the DNV(1) guidelines for the corrosion protection of WTGs (Wind Turbine Generators) stated that “If the inside of piles such as monopiles is ensured to be airtight, i.e. there is no or very low content of oxygen, corrosion protection inside of the piles is not required.” The LID monopiles followed this design philosophy, i.e. no CP was installed internally nor was any coating applied on the monopile or inside of the TP below the airtight deck.

Later versions of the standard(²⁾ recognizing the concerns from operators that there was a potential risk of corrosion inside the closed compartments, the DNV standard on guidance for wind turbine designers, revised this view with the statement “should have cathodic protection with coating around the “splash zone”.” A visit was carried out in October 2011 to two of the wind turbines and the results concluded corrosion, particularly in the submerged zone, was occurring within the supposed airtight section below the TP.

Corrosion rates measured from retrieved coupons in the submerged sections, varied in corrosion rate. As expected the corrosion rates in the sealed monopile were generally less than in the freely flooded condition. Although the corrosion rate was not a concern, fatigue curves used to calculate the life expectancy of the monopile were based on a corrosion free state. This was not the situation and leaving the monopile without any corrosion prevention system would heavily reduce the theoretical life.

Rather than risk not having a corrosion prevention system, a fast-tracked strategy was implemented. In addition to J-tube repairs (re-sealing), the project chose cathodic protection as the most viable method of re-instating a protected structure fatigue curve rather than a freely corroding fatigue, which was now the case.

Papers already published also found this issue on other WTG structures in Europe(³⁾, however, this project is thought to be one of the, if not the first, to have CP installed inside the submerged section of the monopile. Due to the conditions and environment of the monopiles, any retrofit strategy would involve huge costs and require careful planning.

In 2011 there were no specific standards or papers relating to guidance in applying CP inside monopile structures offshore. Specifically, there is little guidance on CP retrofit strategies offshore generally.

Design

CP design is relatively straight forward. Calculate the surface area to protect, apply coating factors (where coating exists, here they do not), apply a current density value (guideline from standards) and one can easily calculate the total current requirement. A few other calculations can then determine anode geometry and quantity to meet the current requirement and use a given anode capacity to determine weight requirements to meet the life.

However, the design had several constraints;

- Weight and handle ability of the anodes offshore,

- How are the anodes going to be mounted?

- Concern over hydrogen evolution in a sealed environment

Surface area is based on a submerged section (primary structure and J-tubes) but also a buried section, up to 30 m seabed depth. How much should be applied (or considered to be lost in seabed)? Biggest question then is what current density to use in the sealed monopile submerged section given that low oxygen diffusion in the water was expected because of the “sealed” airtight hatch on entry into the submerged section of the monopile?

With so many unknown questions and so much riding on the installation of 54 No. structures, optimizing the design was required and a CP trial was implemented in early 2012 with a view to roll out the design in the summer of 2012 with the monopile remedial repairs.

Although much easier and likely more economical to install, the risks from chlorine gas and mixture with damp conditions (hydro-chlorides) and even risky likelihood of hydrogen, the option of ICCP was strongly dismissed as an alternative. Trials have been ongoing and designs are now appearing since this trial was carried out, however based on events this project had neither the inclination or time to consider this option.

The galvanic anodes were to be mounted at several hangoff locations. Anode strings were therefore to be used and to save on wiring, anodes cast on galvanized steel wire were to be used, to be re-assembled immediately prior to installation.

Weight is a major concern from a manhandling perspective. The anodes would need to be loaded on a vessel, taken to the field, lifted (50 kg weight limited) on to the TP platform by a Davit crane, hauled inside, lowered through two decks, assembled, lifted and lowered into position on the lower deck hangoff location. This is not easy.

The operator/ construction crew requested that the choice of anode was either 1 No. 50 kg anode or 2 No. 25 kg anodes connected by the rope, aluminum alloy (Al-Zn-In) or zinc. The electrochemical capacity of aluminum is up to 2500 Ah/kg (although DNV-RP-B401 2010⁴ recommend no more than 2000 Ah/kg in a CP design). This still has 2-3 times greater capacity, i.e. 2-3 less weight, than that required for a zinc alloy. Therefore, on the manhandling and cost of installation alone, zinc anodes were dismissed.

However, how much total weight is required to meet protection, in this case a minimum potential of -800 vs Ag/AgCl (Silver Silver/Chloride) reference electrode was sought. Previous findings from the site initial visit to the field, did not give the project team any cause for concern regarding MIC (Microbial Induced Corrosion).

