South African Journal of Chemical Engineering 48 (2024) 167–183

Contents lists available at ScienceDirect

South African Journal of Chemical Engineering

journal homepage: www.elsevier.com/locate/sajce

Sodium-based flue gas desulphurisation for the South African coal-fired

power industry a review
Dursman Mchabe a, Burgert B. Hattingh b, Lawrence Koech b, c, *, Hilary Rutto b, c,
Hein W.J.P. Neomagus b
a
School of Process Engineering, Faculty of Engineering, Stellenbosch University, Private Bag X1, Stellenbosch 7602, South Africa
b
Eskom Power Plant Engineering Institute (EPPEI) Specialisation Centre for Emissions Control, Centre of Excellence for Carbon-based Fuels & School of Chemical and
Minerals Engineering, Faculty of Engineering, North-West University, Private Bag X6001, Potchefstroom 2520, South Africa
c
Department of Chemical Engineering, Vaal University of Technology, Vanderbijlpark Campus, Private Bag X021, Vanderbijlpark 1900, South Africa

A R T I C L E I N F O A B S T R A C T

Keywords: A comprehensive review of sodium-based flue gas desulphurisation is presented in this article. The paper pro­
Flue gas desulphurization vides a comparison of various sodium-based technologies with regards to sulphur removal efficiency, sorbent
Coal-fired power plants cost, by-product valorisation and overall process economics. More emphasis was placed on dry sorbent injection
Sodium-based absorbents
(furnace sorbent injection, economizer sorbent injection, duct sorbent injection) and semi-dry flue gas desul­
SO2 emissions abatement
phurisation. Particular effort was devoted in identifying state-of-the-art technologies and potential research gaps
in a summarized way. In water-scarce countries, where sodium-based technologies are more economically
feasible, and duct sorbent injection processes are favourable, it is recommended to use sodium-based absorbents
since they will have higher sorbent utilization and improved electrical resistivity (compared to their calcium-
based and/or magnesium-based sorbent equivalents). However, the check-and-balance of the associated ad­
vantages and disadvantages are a pivotal determinant of the sorbent choice. Apart from sulphur dioxide
abatement, the additional advantages of sodium-based absorbents include its high sulphur trioxide removal
efficiency as well as its hydrogen fluoride and hydrogen chloride removal capability. A scarcity in literature,
especially as it relates to application in the South African coal-fired power generation industry, necessitates the
need for further research with particular focus on local sodium-based absorbent markets, the physiochemical
properties and sulphur dioxide reduction potential of locally acquired sodium-based absorbents, modelling of
sulphation reaction kinetics in semi-dry and dry application, as well as the techno-economic feasibility of sorbent
based technologies for the South African power industry.

1. Introduction to be the main contributors to global SO2 emissions mainly as a result of

fossil fuel-based electricity generation. More specifically in 2019, South
In 2019 global anthropogenic activities alone contributed close to 69 Africa ranked 7th with a total anthropogenic SO2 emission rate of close
% of the total SO2 emissions, with coal-fired power generation ac­ to 1200 kt per annum of which 95 % of this value could be related to
counting close to 36 % of the global value (Dahiya et al., 2020). Coun­ coal-fired power generation (Dahiya et al., 2020). A significant reduc­
tries such as the United States of America (Srivastava and Jozewicz, tion (between 37 and 40 %) in global SO2 emissions has, however, been
2001), India (Garg et al., 2002), China (Yan and Wu, 2017), Europe observed over a ten-year period as measured between 2008 and 2009.
(Smith et al., 2011) and South Africa (Pretorius et al., 2015) have shown This marked improvement could be largely attributed to the

