Unexpected halts in plant output can be the result of the seemingly unpredictable corrosion of boilers. But Jaani Silvennoinen, Maria Nurmoranta and Lasse Kauppinen claim one system will make the problem foreseeable.

An unplanned power plant shutdown or unforeseen maintenance will leave an energy producer with an economic loss. One common cause of such an unexpected event is the corrosion of a boiler’s superheater because of the action of alkali chlorides at temperatures above 450°C.

These destructive compounds form because of the presence of fuel components that exist in greater amounts in non-fossil and recycled resources, the use of which is growing, especially in Europe, China and North America. But there is a way to manage the corrosion, make it predictable and, in turn, prevent unforeseen stoppages.

The complexity of corrosion

The corrosion phenomenon is complex andmany factors affect the process. Although the rate of corrosion of the material from which boiler parts are made cannot be measured directly, it is possible to determine how corrosive flue gases are.

Metso employs real-time measurement to determine the sulphur and chlorine contents of the flue gas and calculate their ratio. This allows the estimation of corrosion by using this ratio along with empirical data about how the fuel and the process atmosphere have affected reference boilers.

Generating a reliable estimation requires a wide database of material properties and corrosion rates. The process atmosphere can be manipulated by: controlling combustion and the fuel blend; by dosing with additive – either in the fuel or by injection into the upper furnace; and by controlling material temperatures. Monitoring and controling air emissions and ash quality are also important factors in the overall solution.

Our control package allows the monitoring of the condition of the boiler, prevention of superheater corrosion or to keep the maintenance of it at an acceptable level.

A first step toward any such solution to secure high plant performance is to evaluate the combustibility of the fuel and then to perform a real-time fuel and process atmosphere characterisation. Solid fuels vary widely in their origin and quality. They can be wet as sludge, have as high a heating value as coal, have as high concentrations of potassium and phosphorus as agricultural biomass or have as high concentrations of heavy metals and chlorine as recycled fuels.

Potassium and chlorine in fuels vaporise easily and are known to play a role in bed agglomeration, fouling, slagging and processes related to high-temperature corrosion. While phosphorus is involved in phenomena related to bed agglomeration, fouling and slagging, and an increase in nitrogen content may elevate NOx emissions.

Combustion of recycled fuels such as wood and solid recovered fuels (SRFs) may lead to increased fouling and corrosion of furnace walls, superheaters and economisers. These problems have been associated mainly with chlorine, zinc and lead, while another challenge associated with the combustion of recycled wood is the formation of molten metal at the bottom of the boiler.

Technology of choice

The intention of the plant owner is to capitalise on the investment as widely as possible. In practice there is pressure to have high fuel flexibility without compromising plant availability and without exceeding emissions limits.

Fluidised bed combustion has been the operator’s and owner’s choice of technology because of its natural multi-fuel capability. However, in many cases, precise knowledge of the interactions of the fuels is lacking and has had to be gained through experience by monitoring the process after the plant has been built. Often, the fuel blend fails to meet the specification used in the plant’s planning phase. This can lead to the plant supplier and the owner undergoing an experience-gaining process that is expensive and stressful.

Categorisation of solid fuels according to its main elements that create challenges helps in the process of finding an appropriate technical solution and ensures transparent communication. Different groups of fuels need different technical solutions (see Table 1).

table 1

The level of challenge that a fuel presents lies on a scale of 1 to 10 that derives from experiment-based experience. It is an intuitive element rather than the output of a well-devised equation. Natural gas or heavy oil has a value of 1, whereas Nordic softwood-based fuels have a value of 2 and serve as a reference for solid fuels. Fuels labelled ‘wood+’ have more alkali and chlorine components than ‘wood’ types, and they potentially increase the risk of high-temperature corrosion.

Agro-type fuels have high concentrations of potassium, phosphorus and silicon. Examples are straw and the by-products of the food processing industry, such as hulls or kernels. High concentrations of chlorine, lead, zinc and, sometimes, alkali are characteristic of recycled wood and SRF.

Residues from coal beneficiation processes have ash contents as high as 80 per cent by weight and come with other features that make these resources physically demanding. The main question is how to handle such large amounts of ash.

Petroleum coke or petcoke is a residue from oil refinery coker units or other cracking processes. Its high sulphur content (about 8 per cent by mass) means that limestone has to be used as a bed material. In many fluidised bed boilers fired solely by petcoke some inert additive such as sand is necessary to prevent bed agglomeration in dipleg or loopseal conditions. In principle, circulating fluidised beds are intended for such fuels. Metso’s rating gives such fuels the name ‘fossil+’.