Monitoring system

The monitoring system for the offshore trial consisted of;

- 6 No Dual reference electrodes (Ag/AgCl) measuring the structure

- 1 No Dual reference electrode (Ag/AgCl) measuring one of the anode strings

- pH probe

- Dissolved Oxygen probe

- 3 No. Hydrogen monitors

- 1 No. 1m2 Current Density plate

- 1 No. Remotely accessed (by cell phone data) datalogger

- 1 No. monitor of the ventilation system current input

Offshore trial

A total of four strings of Al-Zn-In alloy anodes were lowered into the confined space of one monopile (The Trial Position - TTP) and hung-off the lower platform. However, of these four anode strings, only two were initially connected to the structure. The trial commenced (connecting two anode strings to the structure) in March 2012 and 4 weeks later, for reasons seen further down, a third anode string was connected to provide additional current. To make life easier and prevention of entering the airtight column, which requires a full rescue team and is very costly, for the trial the anodes were isolated from the structure at the hang-off and 95 mm² cables were run above the airtight deck into the monitoring panel. This allowed relatively easy access to “switch” the anode strings to and from the structure. The latter being a major consideration should dangerous levels of hydrogen be detected.

Six silver/silver chloride dual reference electrodes were located in two positions at three elevations to monitor the structure to seawater potential and an additional reference electrode was placed next to an anode string – identified as Ref. No. 7.

A ventilation system was implemented with a fan, which was monitored to confirm operation, however, this was switched off after a short period of time to see if there was any hydrogen build up, which was one the major concerns over installing a CP system in an airtight facility.

In addition, a current density probe, and hydrogen sensors were installed at the top of the sealed column and one above the airtight hatch.

Further monitoring

As seen in the first set of results below, unexpected potentials were detected. The first plan of action was to connect another anode string to the structure, however, this had no effect and in fact seemed to make matters worse. An intervention into the monopile below the airtight deck was established and calibration checks were undertaken. After water samples were taken it was found that the water had to be replenished and the trial was reactivated with a flushing system implemented.

As of compiling this paper, we are still receiving good data.

A further four monopiles were installed with anode current and potential measurement equipment including a monopile that had been drilled as part of the long-term monitoring strategy.

Results

The first set of results can be seen in the following figure. As can be seen, the structure started to polarize as expected, with initial high current but a reduction in output as the structure achieved more protection. The current density plate also can be seen to polarize very quickly starting at 150 mA/m² reducing to 20 mA/m^2.It should be noted the current density plate was relatively clean sheet steel, as opposed to the corroded uncoated monopile structure.

Figure 2 - Initial offshore trial results

After two weeks of operation, it became apparent from the remote monitoring data that something was wrong as instead of continually polarizing, the structure started to de-polarize and the structure to electrolyte potentials became less negative as can be seen from Figure 2. Furthermore, in addition to the change in structure to electrolyte potential that was observed, there was also an increase in anode current output and cathode current density over the trial period (see Figures 2 and 3). These changes were not what was expected as we believed the anode current output and cathode current density would show a time-dependent decrease as gradual structure polarisation was thought likely to occur.

Figure 3 - Historical CP readings

Figure 4 - Individual anode string current vs time for initial period of trial

Figure 5 - Current density vs time for initial period of trial

Figure 5 shows the effect also occurred on the current density plate.

Hydrogen

At this stage, it should be noted that the ventilation system had been switched off to monitor any rise in hydrogen. An insignificant change in hydrogen levels was detected in that there were no concerns at this stage of hydrogen build up risk within the monopile.

Intervention

The airtight deck was breached two months after commissioning the two anode strings however, what was found was very unexpected. H₂S (Hydrogen Sulphide) alarms were set off and a strong smell of rotten eggs was notable, which also confirmed the alarms were truly measuring H₂S.

The water had a distinctly blue sheen in colour and whitish deposits were noted around the walls of the monopile.

All the equipment was calibrated - in the sense the reference electrodes were checked against a freshly calibrated reference electrode, the current inputs were measured via the shunts using a calibrated voltmeter and a clamp meter and the inputs into the data logger were confirmed it was measuring true readings. The exception was the dissolved Oxygen sensor, which despite efforts to keep agitated in accordance to the manufacturer’s instructions, seemed to fail as the results did not make any sense.

Water samples were taken and an immediate squeeze of a bottle confirmed significant levels of H₂S gases were detected by the personnel gas detectors (required for safe entry in the sealed section of the monopile).

The water samples were taken to a laboratory for testing and very low levels of pH were found as low as pH 4.5. Further samples were removed and tested and a portable pH meter was also used to confirm the laboratory findings.

After examining the options, a flushing solution was introduced by ways of breaching the sealed monopile to the open sea via a couple of small holes, however because the monopile water was so contaminated the water was partly emptied and replenished with open seawater to remove any concerns.

The anodes were never disconnected during this process and re-activated very quickly and polarization was achieved with a lesser initial current, most likely as some calcareous deposits had been formed.

Figure 6 - Potential and current data after flushing

The data has continually been collected since the reactivation of the trial. pH and potentials are being monitored manually to confirm that all the monopiles are protected and flushing is occurring in the monopiles to ensure the pH levels do not decrease. The results are similar as the trialed data with variances in protection seen over time.

Discussion

The offshore CP retrofit trial showed an unexpected phenomenon in that the introduction of aluminum anodes inside a “sealed” Wind turbine structure reduced the level of pH very quickly. Further studies were undertaken by the operator as were other operators started to report similar findings after this trial had been started^5,6,7,8,9,10.