List of abbreviations: ACP, aqueous carbonate process; ANN, artificial neural networks; BET, brunauer, Emmett and teller; CAPEX, capital expenditure; CDS,
circulating dry scrubbers; CFB, circulating fluidized bed; CFD, computational fluid dynamics; DFFE, department of forestry, Fisheries & environment; DSI, dry / duct
sorbent injection; EC SC, emissions control specialisation centre; EPA, environmental protection agency; EPPEI, eskom power plant engineering institute; EPRI,
electric power research institute; ESI, economizer sorbent injection; ESP, electrostatic precipitator; FGD, flue gas desulphurisation; FFP, fabric filter plant; FSI,
furnace sorbent injection; IEA, international energy agency; MES, minimum Emission Standards.
* Corresponding author.
E-mail address: [email protected] (L. Koech).

https://doi.org/10.1016/j.sajce.2024.01.016

Available online 5 February 2024

1026-9185/© 2024 Published by Elsevier B.V. on behalf of South African Institution of Chemical Engineers. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

strengthening of emissions standards and legislation, a reduction in coal techno-economics, co-emission reduction benefits and regional / local
usage as well as an increased application of SO2 emissions control availability of both natural occurring or alternative absorbents (e.g.,
technologies in some of the top emitting countries (Dahiya et al., 2020). sodium-based absorbents) used in semi-dry and dry application but also
The promulgation of more stringent SOx regulations by various regula­ the valorisation potential of the subsequent FGD by-products. The use of
tory bodies across the world has also subsequently prompted the South sodium carbonate (Na2CO3) as a sorbent for acid gas abatement dates
African government to further amend its Minimum Emission Standards back to the 1980s (Kimura and Smith, 1987; Mocek et al., 1983). In the
(MES), in terms of Section 21 of the National Environmental Manage­ 1990s, the approach was widely used in Waste to Energy (WtE) plants
ment: Air Quality Act, 2004 (Act No. 39 of 2004) (NEMAQA), in 2018 (Vehlow, 2015), following the NEUTREC process (marketed by Solvay S.
and 2019 (Department of Forestry, Fisheries & Environment (DFFE), A.) through the injection of NaHCO3 in the flue gas and its subsequent
2004 & 2019; Gubb, 2009; Ross, 2012). solid waste collection (be it by fabric filter or electrostatic precipitator).
For coal-fired power generation industries several pre- and post- Literature review publications that focus more in depth on sodium-based
combustion abatement technologies are currently available to control FGD systems are, however, relatively scarce when compared to similar
atmospheric releases of SO2. Of these available technologies flue gas works based on calcium-based technologies. The objective of this paper
desulphurization (FGD), as a post-combustion abatement solution is was therefore to systematically review the available literature on
mostly favoured due to its technical and economic feasibility as opposed sodium-based FGD as a means of evaluating recent developments as well
to pre-combustion technologies. It is projected that the global market as further research opportunities to contribute to the body of knowledge.
size for FGD will increase by as much as $ 7.2 billion over the next 6 Knowledge gained through this approach is also intended to better
years, growing at a CAGR (Compound Annual Growth Rate) of 4.3 %, as inform not only future technical feasibility studies but also local energy
the demand to install FGD systems or retrofit existing systems increases and environmental policies as well as industry-specific emissions control
to meet the stringent legislative requirements (Bloomberg, 2022). This is technology roadmaps.
especially of significance for developing countries, such as South Africa,
where fossil fuels remain the primary source of energy for the unfore­ 2. Review methodology and approach
seeable future due to the relatively slow implementation of renewable
energy generation alternatives (Koech et al., 2021). Presently, there is A systematic, in-depth review of available literature (past 40 years)
broad range of available commercial regenerative and non-regenerative was performed to compare several technical, operational and econom­
FGD processes in wet, semi-dry and dry application (Carpenter, 2012; ical aspects related to sodium-based FGD. Databases from ScienceDirect,
Lisnic and Jinga, 2018; Srivastava and Jozewicz, 2001; Vernon and Scopus, the Electric Power Research Institute (EPRI), the International
Jones, 1993). Wet FGD (WFGD) is by far the most common technology, Energy Agency (IEA) and the Environmental Protection Agency (EPA)
accounting over 80 % of the total installed capacity worldwide, while constituted the primary resources for this systematic review, whilst news
