Combustion optimization to cut NOx

A combustion analysis and equipment retrofit provided a 48 per cent NOx reduction at a UK coal plant, writes Steve Billett


The Ultra-Low NOx burner

Credit: RJM International

When Rugeley B Power Station was commissioned in 1970, its two state of the art coal-fired 500 MW units were supplying baseload electricity to over one million homes and businesses in the densely populated Birmingham area of the UK’s industrial Midlands.

Originally equipped with front-wall fired Foster Wheeler burners, the plant was upgraded with first generation Burmeister & Wain Energy (BWE) low NOx burners in 1996 to reduce NOx levels to below 650 mg/Nm3 and then in 2007, to ensure the plant fell within the new Large Combustion Plant Directive (LCPD) threshold of 500 mg/Nm3 NOx, a Doosan Babcock boosted over fired air (BOFA) system was additionally installed.

Fast-forward to 2014 and, despite the fact that the plant had been delivering power reliably and efficiently for over 40 years, its future was starting to look less certain – a consequence of volatility in the UK coal market and the government’s commitment to an accelerated decarbonization of the nation’s power generation sector. Set against this challenging background, Rugeley’s management team decided to investigate the possibility of joining the Transitional National Plan (TNP) for large combustion plants to give it additional operational flexibility up to June 2020 and reviewed how it could best begin the process of achieving the new NOx threshold of below 200 mg/Nm3 under the Industrial Emissions Directive – as quickly and as cost-effectively as possible.

Under the terms of the TNP, generators were permitted an annual NOx tonnage allowance. This established a relationship between daily NOx emissions and total running hours and effectively gave power producers three possible options for continued operation:

1) No additional investment in primary NOx reduction technology and maintaining generation with NOx at 400 – 500 mg/Nm3 with hours severely limited post-2015;

2) Some additional investment in primary NOx reduction to achieve around 300 mg/Nm3 with hours somewhat limited post-2015;

3) A more substantial investment in primary and/or secondary NOx reduction technology to reduce NOx levels to below the IED mandate of 200 mg/Nm3, thus ensuring the maximum possible number of generation hours and securing the option for continued plant operation beyond 2020.

Faced with these three scenarios, Rugeley’s management team appointed RJM International to focus on option two.

With Rugeley, RJM followed its tried-and-tested approach already successfully applied to over half the UK’s coal-fired capacity. This consisted of a detailed front-end engineering design (FEED) study, comprising an air distribution analysis (ADA) survey, an assessment of the plant’s Baseline Performance including baseline testing, as well as computational fluid dynamic (CFD) analyses and physical modelling programmes.

Site survey importance

RJM knew from experience that relatively small variations in data can end up having a disproportionate effect on predicting plant performance and so needed to source a sufficient volume of accurate data from which the mathematical and computational models could then be developed.

Plant engineers’ drawings were consulted and validated, other data was gathered using existing plant instrumentation and a third party testing team recorded more data via additional test points across the unit. In addition, historical performance testing reports were consulted to help validate the baseline CFD modelling predictions.

However, once all this data had been gathered and reviewed, it became clear that two key site-specific challenges existed in this project.

Firstly, compared to other similar UK plants where RJM had already completed Ultra-Low NOx Burner upgrade projects, the primary air flow at Rugeley was much greater, meaning that it would be more difficult to maintain an effective low NOx environment within the burner flame as that requires a strong Internal Recirculation Zone (IRZ) which could easily be destroyed by the momentum of the high primary air flow.

Secondly, the site survey confirmed that during combustion, fuel volatiles were being released too close to the BWE burner throat, causing elevated burner temperatures resulting in thermal destruction of the burner. This in turn meant that as the burners were not performing as intended, they would be difficult to accurately represent within the CFD model.

Baseline CFD modelling

Following a detailed site survey, normally the next stage in an RJM upgrade project is to enter all the key baseline data into a CFD model to faithfully replicate current plant performance.

Figure 1. Full furnace CFD model

Credit: RJM International

However, in this case, bearing in mind that the baseline burner performance could not be relied upon, RJM also consulted the 1997 burner performance test data which was gathered when the BWE burners were new. These performance tests for the new BWE burners were conducted prior to the addition of the BOFA upgrade.

Whilst helpful, this historic data also introduced the potential for a number of inaccuracies as, for example, the differences in combustion air distribution and fuel distribution between the burners was unknown; boiler air ingress and PF grind data were unverified and other test data could not be validated due to the lack of primary source calculations.

Aware of these limiting factors, RJM sought to replicate 1997 actual performance within its own CFD model, setting itself a target of a predicted NOx value of +/-15 per cent as one of the thresholds for validating the model. In fact, the difference was just 2.8 per cent: 640 mg/Nm3 1997 actual, versus 622 mg/Nm3 recorded in RJM’s baseline CFD model. This helped to confirm that the CFD model was performing accurately.

