In conjunction with BG plc, Scottish Power, a leading UK multi-utility company, commissioned a feasibility study in 1993 to identify advanced NOx reduction technologies for their 4 x 600 MWe Longannet power station in Scotland. This study strongly favoured the use of gas reburn over other technologies, but at the time gas reburn technology had only been applied to much smaller boilers, mostly in the USA. Due to the technical and financial risks, several boilermakers and utilities formed a partnership and was awarded a European Union Thermie demonstration grant. The demonstration project was initiated in August 1994, and results indicate that gas reburn gives NOx reductions of 50 per cent over that which can be achieved with low NOx burners, with little or no increase in unburned carbon levels, and no significant effect on the general boiler performance.

Environmental concerns about the long term operation of coal based power stations, particularly over the emissions of oxides of nitrogen (NOx) has created a new, world-wide business opportunity for clean coal technology. This is demonstrated in this instance by the application of gas reburning technology, which is capable of achieving a reduction of about 70 per cent when used in conjunction with low NOx burners.

Studies, particularly in the USA, have shown gas reburning to be a cost effective technology, with demonstrations on units of up to 300 MWe in capacity. A number of European companies are collaborating to demonstrate gas reburn on a 600 MWe coal fired boiler at Scottish Power’s Longannet power station.

The other partners in the project are Italy’s ENEL SpA and Ansaldo Energia SpA, Electricité de France, Electricity Supply Board of Ireland, and Mitsui Babcock Energy Ltd. and BG plc of the UK. The Longannet demonstration is recognised as a significant step towards the commercialization of clean coal technology and was granted financial support from the European Union’s Thermie programme.

Reburning is a hybrid combustion modification technique. In gas reburning, natural gas is injected into the furnace after the primary, coal combustion zone to produce a fuel-rich region where the NOx is reduced by reactions with hydrocarbon radicals.

This is termed the reburn zone. Further downstream, in the burnout zone, air is injected into the boiler to complete combustion of any unburned coal from the primary zone and the carbon monoxide formed in the reburn zone. This air is known as overfire air.

Although any hydrocarbon can, theoretically, be used as the reburn fuel, natural gas has distinct advantages which make it particularly attractive for large utility boilers. Reductions in NOx emissions of between 50 and 60 per cent have been reported when around 20 per cent of the thermal input is replaced by natural gas.

The principal aims of the Longannet project are to demonstrate over a period of twelve months the use of gas reburning technology on a 600 MWe coal wall-fired boiler and to develop and validate comprehensive engineering and process design tools to enable the technology to be replicated on similar boilers throughout the world.

Longannet has four 600 MWe coal fired units and is the second largest power station in the UK. The units were commissioned between 1969 to 1971 and, with the planned boiler refurbishment programme including the installation of gas reburning, it is anticipated that Longannet will continue to be operational until 2020.

The boilers were originally designed to burn low sulphur, medium volatile coal and to supply 500 kg/s of steam at 169 bar and 568°C, with reheat of 359 kg/s steam to 568°C.

Each boiler has 32 pulverised coal burners mounted on the front wall, arranged in four horizontal rows of eight burners. However, for maximum continuous rating (600 MWe) only 24 burners are necessary. The original burners have now been replaced with low NOx burners, with the last replacement being undertaken in parallel with the gas reburn modifications on Unit 2.

The maximum NOx emissions guaranteed with these burners is 650 mg/Nm3 (at six per cent O2).

Design conditions

It is well recorded that the conditions in the reburn zone have the greatest impact on overall NOx reduction, this has been investigated by Mitsui Babcock Energy Limited (MBEL) for a range of coals. Moreover, defining the reburn gas input and stoichiometry also fixes the primary zone heat input and stoichiometry.

The required furnace stoichiometries and residence times, and hence the elevations for the gas and overfire air injectors, were determined primarily from the tests undertaken on a 160 kWth down-fired test furnace operated by MBEL at their Technology Centre in Renfrew, Scotland.

The tests were carried out using coal from the Longannet mine complex. The starting point for these tests were published data on optimum reburn zone stoichiometries which suggest stoichiometries ranging from 0.8 to 0.9 for gas heat inputs ranging from around 10 to 20 per cent.

For the Longannet project, 20 per cent gas reburn was selected as the design condition. The tests at MBEL indicated a reburn zone stoichiometry of 0.9 and consequently a primary zone stoichiometry of 1.12. As 1.17 is the nominal baseline condition, this was selected as the burnout zone stoichiometry. The tests at MBEL also indicated that increasing both the primary zone and reburn zone residence times reduced the final NOx levels.

Factors which influence mixing efficiency include injection velocities, momentum, number of jets, shape of the nozzles and their angle and position around the furnace perimeter.

These factors were studied by ENEL and Ansaldo using 3-D Computational Fluid Dynamic (CFD) codes. To compare mixing effectiveness between the many options a Dispersion Index was used which gave a global indication of the mixing rate. The index had been successfully used previously to design oil reburn systems.

