Abstract

Many countries have strict emission limits for nitrogen oxides (NOx); thus NOx control
systems are widely deployed. India has recently introduced NOx emission limits, which
require that pollution control technologies must be installed on most coal-fired plants.
However, operational experience with NOx control systems in India is limited to primary
measures only. Additionally, Indian coals have a high level of inherent ash, which is also
highly erosive and can influence the behaviour of some NOx control systems and thus affect
the selection process. Primary measures for NOx control from coal-fired power plants
include low NOx burners (LNBs), overfire air systems (OFA), fuel reburning, flue gas
recirculation, fuel biasing, low excess air and combustion optimization. Secondary
NOx control includes selective catalytic reduction, selective non-catalytic reduction and
multi-pollutant control systems. Retrofit of primary measures (LNBs and OFA) in India has
been recommended to take place during next-scheduled plant outages; for many plants this
could occur by 2019. However, control strategies for individual plants will be needed to
ascertain the appropriateness of installing post-combustion technologies or various
combinations of NOx control measures. Hence installation of secondary NOx controls is not
expected before test results from secondary controls on selected Indian power plants units
are known.
high-ash coal, NOx controls, SCR, SNCR, low NOx burners, overfire air, India emission
standards
Issue Section:
 Articles

Introduction

Since the 1970s various technologies to control emissions of nitrogen oxides (NOx) have
been used in countries with relevant emission standards. However, this is not the case in
India, which has high-ash coals and until recently did not have NOx emission standards.
NOx emissions control legislation has been introduced and NOx control retrofits are
expected on a large scale. As some of the NOx reduction (DeNOx) technologies (secondary
measures) have not been used in Indian conditions (up to 80 g/m 3 of ash in the flue gas),
utilities are concerned about their applicability to Indian power plants. At the same time,
vendors are confident that their systems can operate in such an environment as they have
been successfully used in high-ash applications, including lignite-fired plants and cement
kilns where ash loading in the flue gas can be as high as 100 g/m 3.

1 India—new norms and challenges

The new Indian emission standards for thermal power plants (TPPs) regulate water
consumption, particulate matter (PM), SOx, NOx and mercury (Hg) emissions, and are
obligatory from 1 January 2017 for new plants. Existing units have two years in which to
comply. Regarding NOx, the proposed new regulations stipulate that thermal power plants
installed before 31 December 2003 emit no more than 600 mg/m 3 NOx, plants installed
between 31 December 2003 and 31 December 2016 emit no more than 300
mg/m3 NOx and plants installed from 1 January 2017 onward limit NOx emissions to 100
mg/m3 (see Table 1).

Table 1

New emission norms in India [1]a

TPPb (units) TPP (units) installed TPP (units) to be

Emission installed before after 31/12/2003 and installed from
parameter  31/12/2003  before 31/12/2016  01/01/2017 

Particulate
matter (PM)  100 mg/m3,c  50 mg/m3  30 mg/m3 

600 mg/m3 for
units <500 MW; 600 mg/m3 for units
200 mg/m3 for <500 MW;
Sulphur units 500 MW and 200 mg/m3 for units
dioxide (SO2)  above  500 MW and above  100 mg/m3 
 TPPb (units) TPP (units) installed TPP (units) to be
Emission installed before after 31/12/2003 and installed from
parameter  31/12/2003  before 31/12/2016  01/01/2017 

Nitrogen oxide
(NOx)  600 mg/m3  300 mg/m3  100 mg/m3 

Cooling tower to be installed in place of Maximum water

once-through cooling; maximum water consumption, 2.5
Specific water consumption ≤3.5 m3/MWh. Water m3/MWhd; zero
consumption consumption ≤3.5 m3/MWh for all existing wastewater
limit  cooling tower-based plants.  discharge 

1. TPP (units) installed before 31/12/2003

(500 MW and above capacity)
2. TPP (units) installed after 01/01/2003 up
to 31/12/2016
Mercury (Hg)  3. TPP (units) installed from 01/01/2017  0.03 mg/m3 
a
These norms are mandated to be met within 2 years by all existing TPP units and from
01/01/17 onward by all new TPP units. bTPP = thermal power plants. cmg/m3 = milligrams
per normal cubic meter at 6% O2, 273.15 K and 0.1013 MPa. dAmended in 2018 to 3
m3/MWh.
Open in new tab