semi-dry and dry processes are reported to only account for less than 10 articles, OEM (original equipment manufacturers) brochures, govern­
% (Carpenter, 2012; Cordoba, 2015; Srinivasn, 2004). In addition, lime mental policy documents, engineering contractor’s case studies and
(Ca(OH)2 or CaO) and limestone (CaCO3) are mostly favoured as suit­ business cases were also occasionally consulted. The paper commences
able absorbents for WFGD due to its abundance and availability.(Car­ with a brief overview of commonly known sodium-based FGD processes,
penter, 2012). However, growing concerns with respect to the high with specific emphasis placed on direct sorbent injection (DSI) and its
process water consumption rates for WFGD as well as the effective associated technical and operational aspects affecting absorbent effi­
management of FGD by-products has sparked renewed interest in dry ciency and downstream processing. The review then proceeds further by
and semi-dry technologies, especially in water-scarce or semi-arid re­ reflecting on recent developments with respect to non-conventional
gions, to minimize operational expenditure and utility demands (Koech sodium-based absorbents as well as recent advances for elucidating
et al., 2021). and improving sodium-based absorbent performance. Further consid­
Although there is a strong drive for South Africa’s energy sector to erations with respect to sodium-based DSI are then summarized, fol­
transition to a more sustainable and renewable future, it is foreseen that lowed by a critical assessment of the research carried out to date. The
coal-fired power generation will remain a key component of the energy latter was aimed at identifying potential research gaps to stimulate
mix for at least the short to medium term (Strickroth et al., 2020). The further research in the field of sodium-based FGD and its applicability
efficient control of emissions (within regulated limits and standards,) and or feasibility within the South African coal-fired power generation
during pulverized coal combustion will therefore become ever increas­ sector. In comparison to other publications in the field of FGD, the re­
ingly more important as the country battles to address the persistent view presented in this paper follows a holistic approach by not only
threat to energy security, whilst also ensuring environmental compli­ focusing on absorbent characteristics and performance but also taking
ance. To this extent South Africa’s major state-owned utility company, into consideration other facets of technology selection, such as the
Eskom, implemented six wet flue gas desulphurization (WFGD) units valorisation potential of produced FGD by-products.
(the first in South Africa and Africa) as part of its new 4800 MWe Kusile
power station in the Mpumalanga region. To date 4 of the 6 units at 3. A review of sodium-based flue gas desulphurization processes
Kusile have achieved beneficial operation. Furthermore, two retrofit
technology selection studies performed in 2014 and 2018 also recom­ 3.1. General overview of flue gas desulphurization
mended the use of WFGD technology for Eskom’s operational 4800 MWe
Medupi power station in the Waterberg region (Bagus et al., 2018; Commercial flue gas desulphurization (FGD) processes are
Eskom, 2018; Harris et al., 2014). Both studies considered limestone and commonly based on the contact and/or reaction of an acidic gas (e.g.,
lime (quick and hydrated) as absorbents, while only spray dry absorbers SO2, SO3, HF, HCl etc.) with an alkaline sorbent (e.g., calcium-, sodium-
/ scrubbers (SDA/SDS) and circulating fluidized bed (CFB) absorbers or ammonium-based) to produce sulphates and/or sulphites and by-
were used in the comparison with WFGD technologies. Although the products. Depending on the fate of the sorbent, FGD can be broadly
focus was more on calcium-based sorbents, a recent feasibility study categorized into two main categories, namely, regenerative and non-
conducted in 2019 further concluded that direct sorbent injection (DSI) regenerative processes. For regenerative processes, the spent sorbent is
was not technically feasible for the Medupi power station due to an recovered via thermal or chemical treatment processes while concen­
uncertainty in sorbent quality and availability, the configuration of the trated SO2 is also generated and further processed to desired products e.
existing dust handling facilities as well as concerns related to waste g., elemental sulphur, H2SO4, etc. In contrast, for non-regenerative
management (Eskom, 2019). The study could, however, have benefitted processes, the absorbents are not recycled (Lisnic and Jinga, 2018; Sri­
from a more detailed account of not only the performance, vastava and Jozewicz, 2001). Classification of FGD processes depends

