As mentioned in my last post, a workshop on the combustion research performed by the UK OxyCAP project was held side-by-side with a meeting of the corrosion-oriented ‘OxyCORR’ working group at Imperial College last week, allowing both groups of researchers to interact and providing some welcome international exposure to the UK project.
Jointly funded by the EPSRC and E-ON, OxyCAP is a research project involving seven UK universities which has been running since 2009 and is concluding this month. The research has included a host of experimental and modelling work aimed at clarifying the complex effect of oxyfuel coal firing on combustion and heat transfer properties. Much of this work is complementary or contributing to the EU Relcom project discussed in my last post, and a large part is based around the 250 kW oxyfuel pilot rig of the UK PACT facility which I visited earlier in the year. Amongst the work presented here was that of the University of Kent on developing both 2D and 3D flame imaging systems for application to this test rig as well as full-scale furnaces. This detailed imaging of flame behaviour helps validate high resolution modelling work such as that presented by the University of Duisburg, where large eddy simulations were used to powerful effect in the dynamic modelling of oxyfuel flames. As accurate CFD models require submodels describing the constituent physical and chemical phenomena, the more fundamental drop tube furnace combustion studies conducted by the University of Nottingham also play a key role. These tests have quantified the influence of effects unique to the oxyfuel atmosphere such as char gasification by CO2 and water vapour; both shown to have a significant effect on char burnout and volatile yields. A pair of closing presentations on analyses of oxyfuel ash (apparently quite similar to air-fired ash) eased the transition into the corrosion focus of the following OxyCORR meeting.
The potential for increased fireside corrosion of waterwall and superheater tubing has been an early concern for oxyfuel carbon capture, as recycling flue gases back to the boiler could make for a more damaging environment than found in air-fired furnaces. A wide range of industrial and academic research is being conducted into corrosion phenomena which are also notoriously difficult to study, so the OxyCORR working group has sought to coordinate and review some of this research since its first meeting in 2009. The atmosphere in an oxyfuel boiler depends greatly on the extent to which flue gas is cleaned before being returned to the boiler, with processes such as desulphurisation and water condensation preventing the concentration of corrosive species in the boiler, but also carrying an energetic penalty which would best be avoided. Whilst SOx species in particular are involved in several high-temperature corrosion mechanisms, the high CO2 and water vapour levels inherent to the oxyfuel process also make for a new environment which requires a rethink of our existing knowledge on coal furnace corrosion.
Metal corrosion tests, whether in pilot furnaces or controlled laboratory environments, are highly difficult to standardise and draw meaningful results from, due to the huge degree of variability surrounding factors such as ash deposit composition. As a result, it is not unusual for research groups to identify quite different behaviour and mechanisms under seemingly equivalent conditions. In spite of some continuing disagreement on mechanistic specifics, the most notable conclusion to come from the meeting was that high-temperature corrosion in oxyfuel is actually very similar to under air-firing, and at worst, appears equivalent to conventional combustion of higher sulphur coals. More surprisingly, even elevated SOx levels arising from hot flue gas recycle do not always appear to have any impact. Whilst they can contribute to stabilising corrosive alkali metal sulphate ‘melts’ over a wider temperature range, in many cases SOx is converted to harmless sulphates of calcium or magnesium within the ash deposits. On the other hand, some groups identified the elevated levels of water vapour as a greater risk, due to its role in impeding formation of protective chromia scales by volatilising chromium. However, it was argued that this too may not be as relevant to real boiler tubing covered in ash deposits. Water vapour also appears to have the beneficial effect of precluding any significant chemical attack by CO2, known as carburisation, which was once thought to be one of the primary hazards of the oxyfuel atmosphere.
Talks from both Babcock and Wilcox and Alstom gave some important insight into corrosion work on the two active oxyfuel demonstration projects, Futuregen 2.0 in the USA and White Rose in the UK. As the FEED study for the American project was concluded earlier this year, Babcock and Wilcox have completed their boiler material selection studies with the favourable conclusion that purely conventional materials such as T91 are appropriate for superheaters. Although the Illinois-based project will not be able to fire exclusively the sulphur-rich local coal (challenging even for conventional plant), blending with 40% Powder River Basin coal and semi-dry FGD of recycled flue gases will both help to reduce SOx levels in the boiler.
With the general conclusion that high-temperature corrosion does not pose a particular barrier to developing an oxyfuel plant, some research institutes have even declared that they will cease working on the problem. The working group suggested instead turning more towards the investigation of oxyfuel corrosion under more demanding conditions such as in load-following plant, biomass cofiring, and at advanced 700ËšC steam conditions, with such work having already received attention from some groups. A discussion on the need to standardise the array of measurement techniques available to corrosion researchers was led by Cranfield University, who have developed their own rigorous statistical method for their own work in this field. As part of this drive, the group suggested that a series of standard tests could be performed by all the participant labs in order to identify the scale of the experimental variation and the reasons behind it.
All in all, the meeting presented a much more upbeat perspective than the previous gathering in Spain last year, with a high level of confidence from all parties that oxyfuel corrosion is not fundamentally different to conventional coal plant, and shouldn’t impede the move to a full-scale oxyfuel plant. However, as with all oxyfuel discussions, the need for such a plant to finally be realised to further our understanding of these phenomena was strongly emphasised. Unsurprisingly, in all three of the oxyfuel meetings I have attended this month there is much anticipation of the forthcoming investment decision of the Futuregen project expected at the end of this year, and that of White Rose expected a year later. The success of one of these projects could put oxyfuel firmly on the table as a carbon capture technology.