Indian thermal power plants fire mostly indigenous sub-bituminous coals and washery
middlings (K. Nandakumar, Fuel Perf Tech Pro LLC, Irving, TX, USA, personal
communication, 2018). The coal has relatively low moisture and low sulphur content but a
high ash content. As reported by Barnes [2], three-quarters of current Indian coal production
has an ash content of 30% or more, with some of the highest ash coals approaching 50%;
coal traded on the international market rarely exceeds 15% ash. Much of the ash is inherent,
which means that it is present as small particles of mineral matter embedded in the
combustible part of the coal, making it difficult to remove to levels below 30% prior to
combustion [2, 3]. Furthermore, ash in Indian coals has a high percentage of abrasive and
erosive solids including silica, aluminium oxide and iron oxide, with silica in alpha form being
particularly erosive. Firing such high-ash fuel can create many problems, including
increased fouling, which causes reduced heat transfer and erosion of heat transfer surfaces
in horizontal pass, second pass and downstream ducting and equipment. Indian utilities
minimize these detrimental effects by appropriate boiler design. Furnaces firing Indian coal
are larger, both in width and in height, than boilers firing imported coal; flue gas velocities
are lower; and the sizing and selection of various auxiliary systems can also differ [4, 5].

In terms of NOx control in India, utilities’ experience is limited to primary measures: low

NOx burners (LNBs), overfire air (OFA) and fuel biasing. There are no secondary
NOx controls on any units to date (May 2018) and the effect of high-ash coal on these
systems is unknown. Consequently, utilities have limited experience with NOx controls and
will rely on equipment suppliers’ expertise. At the same time, the utilities are reluctant to
decide on secondary measures until results for pilot tests in India, currently underway on
units owned and operated by NTPC Ltd, are clear.

As with all retrofits and the introduction of technologies to markets, challenges exist. In
India, these include high-ash, highly erosive coals; technical difficulties including space
constraints; financial issues as it is costly to introduce multiple technologies at the same
time; a lack of local suppliers and a subsequent need to import technologies as well as all
materials and reagents; a lack of local skills and expertise; and a lack of experience among
utilities in continuous emissions monitoring.

2 NOx controls

NOx control technologies can be broadly divided into two main categories: primary or
combustion measures and secondary or flue gas control systems. The rates of
NOx reduction and the costs vary considerably. They can be used alone or in combination,
depending on the level of NOx reduction required.

2.1 Primary measures

Primary measures—including LNBs, OFA, fuel biasing, low excess air, fuel reburning, flue
gas recirculation and combustion optimization—are the least costly approach to obtain an
initial reduction in NOx emissions for any coal-fired plant. By controlling coal combustion
conditions such as the flame temperature, fuel:oxygen ratio and fuel residence time they
reduce NOx emissions. Currently, LNBs are standard DeNOx technology in countries with
relevant standards. It is relatively simple to include combustion controls in the construction
of new boilers, but it is more complicated to modify (or retrofit) combustion controls in
existing boilers because combustion controls can adversely affect boiler operation and lead
to increased emissions of NOx and CO, and more carbon in the ash. Burner flame geometry
also changes and the flame can impinge on the water walls. Hence careful design is
essential, as is the measurement and control of various combustion parameters after
retrofitting primary measures. Advanced sensors and controls can be used that allow
accurate real-time mapping of the furnace conditions

Most primary measures have been already applied to boilers using high-ash coals. In India,
most boilers are the tilting tangential type, have OFA and are supplied by the Indian
company BHEL (Bharat Heavy Electricals Limited). Many also apply fuel biasing in the coal
pipe [6]. However, by the end of 2016, only 24 units had LNBs installed [7]. The new Indian
emission standards mean that the majority of existing and all new power plants (online after
1 January 2017) have to install primary measures in the form of LNBs and OFA systems. As
it is relatively quick to install these systems, it is recommended that they be retrofitted during
the next outages, starting with the newest units. Hence retrofits are expected on a large
scale [8]; in fact, most power plants scheduled retrofits for 2017 and 2018.