Non-predictable failures

Among the elements of concern are those that create alkali chloride, which plays a significant role in initiating the high-temperature corrosion process. Here the corrosion rate can be high and may cause non-predictable failures and an immediate need for shutdown (see Figure 1). Conventional ways to control high-temperature corrosion are via the design of the boiler’s superheater arrangement and through the selection of fabrication materials.

Figure 1
Figure 1: Two views of a section of a superheater tube whose failure resulted from high-temperature corrosion induced by alkali chloride

The threshold material temperature for superheater corrosion induced by alkali chloride is 450°C. Secondary and tertiary superheaters are typically under the same risk too. With high-chlorine fuels, low alloy steels, such as 10 CrMo9-10, can be used below that temperature. Above 450°C, high-chromium alloys that have a chromium content above 25 per cent are preferred, such as TP310HCbN.

The chlorine-induced high-temperature superheater corrosion risk that high-alkali and high-chlorine fuels present can also be reduced by changing the flue gas composition through co-firing with coal, peat and sludge, or additives containing sulphur, sulphate or aluminum silicate. Alkali in alkali chlorides can be sulphated by sulphur dioxides (SO2(g) and SO3(g)), from which hydrochloric acid is liberated.

Our solution

Metso’s solution to manage alkali chloride-induced high-temperature corrosion and the fuel blend or diet comprises three main features.

First is the monitoring of the risk and the chlorine content of the fuel by combining plant measurements with information from our Corrored analyser, which determines the nature of the combustion gas. Second, is the control of the corrosion risk through the use of additives or the management of the fuel blend. Control of carbon monoxide is also effective. Third, is the prediction of the remaining lifetime of the superheater based on the corrosion rate of the particular materials in its individual sections. One option is for the long- and short-term management of fuel diet. This involves an expert application that performs next-to-customer reporting.

This solution opens up the new possibility of monitoring the boiler condition and controlling superheater corrosion. It provides a control package to prevent such corrosion or to keep it at an acceptable level by modifying process parameters, dosing with additive or setting an optimal fuel mix. Making the corrosion risk predictable helps to avoid unforeseen shutdowns because service periods and maintenance actions can be planned in advance. It also provides a basis for the optimisation of fuel costs when considering maintenance costs.

The Corrored analyser comprises the SDG-100 sampler unit and an analyser unit. The analyser performs a real-time measurement of the chlorine and sulphur contents of the flue gas. The sampler is located in the superheater area inside the boiler, typically at temperatures of 550-950°C. The sampler collects sample gas and classifies particles for measurement. Corrosive chlorine and sulphur are in the gas phase or are condensed into small particles of less than 0.5 µm. Bigger particles (greater than 10 µm) are blown back to the boiler via the sampler cyclone. Sample gas is dissolved in deionised water for measurements. A vacuum pump sucks the sample into the unit. The gas condenses and is then analysed.

The analyser unit uses titration to analyse the chlorine and effective sulphur contents. The titrator unit gives results in mg/litre and the results are calculated as mg/m3 of flue gas. The sulphur and chlorine results are used to calculate the sulphur-to-chlorine molar ratio in the flue gas. This value allows the determination of the corrosion risk and rate.

The calculation of the chlorine content of the fuel blend also uses the value of the chlorine content of the flue gas. The measuring interval of the analyser is 10-15 minutes because of the time taken to titrate the sample.

When the corrosivity of the flue gas is known, the corrosion risk can be managed by the use of additives or by controlling the fuel diet. The liquid CorroStop additives are ferric and aluminium sulphates – Fe2(SO4)3 and Al2(SO4)3. These are injected into the upper furnace, upstream of the superheater region. Another option is to dose sulphur granulates into the fuel flow. This is called CorroStop+. The essential parameters for the effectiveness of CorroStop are the full coverage of injected sulphate over the furnace cross section, an optimal temperature window, sufficient residence time and an excess of oxygen.

The Metso DNA system covers all the tasks of the multi-fuel biomass boiler automation, from fuel yard to electrostatic precipitator to emission monitoring and reporting. Steam turbine control and steam network management are also part of the control scheme, while information management applications form an important part of plant performance monitoring.

Process conditions have a significant role in affecting the corrosion phenomenon. In addition to fuel diet and additive control, process parameters such as the steam temperature or the phasing of superheating between superheater sections can be adjusted to a preferred level. The combustion process and boundary conditions are normally complex. Optimising only a single parameter does not provide positive results when the whole process is considered.