168
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

on the aggregation state of the sorbent, namely, wet processes (solution treatment’ which produces a calcium-sulphite-rich sludge (Satriana,
or suspension) which consume the largest amount of water followed by 1981). In 1998 the global installed capacity equipped with wet
semi-dry processes (sorbent with controlled humidity) and dry processes sodium-based (i.e., Na2CO3) FGD was mainly located in the United
(zero humidification) (Mchabe, 2020). Generally, regenerative pro­ States and amounted to 2756 MWe, which represents only a mere 1.4 %
cesses cut across the spectrum of wet, semi-dry and dry flue gas desul­ of the global total of installed wet FGD capacity (Srivastava, 2000). More
phurization processes. Limestone (CaCO3) is the most widely used recently, another 11 sodium-based wet FGD units, associated with a
calcium-based sorbent in FGD processes (Carpenter, 2012; Cordoba, coal-fired power generation capacity of 6322 MWe, were installed at US
2015; Srivastava and Jozewicz, 2001), owing to its availability and power plants (US EPA, 2019).
relatively low cost. Other calcium-based absorbents include quicklime
(CaO) and slaked lime (Ca(OH)2). It is important to bear in mind that 3.2.2. Dual-alkali scrubbers
both quicklime and slacked lime can be derived from limestone through The first dual-alkali units in the United States were brought online in
calcination (quicklime), and calcination and quenching (slaked lime). It 1974 (Donahue et al., 1980; VanNess et al., 1979). The process is based
has been reported by various authors (Carpenter, 2012; Maina, 2012; on two sets of chemical reactions. First, sodium-based alkali solution (e.
Pandey et al., 2005; Poullikkas, 2015; Siagi et al., 2007) that quicklime g., NaOH or Na2CO3) reacts with the SO2 in the flue gas. The second set
is more reactive than limestone. The reactivity was attributed to various of chemical reactions involve the spent alkali solution and lime, to
factors namely: higher surface area, porosity, etc. regenerate and recover the sodium-based sorbent. The process is more
Slaked lime has been reported to have an even higher surface area suitable on particulate-free systems but can be employed for flue gas
and porosity than both quicklime and limestone (Maina, 2012; Sasaoka treatment (high concentration and low concentration) on both coal-fired
et al., 1998), and as a result is generally more reactive. Other common and oil-fired plants (Donahue et al., 1980). The dual-alkali process in­
absorbents used include sodium-based compounds (sodium carbonate, corporates three main unit operations namely: absorption, chemical
sodium bicarbonate, nahcolite and trona), magnesium carbonate and regeneration, and solids dewatering. During absorption, flue gas from
ammonia (Carpenter, 2012; Muzio and Arand, 1981; Pflughoeft-Hassett the facility’s dust collection plant (electrostatic precipitator (ESP) or
et al., 2009; Pilat and Wilder, 2007). These absorbents also react with baghouse / fabric filter plant (FFP)) is introduced at the bottom of a
SO2 and SO3 in the flue gas to produce a mixture of sulphite and sulphate scrubber, and proceeds to the reaction zone, where the gas is thoroughly
salts, and in some instances also produce other sulphate and bisulphate mixed with the alkali solution (fresh makeup NaOH/Na2CO3 solution
compounds. The proportions of these products are dependent on the and the recycled Na2SO3 solution). As the flue gas stream reaches its
process conditions (Carpenter, 2012; Kong and Wood, 2011; Pflugh­ adiabatic saturation temperature, SO2 is removed through the subse­
oeft-Hassett et al., 2009; Pilat and Wilder, 2007). The commonly used, quent production of sodium bisulphite (NaHSO3), sodium sulphite
naturally occurring sodium-based absorbents include trona, nahcolite (Na2SO3) and sodium sulphate (Na2SO4). In regenerative mode the spent
and sodium bicarbonate (NaHCO3) (Muzio and Arand, 1981; Sahu, scrubber solution is reacted with lime to regenerate the sodium-based
2013). These sodium-based absorbents have been reported to be more sorbents and produce calcium sulphite and sulphate (CaSO3 & CaSO4)
reactive than hydrated lime during dry injection (Carpenter, 2012). Both (Donahue et al., 1980). Hereafter the thickened slurry is dewatered, and