Several manufacturers offer LNBs and OFA systems, some of which are specifically
designed for the Indian market; as such they can operate with erosive, high-ash coals. For
example, Doosan Babcock and Doosan Heavy Industries’ high-ash coal burner (see Fig. 1)
has been designed for wall-fired 800-MW pulverized coal-fired units and is suitable for both
retrofit and new-build applications. As noted by D. Smith (Doosan Babcock Limited,
personal communication, 2017), the objective of the LNBs is to provide excellent
performance in response to market requirements:

Fig. 1

Open in new tabDownload slide

 High-ash coal burner designed by Doosan Babcock and Doosan Heavy Industries for Indian
market (Source of the photo: D. Smith, Doosan Babcock Limited, UK, personal
communication, 2017)

 up to 90-MWth burner size;

 primary NOx emissions below 300 mg/m3;
 very low unburnt loss (UBL);
 low operating excess air <15%;
 >3:1 burner turndown; and
 • improved burner life and wear resistance to more than 16 000 hours.

2.2 Secondary measures

Secondary NOx control measures include selective catalytic reduction (SCR), selective non-
catalytic reduction (SNCR) and a combination of the two. Both technologies neutralize
NOx into nitrogen and water, either with the presence of a catalyst (SCR) or without
(SNCR). Generally, secondary measures are applied to plants that need to reduce
NOx limits beyond the capability of primary measures.

2.3 SCR

SCR achieves the highest NOx removal rates at 80–90%, but it is also the most expensive
option [9]. SCR has three main configurations for a coal-fired power plant (see Fig. 2). In the
first configuration, known as hot-side, high-dust, an SCR is installed upstream of the PM
control, between the economizer and the air heater. This is the most widely used version in
coal-fired power plants, as the temperature of the flue gas between the economizer and air
heater is ideal for the SCR reaction. In this configuration, the catalyst is exposed to the fly
ash and chemical components of the flue gas.

Fig. 2

Open in new tabDownload slide

SCR configurations [10]

In the low-dust arrangement, an SCR is installed downstream of the PM control equipment.
Such a configuration, also referred to as hot-side, low-dust, reduces degradation of the
catalyst by fly-ash erosion, but needs a costly hot-side electrostatic precipitator (ESP) or a
flue gas heating system to maintain the optimal temperature required for the catalytic
reaction of NOx and to avoid formation of ammonia.

In the tail-end configuration, also referred to as cold-side, low-dust, the SCR is downstream
of the flue gas desulphurization (FGD) unit. It can be retrofitted on installations with limited
space, but it also requires heating of the flue gas.

There are many SCR suppliers and SCR design is highly site-specific. Consequently, SCR
systems differ between power plants, and the capital and operating costs vary.

2.4 Solutions for hot-side, high-dust SCR applied to high-ash coals

According to Moulton [11], more than 85% of global installed SCR systems have the hot-
side, high-dust configuration. Some of these are on lignite-firing plants that can burn a fuel
with an ash content of 30% or more. For example, China has had a positive experience with
SCR operation in coal-fired plants, with an ash loading in the flue gas as high as 70
g/m3 [12]. SCR units in this configuration have also been successfully applied in cement
kilns in Europe since 2001 and are reported to achieve control efficiencies of approximately
80% with an inlet dust loading up to 100 g/m 3 [13]. Thus, although Indian utilities lack
experience with these systems, it is reasonable to expect that, with some modification,
proper dust-cleaning systems and careful management, such SCR units could be applied to
Indian high-ash coal plants. This view is confirmed by recent, successful pilot tests of SCR
catalysts in BHEL’s facility, which used Indian coals with an ash content of ~40%, of which
65% is silica [10].

In India, around 120 GW of existing capacity (279 units) and 72 GW under construction (73
units) may require SCR, according to Kumar [14].
 Ash loading in the flue gas of Indian plants can be as high as 80 g/m 3. Hence if SCR is
placed in the hot-side, high-dust configuration, the highly erosive ash components such as
silica and alumina will have an impact as they contribute to excessive wear on the ductwork,
large particle ash (LPA) screens, ammonia injection grid (AIG) nozzles, flow distribution
devices and the SCR catalyst. The ash components also lead to poor distribution of velocity
into the catalyst, accelerate its deactivation and increase catalyst management costs. The
more ash there is in the flue gas, the greater the effect on the SCR system. Various
measures are available to alleviate the detrimental effects of highly erosive coal, including
the use of abrasion-resistant coatings, erosion-resistant wear plate, and wear shields on
AIG lances. Proper reactor sizing and catalyst module shape and pitch will also minimize
these effects [11, 15].