The required dose of the liquid additive to the upper furnace upstream of the superheater region or solid additive to fuel flow depends on the process conditions, which real-time measurement can determine. Emission levels must still be held at an acceptable level after the dosing of the additive. The advanced combustion control package is a solution to solve the problem of complicated restrictions and dependencies.

Our Metso DNA system produces comprehensive corrosion information based on the Corrored analyser signal, experience of the boiler’s operation, the behaviours of fuels and material properties. The monitoring and reporting of key indicators relating to corrosion are shown as a part of the overall monitoring of plant performance.

Real-time information about the corrosion rate ensures the right actions are taken to control corrosion. The technology being developed deals with quantifying the corrosion risk level and rate, expected material loss and approximated remaining lifetime of various superheater sections in a power plant. In addition chlorine content of the fuel blend can be measured and reported.

Fuel management

Information about the properties of the fuel and its availability allows the energy producer to plan the fuel diet of the boiler. The fuel data management system stores and processes fuel data and further optimises the fuel diet. Figure 2 shows how it offers a real-time window for monitoring the deliveries of the fuel. The system merges information about the delivery of the fuel from the weighbridge and laboratory data on the quality of the fuel.

Figure 2
Figure 2: The fuel data management system

As well as what can be seen on the DCS screen, the measured fuel chlorine content can be reported and stored in the fuel data management system to be used in production planning. The second novel alternative is to replace the time-consuming laboratory analysis with a quick and reliable off-line measurement when it comes to analysis of fuel moisture. Integrated moisture measurement, such as our MR Moisture, ensures operators and plant managers receive important fuel moisture information without delay.

Process data can be combined with fuel data. Data are accessible through a web user interface. The potential that the DCS has for integration boosts the efficiency of the fuel handling controls. Planning of the boiler fuel diet can be based on fuel quality information.

The fuel diet and corrosion management application can be integrated with any type of DCS. However, with a non-Metso DNA system a separate server is needed to perform all calculations and handle and store data.

Solution in action

A CHP plant with a multi-fuel boiler that burns recycled and biomass fuels has been using our solution for controlling this high-temperature corrosion. The plant employs bubbling fluidised bed combustion technology under the conditions of 8000 kPa, 500°C, 30 kg/s and 80 MWth. The plant’s fuel is 50-60 per cent SRF, 34-44 per cent woody biomass and about 6 per cent paper mill sludge.

The SRF is mainly commercial and industrial waste. It joins the woody biomass to feed the furnace from its right and left walls. Drying of the papermill sludge reduces its moisture content to about 10 per cent by weight before it is fed into the furnace rear wall through a separate line. The flue gas cleaning system of the boiler comprises an electrostatic precipitator and a wet scrubber.

The system had been in service at the power plant for 15 months, during which it was observed for a four and a half month period. During this a few modifications were made to the sampling device and the analyser itself.

Figure 3 shows how the corrosion risk and chlorine content of the fuel blend varied over the four and half months. It confirms the alkali chloride-induced superheater corrosion risk is very high. Over a period of ten days CorroStop+, or granular sulphur, was added at a feeding rate during the first three days of 60 kg/h. This rate was later reduced. The constant addition of the granulate held the corrosion rate at a safe level – decreasing the chlorine content reduced the risk too.

Figure 3
Figure 3: Real-time variation of the corrosion index, and hence corrosion risk, and fuel chlorine content

Figure 4 shows the real-time corrosion rate for the secondary and tertiary superheater. The average measured corrosion rate is about 0.5 mm/annum. This is very high. Only during the addition of CorroStop did the rate remain at an acceptable level of 0.1 mm/annum. And Figure 5 shows the cumulative material loss over the four and a half months. Looking at the measured cumulative material loss gives an estimated lifetime of the superheater of only 3.5-4 years. The wall thickness measurement confirms the measured corrosion rate and this estimate.

Figure 4
Figure 4: Real-time corrosion rate of tertiary and secondary superheaters made of austenitic stainless steel
Figure 8
Figure 8: Cumulative material loss of tertiary superheater made of austenitic stainless steel

In a world in which greater use of non-fossil fuels will cause more corrosion problems associated with alkali chlorides, a system that determines how such damage is progressing in a boiler will clearly benefit energy producers.

Jaani Silvennoinen is Manager, Plant and Process Solutions at Metso Power, Finland. Maria Nurmoranta is Development Manager and Lasse Kauppinen is Product Manager, Analyzers, both at Metso Automation, Finland. For more information, visit www.metso.com

The article is based on a winner at this year’s POWER-GEN Europe Best Papers Awards.

More Power Engineering International Issue Articles
Power Engineering International Archives
View Power Generation Articles on PennEnergy.com