sodium bicarbonate and trona have also been found to be more reactive the filtered liquor recycled back to the system. High SO2 removal effi­
to SO2 than sodium carbonate (Na2CO3) due to a higher surface area for ciencies in the order of 85 to 95 % have been reported by various
reaction as a result of a phenomenon known as the “popcorn” effect (the workers (Córdoba, 2015; Donahue et al., 1980; Poullikkas, 2015; Sri­
generation of micropores during water and CO2 generation and subse­ vastava and Jozewicz, 2001). The advantages of the dual-alkali process
quent release from the sorbent particles). include, amongst other factors: (1) a high availability as the process is
relatively free from plugging and scaling, (2) a low energy requirement
3.2. Wet sodium-based FGD processes (2 to 3 % of the total plant energy input), (3) a relatively pure gypsum
by-product, as well as (4) minimal corrosion and erosion problems.
3.2.1. Soda ash wet scrubber However, drawbacks of this process include: (1) process complexity, (2)
A variety of different wet scrubber technologies are currently avail­ high operational cost due to reagents, (3) the formation of sodium salts
able for FGD and include spray towers, venturi scrubbers, packed towers on the filter cake, (4) particulate emissions due to misting, (5)
as well as trayed towers. Spray towers are commonly more preferred for complexity in process retrofit and additional capital expenditure due to
large-scale FGD application given its better slurry handling capability the requirement of extra auxiliary equipment as well as (6) blockages in
(US EPA, 2003). An example of a sodium-based, non-regenerative wet areas where low temperatures are common.
FGD process is the soda ash scrubber, of which an illustrative schematic
can be found in the works of Pandey et al. (2005). A soda ash scrubber 3.2.3. Wellman-Lord’s sodium sulphite process
operates on similar principles to that of a wet limestone scrubber i.e., Over the past two decades publications on the Wellman-Lord process
SO2 and SO3 species in the flue gas dissolves in water to produce focused primarily on gas purification from harmful components
sulphurous and sulphuric acids that consequently reacts with the dis­ (Dzhonova-Atanasova et al., 2013; Yildirim et al., 2012; Srivastava and
solved alkaline sorbent to produce sulphite and sulphate salts. The Jozewicz, 2001) and has been reported to be applied in about 40
produced salts precipitate out of solution, depending on their respective operating installations in Japan, USA, and Germany. The main advan­
saturation points and relative solubilities (Pandey et al., 2005). When tage of the process is the relatively low sorbent demand and low waste
both reactants and products are soluble, the need for technologies to production (Yildirim et al., 2012). In this process, either sodium sulphite
handle solid waste products is eliminated (Schnelle et al., 2015). For (Na2SO3) or sodium sulphate (Na2SO4) solution is employed to absorb
both soda ash and caustic soda the sulphation process entails several SO2 from the flue gas. The process constitutes the following unit oper­
dissolution, dissociation and aqueous phase reactions (more details ations: a flue gas pre-treatment step, a SO2 absorption unit as well as an
provided in the works of Pandey et al. (2005)) to produce more specif­ absorbent regeneration and sulphate removal section. Flue gas
ically sodium bisulphite (NaHSO3) and or sodium sulphite (Na2SO3) compression is facilitated via the venturi pre-scrubber where after the
(Pandey et al., 2005). It has been reported that this process is able to waste stream is neutralized (by alkaline fly ash and/or lime/limestone)
achieve SO2 removal efficiencies of between 90 and 95 % when and deposited in a precipitator as a 5 wt. % suspension. Regeneration of
sodium-based alkalis are used (US EPA, 2021). However, soda ash or Na2SO3 produces a concentrated SO2 stream, which may be converted
caustic soda scrubbing processes are costly (Pandey et al., 2005) but to into a saleable product such as liquid SO2, sulphuric acid, or elemental
offset significant operational expenditures, lime or limestone may be sulphur. It is important to bear in mind that, despite the targets of
introduced into the process, thereby implementing ’dual-alkali regenerating all the sodium sulphite and/or sodium bisulphite, minor