As noted by Moulton [11], one drawback of burning high-ash coal can be the formation of
LPA of 5–10 mm or more in the upper convective heat exchanger surfaces of the boiler.
These particles are conveyed in a high-velocity flue gas to the SCR catalyst, resulting in its
erosion and consequently a decreased rate of NOx removal. The damage can be avoided
by installing LPA screens upstream of the AIG to capture the particles, which can then be
removed in the economizer hopper. Appropriate materials such as abrasion-resistant plates
should be used to increase the inherent screen strength and minimize the effects of erosion.
Additionally, wear-resistant coatings, such as those made with chromium oxide and
tungsten carbides, should also be used to extend the screen life. Alternatively, an extra
screen can be provided if a wear-resistant coating seems to be uneconomical. In such
cases the extra screen should be treated as a consumable with a specified replacement
schedule [11].

The AIG nozzles are an easy target for erosive components of the flue gas. According to
Moulton [11], their life can be prolonged by

 designing the lances with an additional 0.32 cm (1/8 inch) of materials to account for
anticipated wear;
 using a more wear-resistant material such as 304 stainless steel; and
 installing sections of pipe 0.3 m (~1 foot) deep into the duct around each AIG header
pipe, as such shields can neutralize the effects of highly localized velocities near the duct
wall where the AIG enters into the flue gas.

Flow distribution devices such as turning vanes, static mixing elements and distribution
plates are also subject to erosion. The role of the plates is to ensure optimal ammonia
distribution and uniform flue gas velocity before it enters the catalyst bed. Hence they need
to be of a durable design, which includes using erosion-resistant protective coatings and/or
an additional thickness of 0.16 mm (1/16 inch) [11].

The size of the SCR unit for each application needs to be determined carefully and the
erosive nature of the flue gas components must be considered. This is achieved by
computational fluid dynamic (CFD) modelling. In contrast to low-dust applications, the
reactor should be designed with a larger cross-sectional area so the catalyst faces velocities
on the lower end of the design range at 2.4–3.0 m/s. This should ensure sufficient contact
between the flue gas and the catalyst for effective NOx reduction reaction. For a low-dust
environment the typical velocity through the catalyst is in the range of 3.7–5.5 m/s.

CFD modelling is also used to determine the distribution of velocity to minimize pressure
drop and particulate dropout. Identifying localized high velocities allows prediction of high-
wear sites within the SCR unit and the ductwork. Based on this, additional erosion protection
can be added to the most affected locations. Furthermore, CFD modelling can help assess
potential operational issues, including those that may occur due to the physical gaps
between the catalyst modules. High-velocity zones can occur in such gaps and lead to
accelerated deactivation of the catalyst and increased maintenance costs [11]. Some
companies have already developed technologies to mitigate these issues. For example,
Mitsubishi Hitachi Power Systems (MHPS) have patented a technology to minimize the gas
angle and velocity maldistribution using a guide vane and rectifier. This is believed to be
appropriate for high-ash coal applications, including in the Indian market [12].

Three main types of SCR catalyst are in use: corrugated, honeycomb and plate. The last
two are the most commonly used in coal-fired power plants. Selection of the appropriate
catalyst type is important for effective SCR operation. According to Schirmer [16], there is a
‘common perception’ that the plate catalyst experiences less plugging than the honeycomb
type, but this is not the case. Both types of catalyst can be used in high-ash environments if
they have the correct size of pitch. In general, a larger catalyst pitch is less susceptible to
plugging than a small one. However, as shown in tests and commercial operation of
CORMETECH’s SCR catalyst, there is a cut-off in the optimum size of the pitch, as the
inherent performance of a very large pitch (>11 mm) is a compromise between no catalyst
clogging and a higher rate of replacement, higher local ammonia slip and lower degree of
NOx reduction. Appropriate particle and dust removal systems are necessary so that fly ash
and LPA are properly managed and do not accumulate. Otherwise, ash plugging of the
catalyst will occur and increase the pressure drop across the SCR catalyst and
consequently diminish its performance. LPA can be stopped by screens and baffles and
then removed in the hopper. Fly ash can be removed by soot blowers and sonic horns and
moved to the hopper [15]. Dust removal systems are generally installed between each
catalyst layer and operate on a periodic basis, such as once a week [13]. However, for high-
ash situations, frequent or continuous cleaning of the catalyst may be required. As noted by
Schirmer [16], maintaining the frequency of cleaning is critical to avoid accumulation of the
fly ash. Coating the catalyst edge with a reinforcement material is another important
measure in high-ash situations as it prevents erosion by the ash and extends the life of the
catalyst [15].