The predictions indicated that it would be necessary to utilise flue gas recirculation (FGR) in order to increase the momentum of the gas jets. The mixing of the reburn gas (plus FGR) with the coal combustion products in the reburn zone for the final arrangement of eight gas injectors on the front burner wall and 16 injectors on the rear wall.

The better the mixing the more the Dispersion Index tends to zero.

CFD models were extensively used to determine the number and position of overfired air injectors.

The results indicated that 12 injectors on the front wall with two injectors on each of the side walls, giving 16 in total, would give the desired mixing.

Predicting reductions

The final NOx emission levels were predicted by BG Research and Technology Centre using detailed chemical kinetics modelling based on models developed in-house using schemes available from the literature and validated against experimental data where possible. The Longannet furnace was again split into the three zones: primary, reburn and burnout zone.

The kinetic modelling was undertaken for a wide range of operating conditions taking the zone stoichiometries, residence times and furnace temperature predictions from the various process design modelling activities. These predictions gave NOx reductions up to 65 per cent, although there was a degree of uncertainty particularly surrounding factors such as zone temperatures and the degree of coal combustion in the primary zone. Main coal burnout can have a noticeable impact on the actual reburn zone stoichiometry, which in the extreme, could result in the zone becoming fuel lean.

Oxidising conditions in the reburn zone severely impair the NOx reduction efficiency of the reburn process. This was shown for various reburn temperatures, and the results arising from the 160 kWth test furnace confirmed this finding.

Designing and integrating the reburn control and safety interlock systems into the automatic boiler control system was undertaken by ESB and ENEL using the experience gained from ENEL’s oil reburn work. The new system is fully automated.

Positive results

The objective of the test programme was to confirm the performance of the gas reburn system at design conditions by baseline testing and to investigate the sensitivity to the various design parameters by optimization testing.

The test programme is currently 70 per cent complete and the results of the first phase (baseline testing) have been fully analysed. The high load tests were carried out at 530 to 550 MWe whilst the low load tests were carried out at 420 MWe.

The baseline NOx emissions under single stage combustion were measured as 760 and 670 mg/Nm3 at six per cent O2 for high and low firing patterns respectively for a primary combustion zone stoichiometry of 1.1.

With gas reburn in service and 20 per cent of the heat input being injected through the reburn ports the NOx emissions varied from 440 to 275 mg/Nm3 at six per cent O2. The reason for the wide variation is due to the sensitivity of the performance to firing pattern and reburn zone stoichiometry. The overall NOx reduction performance can be summarized with respect to reburn zone stoichiometry.

The combustion performance of 530 MWe baseline single stage combustion at a stoichiometry of 1.1 in the primary combustion zone with respect to CO and carbon in ash gave less than 20 mg/Nm3 at six per cent O2 of CO and carbon in ash levels of 4 to 5.5 per cent for high and low firing respectively.

This was somewhat unexpected as higher firing patterns normally give higher carbon in ash levels.

The reason for this anomaly is not clear, but it is possible that different mills producing different coal product fineness may have been part of the reason.

The CO emission levels at 550 MWe with gas reburn in service show similar levels (less than 20 mg/Nm3 at six per cent O2) to single stage combustion at a reburn zone stoichiometry of 0.9. As the reburn zone stoichiometry reduces the CO level increases with levels of 140 and 350 mg/Nm3 at six per cent O2 at reburn zone stoichiometries of 0.85 and 0.825 respectively.

The carbon in ash levels at 550 MWe with gas reburn in service show that levels increase as the gas reburn stoichiometry decreases, again with the lower firing pattern showing higher carbon in ash levels than the higher pattern. At a gas reburn stoichiometry level of 0.9 the carbon in ash is five and seven per cent compared to four and 5.5 per cent for single stage combustion at a primary combustion zone excess air level of ten per cent.

This is a very modest increase in carbon in ash considering that a reburn zone stoichiometry of 0.9 gives a NOx reduction of 42 to 49 per cent. At a reburn zone stoichiometry of 0.85 the carbon in ash increases to 7.8 to 9.0 per cent and approaching the point where the carbon in ash is increasing at a faster rate than the NOx is decreasing. This would tend to indicate that the optimum reburn zone stoichiometry lies between 0.85 and 0.9.

Market opportunity

Scottish Power has already indicated that they will implement gas reburning on the remaining three boilers at Longannet if the demonstration shows that the technology is a cost effective method of reducing the NOx emissions.

The future prospects for commercial exploitation of this technology are being investigated. A world-wide survey of potential coal fired retrofit boilers indicates a substantial market opportunity, particularly in the first quarter of the next century when more stringent environmental legislation is expected.
Tables

Table 1. Impact of primary zone burnout on reburn zone stoichiometry
Table 2. Test programme
Table 3. Gas reburn NOx reduction performance