169
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

irreversible oxidation reactions may occur, leading to the formation of precipitation and removal (filtration) of heavy metal compounds to in­
sodium sulphate (Pandey et al., 2005). It has been reported by Pandey crease the purity of the final products. The resulting spent solution is
et al. (2005) that this process is able to achieve a SO2 removal efficiency then finally mixed with ammonium bicarbonate (NH4HCO3) to not only
of up to 95 %. Recently a review conducted by Dzhonova-Atanasova regenerate the NaHCO3 through the process of precipitation but also to
et al. (2013) provided useful insights into improving the energy effi­ produce ammonium sulphate as a high-quality fertiliser by-product
ciency of the Wellman-Lord process. From this review it is noteworthy to (Carpenter, 2013). Apart from the regeneration benefit, the Airborne
mention that the final concentration of sodium bisulphite in the process™ also has the following additional advantages: (1) lower total
absorber also determines the process efficiency and is dependent on the power consumption (3 %), (2) lower capital and operating costs than
initial quantity of sodium sulphite fed into the absorption column. separate conventional pollutant control technologies (SCR, limestone
Operational concentrations are therefore usually close to saturation. The wet scrubber, wet ESP and activated carbon injection) as well as lower
lower solubility of Na2SO3 (as compared to NaHSO3) also creates the water consumption than wet limestone scrubbers (Carpenter, 2013;
possibility of further saturating the partially unreacted solution to in­ Mortson and Telesz, 2001; Zhu 2010).
crease the NaHSO3 concentration resulting in a subsequent reduction in
the energy demand (i.e., steam consumption in this case) per unit of SO2 3.3. Semi-dry, sodium-based FGD processes
absorbed (Dzhonova-Atanasova et al., 2013). The process has the
advantage of producing a relatively pure SO2 product stream via the The semi-dry desulphurization process emanates from taking
regenerative step, while the largest drawbacks of this process include the advantage of the sensible heat of evaporation of the desulphurization
high capital investment for additional equipment required for the slurry moisture from the sorbent particles. A predetermined amount of
recirculation of the absorbent as well as the significant amount of steam water is introduced, either to form a film around solid particles or as to
required for solution regeneration (Dzhonova-Atanasova et al., 2013). introduce the sorbent as a highly concentrated slurry. During the pro­
Despite this the operational costs have, however, been reported to be a cess, SO2 and SO3 reacts with the solid sorbent directly. To ensure an
similar order of magnitude to that of wet limestone FGD (Pandey et al., efficient process, absorbents need to be characterized by a high specific
2005). surface area and porosity. Similar to the wet FGD process, the products
from semi-dry, sodium-based FGD include sodium sulphites and - sul­
3.2.4. Sodium carbonate eutectic process phates (Na2SO3 and Na2SO4). Generally, semi-dry processes can be
This process takes advantage of the eutectic temperature of sodium categorized into: (1) spray dry scrubbers, (2) duct spray dry process and
carbonate (Na2CO3) (i.e., 425◦ C). The absorption of SO2 proceeds ac­ (3) circulating dry scrubbers (Carpenter, 2012). Other semi-dry scrub­
cording to Reaction R1 (Kumar and Jana, 2021; Pandey et al., 2005). bing technologies include gas suspension absorption (GAS) and Alstom’s
Novel Integrated Desulphurization (NID™) system (Carpenter, 2012).
Na2 CO3 + SO2 →Na2 SO3 + CO2 (1)
Publications on sodium-based, semi-dry FGD processes are relatively
The regeneration of sodium carbonate is given by Reaction R2: scarce in literature with most, only briefly referring to the use of these
absorbents without providing any detailed technical information e.g.,
Na2 SO3 + CO + 2H2 →Na2 CO3 + H2 S + H2 O (2)
SO2 removal efficiencies, average water consumption rates, proven
The H2S produced in Reaction R2 can be processed to yield valuable commercial facilities or demonstrations and cost estimates. The use of