Optimization of the catalyst pore structure is one of the key factors to prevent deactivation
by plugging and masking [15]. This can be achieved by preparing a catalyst with a trimodal
pore structure, which contains pores of three different sizes: micro, meso and macro
[17, 18]. A trimodal porous structure provides a high internal surface area and more active
sites. The micropores provide the required high specific surface area; the mesopores deliver
resistance to arsenic and other poisons such as CaO, as they can accommodate large
amounts of these substances; and the macropores help to enhance gas-phase diffusion of
the reactants, NOx and ammonia, into the catalyst, and thus assist overall catalyst activity
[17].

Minimizing the angle of the flue gas entering the SCR catalyst and minimizing velocity
maldistribution at the catalyst inlet will prevent erosion of the catalyst [12] by the use of flow
distribution devices such as turning vanes, static mixing elements and distribution plates.
However, such devices can be the subject of erosion themselves, so they must either be
made of erosion-resistant materials or coated with them.

It is possible to design a catalyst fit for the Indian market, when these factors are taken into
consideration. For example, BHEL have developed, in house, the Ti-V-W catalyst, which
has been tested at a pilot scale at its Fuels Evaluation Test Facility. The honeycomb-shaped
catalyst was tested and compared with an imported system. The tests were performed at
different temperatures (290–400°C) and velocity ranges as well as with gaseous and liquid
ammonia at 25% concentration. The average characteristics of the Indian coal used for the
test are shown in Table 2.

Table 2

Characteristics of Indian coal tested in BHEL SCR pilot facility [10]

Proximate analysis, %  Ultimate analysis, %  Ash composition, % 

Moisture, 3.6  Carbon, 45.64  SiO2, 65.6 

Volatile matter, 25.7  Hydrogen, 2.98  Al2O3, 23.6 

Fixed carbon, 33.13  Sulphur, 0.61  Fe2O3, 5.2 

Ash, 37.13  Nitrogen, 1.01  TiO2, 1.7 

Calorific value, 4385 kcal/g  Oxygen, 4.91  CaO, 0.8 

  MgO, 0.4 
 Proximate analysis, %  Ultimate analysis, %  Ash composition, % 

Na2O, 0.2 

K2O, 1.7 

SO3, 0.4 
Open in new tab

During the initial 100-hour test using coal with 37% ash content, NOx reduction rates of 80–
95% were achieved, no plugging of catalyst cells by fly-ash particles was found, and
ammonia slip was kept below 5 ppm [12].

2.5 SCR in low-dust configuration

Placing the SCR in a low-dust position, downstream of PM control, will result in essentially a
particle-free flue gas stream. This will eliminate the requirement for LPA screens, extend the
life of the catalyst and reduce the maintenance costs associated with abrasion and erosion.
On the other hand, it will require heating of the flue gas to keep the catalyst at the right
temperature for the complete NOx reduction reaction, and to avoid formation of ammonia.

2.6 SCR in tail-end configuration

Another option for high-ash coals would be to place an SCR system in the tail-end position,
according to Nakamura from JGC Catalysts & Chemicals, Japan [15]. Such systems have
been used since the late 1980s and applied in retrofits of many coal-fired power plants in
Europe, as they are often easier and less complex to retrofit than other SCR-type
configurations. Thus, their installation causes less disruption to power plant operation. Also,
in some cases, especially where there are space constraints, the capital cost for retrofitting
tail-end SCR can be lower than for high-dust SCR. Furthermore, tail-end systems are
available in modular forms designed to be installed with minimal disruption to plant
operation. However, as mentioned earlier, the flue gas reheat is needed in this configuration
for effective NOx reduction reaction [13]. In Nakamura’s opinion [15], one solution for Indian
coals would be to place an SCR after a dry FGD that uses an SOx adsorbent made of fly
 ash, calcium hydroxide and the used SOx adsorbent as adhesive; as such the FGD can
remove more than 85% of the dust. This, in combination with an ESP, would mean a low-
dust environment for an SCR and would avoid problems related to a high-ash environment,
particularly erosion, plugging and clogging of the catalyst.

2.7 Role of SCR catalyst in mercury capture

Of increasing importance, an SCR catalyst can be a key component for mercury oxidation,
facilitating its removal in downstream pollution control systems [19]. Oxidized mercury
species, such as HgCl2, are typically water soluble and can be removed in wet FGD
systems. Oxidized mercury can also be bound to particulate matter and so removed in PM
control systems. In contrast, elemental mercury vapour passes through sulphur and
particulate controls uncaptured. A number of factors affect mercury oxidation in the SCR
catalyst, including

 fuel composition, particularly the levels of mercury and halogen present;

 supplementary halogens, that is the addition of bromine and chlorine;
 catalyst design, including catalyst pitch, the number of layers and their configuration,
catalyst volume and composition;
 catalyst age;
 temperature, as a lower temperature favours oxidation of mercury;
 flow rate; and
 presence of ammonia [19].