A second NOx comparison test to evaluate to what extent actual burner damage was impacting on performance showed a difference of just 12.8 per cent: 450 mg/Nm3 recorded in RJM’s site survey, versus 392 mg/Nm3 in its CFD model.

These two performance comparison tests gave RJM’s engineers the confidence that their Rugeley baseline CFD model was faithfully replicating the plant’s actual performance.

The next stage was to run that model and use it to help identify what hardware changes would need to be delivered and installed to meet the required NOx reduction.

Upgrade CFD modelling

RJM knew from previous experience that to deliver significant NOx reductions on a front wall, bituminous coal-fired boiler, three key parameters had to be met.

Firstly, the burner stoichiometry would need to be reduced so that the amount of available oxygen in the burner flame minimizes the opportunity for NOx to form at the burner.

Secondly, the dynamics and control functions of the new Ultra-Low NOx burner must be able to maintain a stable and robust flame when firing under the required sub-stoichiometric conditions.

Thirdly, air would have to be reintroduced elsewhere in the furnace so that full combustion could be completed, thus mitigating the unburnt carbon and CO created as a result of firing the burners sub-stoichiometrically.

With these factors in mind, RJM set about building a single burner model within the CFD programme. Focusing on a single burner first not only allows for greater granularity and detail within the CFD model, but also means the burner can be analyzed in isolation, so that key design elements of the burner design required to control combustion can be evaluated in detail.

The single burner CFD model confirmed how a tightly-controlled and sculpted IRZ could be created in such a way as to create a stable flame, whilst also ensuring that the burner’s exposure to the flame itself would be limited, thus resolving the issue of burner thermal damage.

Having matched the baseline single burner performance and demonstrated its upgrade burner solution, RJM then constructed a full furnace CFD model of all 28 burners and ran that model through a series of design iterations, before arriving at the final model.

The CFD image in Figure 1 highlights that a number of key performance criteria had been successfully met:

ࢀ¢ The blue to white path-lines of overfire air and curtain air show how oxygen is consumed during combustion;

ࢀ¢ The curtain air ports show strong penetration, providing protection against fire-side corrosion;

ࢀ¢ The overfire air ports can be seen to provide sufficient penetration across the upper furnace to prevent CO from reaching the superheater tubes and the convective pass, whilst ensuring full coverage, thus preventing the majority of the unburnt fuel from escaping into the convective pass;

ࢀ¢ The burners are shown to work successfully with the other air delivery systems and to be capable of controlling the high primary air and fuel momentum with a short and tight flame.

In terms of data results, the CFD model predicted the following results, including a significant reduction in NOx of 41 per cent from the original low NOx baseline.

The outputs from this CFD model were then used in the detailed design and mechanical engineering of the upgrade firing system solution. These changes were confirmed following a technical and commercial review process with the plant managers at Rugeley.

The key deliverables that formed the engineering elements of the upgrade, included full replacement of the existing BWE burners with 28 new RJM Ultra-Low NOx CleanAir Burners; a new curtain air system to protect the furnace wall whilst operating at lower stoichiometry; additional BOFA ports and modifications to the existing BOFA ports. Some smaller changes were also made, including the fitment of new air ducts and oil burner modifications, further helping to keep costs and downtime to a minimum.

NOx down 48 per cent

Following the RJM upgrades, the unit was commissioned in just 24 hours, providing almost immediate generation at full load. And after just one week of optimization, the unit was delivering NOx levels of between 178 mg/Nm3 and 209 mg/Nm3 across the load range, with good char oxidation, as well as producing saleable ash.

This result replicated the strong reduction predicted by the CFD model of 184 mg/Nm3. In fact, the upgrade combustion system was able to control reported NOx values to within the IED mandate of <200 mg/Nm3 from 50 per cent to 90 per cent load, with CIA levels maintained with the range of 6-8 per cent.

Overall, the combustion optimization study and ultra-low NOx equipment retrofit installation provided a 48 per cent NOx reduction across the load range, whilst retaining excellent combustion efficiency and ash product within saleable CIA limits.

Enabling Rugeley Unit 7 to operate within the IED NOx mandate signified a step change in the potential NOx reduction now attainable, using primary measures only. No longer are generators confined to the difficult commercial decision of investing heavily in a plant’s future by implementing a secondary measures NOx reduction system. Now the opportunity exists to extend plant life under the IED regulations, with only a relatively small investment in primary measures technology.

Due to market pressures, Rugeley Power Station ceased market operations in June this year. However, the equipment retrofit confirms what can be achieved on primary measures alone, if one takes the time to understand – in great detail – what is actually going in in terms of the process of combustion within the furnace.

Steve Billett is Senior Project Manager at RJM International.


No posts to display