products (e.g., elemental S by Claus process). sodium-zinc sorbent as a substitute for the traditional semi-dry FGD
technologies has, however, been proposed by Zhang et al. (2015).
3.2.5. Airborne process™
One of the emerging technologies, as it relates to sodium-based FGD, 3.3.1. Spray dry scrubbers
is the Airborne process™. This emissions abatement process is an Carpenter (2012) indicated that in the spray dry scrubbing (SDS)
amalgamation of the wet sodium scrubbing process and dry sorbent processes, a concentrated lime slurry (a few systems use a sodium car­
injection of sodium bicarbonate. Mortson & Xia (2006) reported bonate solution) is introduced into the top of the absorber vessel through
removal efficiencies of 99.9 % for SO2, 99.9 % for SO3, 99 % for NOx, 99 rotary atomizers or dual fuel nozzles. These atomize the slurry creating a
% for mercury from their bench-scale and pilot scale demonstration fine mist of droplets containing the reagent, which reacts with SO2 and
tests. Carpenter (2013) reported that the proprietary NaHCO3 regener­ SO3 in the downward flowing flue gas to form calcium sulphite and
ation component has been demonstrated at a commercial sodium sul­ sulphate (CaSO3 and CaSO4). Although the work of Dantuluri (1988)
phate mine in Ormiston, SK, Canada. However, the integrated Airborne focused primarily on calcium-based absorbents, it was also stated that
technology still needs to be demonstrated on commercial scale. The sodium carbonate solution (Na2CO3) can be utilized during the semi-dry
Airborne Process™ consists of four main unit operations, namely dry scrubbing process (e.g., spray using an atomizer into a drying vessel).
sorbent injection, wet scrubbing, oxidant wash and sorbent regener­ Srivastava and Jozewicz (2001) as well as US EPA (2003) highlighted
ation/fertiliser production (Carpenter, 2013). In the first step, dry that the spray dry scrubbers typically use lime (CaO) as sorbent, owing
NaHCO3 is injected into the flue gas downstream of the particulate to its lower price as compared to its sodium-based sorbent counterpart.
control device employed within the combustion process (e.g., fabric Although cheaper than wet scrubbing, SDSs are normally restricted to
filter plant or electrostatic precipitator). Here the NaHCO3 thermally power plants with a capacity lower than 400 MW as larger power plants
decomposes to form porous sodium carbonate particles with a large require multiple absorber systems which may reduce its capital cost
surface area which are highly reactive to acid gas species. It has been advantage (Carpenter, 2012; US EPA, 2003). Further advantages offered
found that Na2CO3 produced in this way is a significantly better absor­ by semi-dry FGD includes low water consumption (~60 % less than wet
bent than commercially produced analogues. SO2 and SO3 react with the FGD), low energy consumption (0.3 – 1.0 %), opportunity to use low
Na2CO3 and NaHCO3 to form Na2SO4. After the direct sorbent injection quality steel, minimal scaling and plugging problems, no need for sludge
step, a mixture of flue gas and sorbent is then introduced at the bottom handling equipment, etc. However, operational expenditure for SDS is
of a wet scrubber unit where the flue gas contacts the sodium scrubbing higher due to low sorbent utilization, increased frequency of mainte­
solution counter currently. The formed Na2SO4 is dissolved into the nance on FFPs because of greater wear as well as higher sorbent and
solution whilst any unreacted SOx are also removed. The spent sodium waste disposal costs when compared to wet FGD (Carpenter, 2012). A
scrubbing solution is collected and withdrawn from the bottom of the critical review conducted by Koech et al. (2021) on recent developments
scrubber into an oxidizing unit where aeration is carried out to convert with respect to spray dry FGD, albeit more focused on the use of tradi­
sodium sulphite (Na2SO3) in the spent solution to Na2SO4. Hereafter, the tional calcium-based sorbents, identified the need for further research as
solution is conditioned with Na2CO3 to adjust the pH to enable the it relates to droplet–gas absorption kinetics, accurate process