Several manufacturers offer SCR reactors with a catalyst that can oxidize mercury. For
example, Mitsubishi Hitachi offers an SCR system, with TRAC (TRiple Action Catalyst).
Such a system is proposed for the Turów 11 lignite power plant currently under construction
in Poland. According to Modern Power Systems [20], this will be the world’s first lignite
power plant to meet the new EU Best Available Technology Reference Document (BREF)
standards. For a new lignite-fired power plant, it means that daily average emission levels of
NOx must be 80–125 mg/m3 [21]. This will be achieved by using ammonium chloride
(NH4Cl) as the reagent, delivered via Delta Wing technology, and a plate-type TRAC
catalyst, which will have a 2 + 2-layer arrangement. This will result in a NOx level at the
SCR outlet of 85 mg/m3, high mercury oxidation and low SO2 to SO3 conversion [21].
Mitsubishi’s SCR has already been installed on various lignite plants in Austria, China,
Germany, Thailand and the USA, so it is reasonable to expect that the Turów 11 power
plant will achieve the NOx reduction predicted. Turów 11 will burn locally mined lignite with
an ash content of ~30%. Other catalyst suppliers which offer products that enhance mercury
oxidation include Johnson Matthey, CORMETECH and Haldor Topsøe.

3 SNCR

On its own, SNCR reduces NOx by 30–50%, whereas SCR levels of performance can be
achieved if it is applied in conjunction with other combustion controls. Historically, the
effectiveness of SNCR has been limited, especially in utility-scale boilers, due to a lack of
accurate real-time tools to measure the temperature and CO profile within the boiler—
parameters that are important for an effective NOx reduction reaction. However, recent
developments in measurement systems allow the effective use of SNCR even on large
furnaces (>400 MW).

Although SNCR technology has not yet been used with Indian coals, it has been proven with
other high-ash coals, such as coal-fired plants burning lignite and bituminous coal with an
ash content of ~30%, as well as on circulating fluidized bed boilers and cement kilns. In the
latter, SNCR has been reported to achieve a reduction in NOx of 12–77% for systems using
ammonia and 25–90% for those using urea as the reagent [13]. However, different rates of
NOx reduction are reported for coal-fired plants: up to 50% for an SNCR unit on its own and
up to 90% for an SNCR unit combined with primary measures [9]. SNCR manufacturers with
long experience of high-ash coal environments include Fuel Tech Inc., which systems have
wall injectors and multi-nozzle lances, and Mobotec, which developed ROTAMIX.

Based on this experience, various experts and equipment manufacturers are confident that
SNCR can be successfully used in Indian coal-fired plants, especially for the units that must
reduce their NOx emissions below 300 or 100 mg/m3 (J.M. Boyle, Fuel Tech Inc., USA,
personal communication, 2017; P. de Havilland, Fuel Tech Srl, Italy, personal
communication, 2017; R. Żmuda, SBB Energy S.A., Poland, personal communication,
2017).
SNCR applied in combination with primary NOx measures such as LNBs and OFA will meet
the 300 mg/m3 limits and is an alternative to SCR. However, for new plants that have 100
mg/m3 NOx emission limits, an SNCR in combination with a small SCR is an option. In
Boyle’s opinion (Fuel Tech Inc., USA, personal communication, 2017), such a combination
applied to Indian boilers would reduce the cost of an SCR catalyst as fewer layers would be
required. A smaller SCR unit with fewer layers of catalyst in the duct or standalone has
many benefits. First, at low load, the temperature of the flue gas at the SCR reactor may be
too low for its effective operation. In this case, an upstream SNCR system can still provide
some level of NOx control. Second, catalyst poisons may force replacement of catalyst
modules, which is expensive. Fewer layers make the risk of catalyst failure less expensive.
Third, each added catalyst layer causes more oxidation of SO 2 to SO3 and requires greater
fan capacity to overcome the pressure drop in the SCR reactor. Fewer layers of SCR
catalyst would also make ash removal easier and would consequently prolong the life of the
catalyst (J.M. Boyle, Fuel Tech Inc., USA, personal communication, 2017).

