In recent years, RWE Power has carried out extensive R&D activities to optimise existing flue gas treatment systems as well as to examine new technologies for effective mercury control in order to further reduce Hg emissions and to be able to comply with future emission limits.

Mercury control techniques have been evaluated and optimised in a holistic way, which covers the entire process chain of the power plant and all relevant incoming and outgoing media streams as well as the safe disposal of the captured mercury.

Challenging EU Hg emission limits

With their existing flue gas treatment capabilities, RWE’s coal-fired power plants will safely comply with both the current and the upcoming German national mercury emission limits, which will become effective in 2019.

However, the introduction of EU-wide flue gas mercury emission limits via the BREF-LCP process, with a bandwidth of Hg < 1-4 μg/m3N for hard coal and Hg < 1-7 μg/m3N for lignite (annual average for existing large combustion plants >300 MWt), will again drastically tighten the framework conditions for the operation of RWE’s power plant fleet

from 2021 onwards. Currently, no mercury control technique is commercially available to safely and reliably achieve emission levels below the upper limits of the bandwidth of 4 μg/m3N for hard coal and 7 μg/m3N for lignite in an industrially applicable and economically affordable way.

Lignite-fired power plants usually operate mine-mouth, ie, are tied into local mines and have very limited or no options for changing fuel quality, unlike hard coal units using imported coal. Mercury emissions vary from site to site, depending on lignite mercury content and power plant configuration. In the BREF-LCP process, these differences were not sufficiently considered, and the plants used as a reference, such as IGCC plants or fluidised bed combustion systems with fabric filters, were not representative. Only three lignite plants with pulverised fuel firing considered in the LCP-BREF process reported comparably low Hg emissions, but it remains uncertain whether these emission levels can be maintained over extended periods and under all operating conditions.

In the USA, the Environmental Protection Agency (EPA) had a similar experience when assembling a database for the Mercury and Air Toxics Standards (MATS), which became effective in 2012.[1]

For lignite-fired power plants, EPA acknowledged significant differences in Hg speciation and control options compared to hard coal, and as a consequence the MATS Hg emission limits for lignite are less stringent than for hard coal (lignite, 4.0 lb/ TBtu, or 5.0-5.4 μg/m3N; hard coal, 1.2 lb/ TBtu, or 1.5-2.2 μg/m3N).[2]

However, several commentators on MATS stated that this so-called ‘beyond-the-floor’ Hg emission limit for lignite was derived from three samples only, taken during a single test at one lignite-fired power plant.1 If EPA had followed the normal procedure for setting emissions standards, the Hg emission limit for lignite would have been 11 lb/TBtu, or around 14 μg/m3N – based on the emissions reductions achieved by the average of the USA’s top 12% best controlled lignite-fired power plants.

Activated lignite, HOK®

Based on experience and taking into consideration available technology options for Hg control in lignite-fired power plants, RWE has opted for activated carbon injection (ACI) in an entrained flow process. RWE’s affiliate Rheinbraun Brennstoff GmbH produces an activated lignite called HOK® (“Herdofenkoks”), produced from Rhenish lignite via the so-called rotary hearth furnace process. The special properties of Rhenish lignite, coupled with the activation conditions in the production process, yield an activated carbon which has the capability for retention of a large number of pollutants due to its large specific surface area, its favourable pore radii distribution and its catalytic behaviour. HOK® activated lignite has been commercially used for many years as an adsorption and filter agent for waste gas and effluent treatment in a vast range of applications, and several American power producers also use HOK® in their ACI systems.

Consequently, RWE’s R&D programme focuses on a mercury control technique based on HOK® optimised for lignite-fired power plants in the Rhenish lignite area. The HOK® injection into the flue gas ducts had to be adapted to larger dimensions and higher flue gas flows. For example, the flue gas volume flow in the 1000 MW BoA1 lignite fired power plant is approximately 3.2 million m3N/h in two separate flue gas trains (1.6 million m3N/h each) and the flue gas duct dimensions at the injection ports are 13 x 6 m (Figure 1). Due to the abrasive nature of the sandy fly ash, internals in the flue gas duct have to be avoided.

As an alternative to the well-established dry injection of HOK® dust, a new milling and injection concept was developed, where HOK® is first mixed with water and then wet milled (Figure 2) to obtain a HOK® slurry of finely dispersed particles in water, similar to black ink, which is finally sprayed into the flue gas stream. In this concept, the diameter of the HOK® particles is reduced to less than 10 μm (d50 value), creating conditions for an expected improved transfer of mercury from the gas phase to the particles.

Extensive testing

Since 2014, Hg emission reduction trials have been conducted in RWE’s flue gas treatment pilot plant at the Coal Innovation Centre, Niederaussem. The pilot is installed in a flue gas slipstream from the BoA1 power plant and treats approximately 30 000 m3N/h, or 1% of BoA1’s flue gas volume flow, with a combination of limestone- based wet flue gas desulphuriser and wet electrostatic precipitator. The experiments included both the adsorptive binding of Hg to activated carbons and HOK® in the entrained flow process as well as the fixation of Hg in the scrubber sump. Both dry and wet injection of adsorbents was tested, allowing a first comparison of both process variants.