A low pH reduces the efficiency of the anodes, makes it difficult to produce the calcareous deposits required for adequate CP and increases the risk of hydrogen production.

An explanation of the mechanisms to how a low pH could exist is explained by the authors in a separate paper, as the mechanisms are extremely complex and has wide reaching consequences for the CP industry, as this phenomenon has never been knowingly reported before this trial was undertaken¹¹.

The project team looked at a number of options including introducing chemicals into the water within the monopile to counteract the low pH. However, two concerns prevented the project team from executing this solution. Firstly, carrying chemicals especially in either large doses or concentrated doses poses a HSE (Health and Safety) risk on the vessels and on the transition piece and secondly on environmental grounds it would be challenging in an offshore environment to have a mixture of chemicals.

The end solution was to the breach the monopile to expose the seawater inside to the outside but in such a way the tides, especially the Spring tides, did not reach the lower platform within the monopile, which would strongly impact its integrity. The holes were carefully designed to allow a 5% change in water per day. This has proved extremely effective although constant monitoring is still required to ensure the flushing continues. Ongoing portable pH checks have confirmed that most of the monopiles now have pH levels above pH6 which is considered satisfactory, although the vast majority exceed this into pH7.

The H₂S levels have abated over time. This is likely to be a result that any residue sulfide ions present in the water, left over from living organisms and microbial activity, have reduced due to the evolution of H₂S but also as the pH had returned to normality. It should be noted H₂S was not found in all the remaining monopiles but was detected in 10% of the piles during the first year of operation of the CP system. H₂S has only been detected on those monopiles where the pH was less than pH6 at one stage. In those piles, the water was also found to have a blue sheen and bubbling an unidentified gas, not thought to be hydrogen as no gas alarms were activated other than low oxygen levels, indicating this could have been CO₂(Carbon Dioxide).

The whitish deposits noted on the trial monopile were probably early signs of calcareous deposits.

The trial monopile continues to be monitored. The data shown in Figure 5 shows some interesting seasonal variations. Further examination of the long term data is required.

Conclusions

The offshore CP trial within a Wind turbine monopile structure was extremely beneficial in establishing current requirements and longevity of the anode strings, which confirmed the original design premise but also highlighted an unexpected phenomenon with a reduction in pH utilizing aluminum anodes.

No significant hydrogen levels were ever measured, however, the introduction of CP in monopiles has found that H₂S gas can be evolved in the initial period of polarization, from the sulfide ions that are initially present in the water.

The additional monitoring in the other monopiles not only gives the operator peace of mind that the CP is working correctly, this data also aids the operator in proving to certification bodies such as DNV GL that the CP system is operational and protection is continually being maintained.

Replenishment or flushing of the stagnant seawater inside the monopile is necessary when using aluminum anodes to prevent a reduction in pH.

Acknowledgements

The authors of this paper would like to thank Centrica PLC for its financial support and for the permission to publish this paper. Thanks are also due to David Kerr and Nigel Terry of RES Ltd for their support on the offshore activities and a huge thank you for the construction team that contributed to a safety award as the optimized design was implemented on the remaining piles without incident.

References

1. DNV-OS-j101^([1]) (2004), “Design of Offshore Wind Turbine Structures (Det Norske Veritas AS).

2. DNV-OS-j101 (2007), “Design of Offshore Wind Turbine Structures (Det Norske Veritas AS).

3. L Hilbert et al, “Inspection and Monitoring of Corrosion Inside monopile Foundations for Offshore Wind Turbines”, EUROCORR 2011, paper 4730 (Stockholm: Eurocorr 2011)

4. DNV-RP-B401 (2010), “Cathodic Protection Design” (Det Norske Veritas AS)

5. A R Black, L R Hilbert, & T Mathiesen. “Corrosion Protection of Offshore Wind Foundations”. CORROSION 2015. Paper C2015-5896

6. S Briskeby, S M Hesjevik & L Borvik. “Cathodic Protection In Closed Compartments – pH Effect and Performance Of Anode Materials” CORROSION-2015 paper 2015-5657

7. Isaac Tavares et al, “Internal Cathodic Protection of Offshore Wind Turbine Monopile Foundations”, ICorr – Corrosion Management, Issue 123, Jan/Feb 2015

8. Isaac Tavares et al, “Internal Cathodic Protection of Offshore Wind Turbine Monopile Foundations”, Eurocorr 2015, Paper 442;

9. Isaac Tavares et al, “Internal Cathodic Protection for Offshore Wind Turbine Monopile Foundations”, Australasian Corrosion Association 2016

10. I Tavares et al. “Corrosion Threats to Offshore Wind Foundations” EWEA Offshore 2015. Copenhagen 10-12 March 2015

11. I Tavares, A. Delwiche, P Lydon, “Concerns Over Utilizing Aluminium Alloy Anodes in Sealed Environments,” CORROSION-2017, (New Orleans, NACE, 2017) paper no. C2017-8956

[1] DET NORSKE VERITAS, now known as DNV GL, Veritasveien 11363 Høvik, Norway