Both ammonia and urea can be used as a reagent in high-ash coal plants. However, as
noted by P. de Havilland (Fuel Tech Srl, Italy, personal communication, 2017), using SNCR
with urea would be more beneficial in Indian applications. This is because urea is non-toxic,
unlike ammonia, and its use would avoid significant environmental and health hazards.
Additionally, as noted by Boyle (Fuel Tech Inc., USA, personal communication, 2017), urea
is much more effective than ammonia for SNCR NOx reduction on large furnaces (>400
MW), because aqueous urea droplets can be propelled into the furnace with no reaction
until the water is evaporated. This then releases the urea salt and reaction begins with the
disassociation of urea into ammonia (NH3) and isocyanic acid (HNCO), both of which then
react to reduce NOx. Although some competitors have suggested that HNCO must first
convert to ammonia, in Boyle’s opinion this is not the case and the HNCO reaction path
does not include ammonia, so the likelihood of residual unreacted ammonia, or ammonia
slip, is reduced.

Boilers firing Indian coals are taller and wider than those firing imported fuel due to the
significantly higher ash content. Consequently, an SNCR system must cover a greater area
and thus not all types of spraying nozzles and/or SNCR systems may be applicable. This
can be verified using CFD modelling and field tests.

There are a number of SNCR suppliers. Fig. 3 shows a multi-nozzle lance from Fuel Tech’s
advanced SNCR. Such a system has been used on a Chinese power plant firing lignite with
an ash content of 27.5% since 2007, and no erosion or plugging has been reported.

Fig. 3

Open in new tabDownload slide

Multi-nozzle lance during test (Source of the photo: P. de Havilland, Fuel Tech Srl, Italy,
personal communication, 2017)

3.1 Multi-pollutant and emerging NOx control technologies

As noted by Sloss [22], pollution control technologies are expensive and require significant
time to install, which in turn can disrupt power generation. India already experiences power
shortages, and has to introduce multiple technologies at once, so it would make sense to
coordinate installation of pollution control systems and consider multi-pollutant control
systems.

The definition of a multi-pollutant control system can be confusing because all pollution
control systems offer the potential to reduce a few pollutants simultaneously. In recent
years, many technologies that have emerged in the market are defined commercially as
multi-pollutant control systems. Although many of them can be used to remove several
pollutants simultaneously, they appear to be sold for use as an addition to the usual control
systems, such as FGD, ESP, and SCR, with the aim to reduce emissions of specific
pollutants further [22].

A few multi-pollutant systems can remove NOx. Some of them, such as ReACT, have been
used in coal-fired power plants for several years, notably at full scale in Japan since the
1990s. Others are deployed in non-coal applications but have the potential to be applied to
coal-fired plants and are in various stages of testing and demonstration.
 Although no multi-pollutant systems have been developed specifically for high-ash coals, it
appears that the existing ones, such as ReACT and LoTOx, can be applied successfully.
For India, both cost and water availability are important considerations in the selection of
technology. Hamon Research-Cottrell’s ReACT is a regenerative activated coke dry-type
capture technology that captures SOx, NOx and mercury while only using 1% of the water
required by a conventional wet FGD. The system can be configured for different levels of
NOx control, described as

 co-benefit NOx control at 20–50%;

 enhanced NOx control of 40–60% with ammonia-dosed activated carbon
regeneration; and
 higher NOx control to 80% available with two-stage adsorption.

The ReACT technology is placed downstream of the ESP; hence the system is not subject
to fly-ash loading and is suitable for units firing high-ash coals, according to J. Peters
(Hamon Research-Cottrell, New Jersey, personal communication, 2017). For such
applications, it can be used in a single-stage configuration, which allows removal of ~50% of
NOx. Alternatively, depending on economic and performance factors, it can be used in a
two-stage configuration that enables control NOx control of up to 80%. Typically, the inlet
NOx concentration will be around 550 mg/m3 or lower for ReACT applications. This level of
concentration of NOx is achieved easily in coal-fired units with low NOx burners.