Important process parameters such as adsorbent residence time, various nozzle geometries, particle sizes and droplet spectra, and different concentrations in the flue gas, were studied. The tests confirmed that both gaseous elemental Hg and oxidised Hg compounds are adsorbed, which is important, as Hg is predominantly elemental in the gas phase in the flue gas of the lignite power plants in the Rhenish lignite area.

The new, proprietary wet process route with the injection of a HOK® slurry into the flue gas passed the ‘proof-of-concept’, although the observed Hg reduction achieved by the dry process route was slightly better in this first test. Also, the possibility of adsorbing dissolved Hg in the scrubber sump and accordingly reducing the concentration of Hg in the liquid phase was confirmed. 


In a next step, larger scale test campaigns were carried out, in 2016 and 2017, in a fluidised bed boiler at the Berrenrath lignite refining plant (which produces lignite dust for industrial applications). This boiler combusts lignite and sewage sludge and has a flue gas volume flow of about 400 000 m3N/h. Due to the sewage sludge co- incineration, the plant was already equipped with conventional dry HOK® injection to control mercury and other contaminants. The results of the wet injection test campaigns could be compared directly with those of the conventional technique. The experiments again confirmed the feasibility of the wet injection technique. During the wet HOK® slurry injection, the Hg concentration in the cleaned gas could be kept at the same level as with the dry HOK® injection using air, but at a dosing level 20% to 30% lower. No negative effects, on other emissions, ash removal, fouling of flue gas ducts or arcing in the electrostatic precipitator, were observed.

To validate the technique of wet milling and injection of HOK® slurry, a large-scale pilot plant (Figure 3) was constructed and commissioned in 2017 on one flue gas train of the BoA1 unit, representing a final scale- up to a flue gas volume flow of approximately 1.6 million m3N/h. The pilot plant was developed in-house and existing equipment was used where possible to minimise the planning and procurement period.

In total, 16 injection ports were retrofitted into the flue gas duct of BoA1 to study the influence of HOK® distribution on Hg adsorption as well as variations in HOK® slurry properties and other injection parameters. To measure the Hg reduction, a continuous Hg analyser with two heated sample lines was installed, taking samples alternately from the two flue gas trains of BoA1 with and without HOK® slurry injection.

The pilot plant for Hg reduction was operated from October 2017 to June 2018 in the Niederaussem Coal Innovation Centre. Its main purpose was optimisation of HOK® slurry production, handling and injection. Test runs usually lasted one to three days, too short to obtain a final answer to the question: which Hg levels are achievable in commercial operation? As expected, Hg reduction changes with the injected quantity of HOK® and it was confirmed that the correlation is non-linear, ie, doubling the HOK® injection quantity does not double the Hg reduction. There was little difference in Hg reduction with changes in the number of operating injection nozzles, the injection pattern, the HOK® content in water or the HOK® particle size, as long as the total injected HOK® quantity remained the same.

The test programme conducted at the full- scale Hg reduction pilot plant also included Hg balancing. Ash samples were regularly analysed and it was proven that HOK® is removed very effectively from the flue gas by the electrostatic precipitator. It was also confirmed that the Hg content in the power plant residues typically sent to landfill remains low and does not exceed required limits or pose a threat when handling the ash.

From R&D to commercial operation

In parallel with pilot plant operation, planning and construction of a pre- commercial Hg reduction demonstration plant was begun, which will test both HOK® injection processes – wet and dry – to be in long-term operation starting 2019. The demonstrator will collect information on long-term operating experience in terms of mercury reduction, wear, downtime, maintenance and operating costs, which will be decisive in making the choice of preferred HOK® injection method – wet or dry – for the final commercial large-scale Hg control system.

The R&D projects and trials with increasing operating times and growing plant dimensions have led to a better understanding of the adsorption process, especially the mechanisms of mass transfer of Hg into and onto the highly porous adsorbent and its diffusion or binding into active areas of HOK®, which are the limiting factors for a higher loading or utilisation of HOK® and other activated carbon products. However, the ultimate goal of RWE’s R&D programme is to transfer a technology to commercial use which is easy to operate and maintain. We are convinced that HOK® injection has the potential to further reduce Hg emissions, to achieve future emission limits in compliance with BREF-LCP and to enable continued operation of lignite-fired power plants to provide reliable power in times of low wind and no sun. 


[1] National Emission Standards for Hazardous Air Pollutants From Coal and Oil- Fired Electric Utility Steam Generating Units and Standards of Performance for Fossil-Fuel- Fired Electric Utility, Industrial-Commercial- Institutional, and Small Industrial-Commercial- Institutional Steam Generating Units, Federal Register, Vol. 77, No. 32, 16 February 2012

[2] Grenzwerte für Quecksilberemissionen aus Kohlekraftwerken, Kather et al, VGB PowerTech, 12, 2015