The ReACT system can be installed on new and retrofit projects, and appears to be
especially advantageous for units located where water use is an issue. In retrofit
applications, a single-stage ReACT can provide a feasible solution in addition to low NOx
burners and/or overfire air systems. A two-stage ReACT application would also be an option
for new power plants using Indian coal as an alternative to SCR technology, again if
installed in addition to LNBs and other primary measures. The system has other advantages
relevant to the Indian market: it minimizes waste; revenue can be derived from the by-
products of fly ash and sulphuric acid; and the system is operational through start-up and
shutdown. As noted by Peters (Hamon Research-Cottrell, New Jersey, personal
communication, 2017), the system can be considered as a ‘retrofit’ technology that
minimizes station outage requirements for construction, because the ReACT equipment is
located downstream of the plant’s existing equipment and can be tied back into the existing
stack. The potential of the ReACT system for Indian high-ash coals has been already
recognized by the Indian government and utilities [8].

Another example of multi-pollutant capture technology is Linde Gas’s patented LoTOx

process. It uses low-temperature ozone injection to oxidize NOx to highly soluble dinitrogen
pentoxide (N2O5) for capture in an FGD system, where the flue gas moisture and the liquid
convert it to nitic acid (HNO3) [23]. It can also be used with a dry FGD system (B. Stapper,
AECOM, Texas, personal communication, 2017). This technology has been applied in the
refining sector on more than 30 units, installed on a 25-MW coal-fired institutional boiler and
tested as part of a pilot demonstration by the Electric Power Research Institute on a 550-
MW coal-fired plant and it can provide 90% NOx reduction [19, 24]. In LoTOx applications,
ozone is produced onsite by passing oxygen through a conventional ozone generator, so no
ozone storage is required. The ozone is generated ‘on demand’, depending on the
NOx level present in the flue gas and the required removal efficiency. Another advantage of
the system is its ability to achieve mercury oxidation rates of 50–70%; moreover, SO 2 is not
converted to SO3 [23].

Many existing Indian plants must install FGD units to comply with the new SO 2 standards of
200 mg/m3. Units larger than 500 MW with the space for such systems have already been
identified. It has been estimated that pre-installation activities should take about three years,
until around 2020. During this period, it is recommended that other pollution control
measures such as LNBs and OFA can be installed in the next planned outage [8, 24]. For
plants where a wet FGD system is selected, following installation of primary NOx measures
such as LNBs or OFA, it may be worth considering installation of the LoTOx system as an
alternative to an SNCR or SCR. This is because the LoTOx system will allow removal of
NOx and oxidized mercury in a wet scrubber and should not be affected by the high ash
content. It will also remove the need for storage and handling of an ammonia reagent and
consequently should save both space and water, as well as potentially bringing the cost of
NOx control down.
Other multi-pollutant controls and emerging systems may have potential in Indian plants too,
especially those that minimize water usage and cost.

4 Conclusion

NOx controls, both primary and secondary measures, are well-proven mature technologies,
in use for more than 40 years. However, in India, which until recently did not have
NOx emission standards and where the majority of plants fire high-ash coals, secondary
controls (SCR and SNCR) have not yet been commercially deployed and utilities are
concerned about their suitability for Indian coals. Nevertheless, vendors have confidence
that they can supply solutions for Indian boilers as both systems have been used
successfully in high-ash lignite plants and in cement kilns where the ash loading in flue gas
is as high as 100 g/m3. Additionally, initial tests of BHEL’s SCR catalyst also had promising
results. Still, demonstrations are required to establish how to customize these technologies
for the Indian market, so they can be used cost-effectively. Such demonstrations are already
underway on eight NTPC boilers, which are expected to pave the way for the installation of
these technologies in India. For example, for SCR, the tests will show which type of catalyst
works best, its optimal pitch size, how many layers are required, what cleaning system to
choose and how often it should be used. In the case of SNCR, the tests will determine which
type of spaying nozzles is able to cover Indian boilers and which reagent (ammonia or urea)
will work best. Nevertheless, to know the exact effect of the Indian coal on these systems
may take years.

As India is introducing emission standards for more pollutants, and pollution controls are
expensive and time-consuming to install, which disrupts power generation, it would make
sense to coordinate installation of pollution control systems and to focus on multi-pollutant
control systems. There are a few multi-pollutant controls that can remove NOx and are not
affected by high ash content in the flue gas. Examples include ReACT, which has been
used in coal-fired power plants for several years, and LoTOx, which is deployed mainly in
the oil-refining sector but has the potential to be applied to coal-fired plants.

To summarize, NOx controls for high-ash coals are broadly the same as for ‘normal’ boilers,
but they must be customized to local market requirements (high ash content) as already
happened with LNBs. Choosing appropriate NOx control systems for any power plant
requires a site-specific strategy that, apart from cost, performance and safety, should also
consider water availability.