170
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

optimization employing advance parametric studies and computational injection, and using low NOx burners. Johnson et al. (2005) and Kong &
fluid dynamics modelling, spray dry FGD waste management as well as Balland (2011) further reported that Na2CO3 also can remove HCl and
the quantification of multiple pollutant removal efficiencies. HF according to Reactions R14 and R15:
Na2 CO3 + 2HCl→2NaCl + H2 O + CO2 (14)
3.4. Dry sodium-based FGD processes
Na2 CO3 + 2HF→2NaF + H2 O + CO2 (15)
3.4.1. Brief overview of dry sodium-based FGD scrubbers
The application of dry FGD scrubber technology presents a viable In comparison to calcium-based absorbents (e.g., limestone and
option to the power utility and industrial sector (both locally and lime), sodium-based absorbents are significantly more reactive when
internationally, especially in water scarce countries) receiving increased injected as dry powders (Yeh et al., 1982; Zhu, 2010). In general, the
pressure to retrofit their facilities with SO2 abatement technologies, to advantages offered by dry FGD scrubbing includes no or lower process
reduce emissions to the allowable country specific, emission limits. The water requirements, lower capital and operating costs, less space
selection of dry scrubbing technology, like other technologies, is influ­ required than wet FGD systems, etc. However, for sodium-based ab­
enced by unit size, unit age, fuel usage, financial base, geographic sorbents, the disadvantages include, amongst others: (1) leaching of
location, etc. During the dry scrubbing process, the sorbent reacts with landfilled solid by-products, (2) CO2 production (contributing more to
SO2 to form a dry product containing sulphates and sulphites. After the the GHG emissions burden) and (3) lower SO2 removal efficiency when
scrubbing section, the mixture of spent sorbent material and fly ash can compared to wet FGD.
be removed from the gas stream by either ESP or FFP. The spent sorbent
is usually disposed of, but in some cases, it may be regenerable or have 3.4.2. Sodium-based, dry sorbent injection processes
commercial end uses (Miller et al., 1979; Yeh et al., 1982). When nah­ Dry sorbent injection, in which the sorbent (e.g., trona, nahcolite,
colite (or any synthetic form of sodium bicarbonate) and/or trona are NaHCO3, etc.) is injected directly either into the furnace or post-furnace,
injected post-combustion (above 125◦ C), they thermally decompose offers a low cost FGD alternative due to the lower capital cost incurred
according to Reactions R3 and R4 to form sodium carbonate, carbon because of equipment installation and or modification (Muzio and Often
dioxide (CO2) and water: 1987; Often et al., 1987). Dry injection technologies have been under
investigation since the 1960s and are in some cases not considered
2NaHCO3 →Na2 CO3 + CO2 + H2 O (3) popular due to several reasons, e.g., SO2 removal efficiencies not always
greater than 90 %, lower value and solubility of solid by-products
2(Na2 CO3 ⋅NaHCO3 ⋅2H2 O)→3Na2 CO3 + CO2 + 5H2 O (4)
(Muzio and Often 1987). More recently though, dry sorbent injection
The evolution of water vapour and CO2 during the thermal decom­ has been classified (as per the European BREF - best available technol­
position process results in the formation of microporous sorbent parti­ ogy reference documents) as the best available technique for no less than
cles with surface areas 5 to 20 times larger than the original particle seven different manufacturing industries in the European Union, i.e.,
surface area (Carpenter 2012; Kong and Wood, 2010). This relatively ceramics, common wastewater, iron & steel making, large combustion
high surface area increases the reaction rate between Na2CO3 and SO2. plants glass, cement & lime, and waste incineration (Balland et al.,
The produced sodium carbonate intermediate reacts with SO2 or SO3 to 2017). As a measure of its increasing relevance, one might also consider
produce sodium sulphite (Na2SO3) or sodium sulphate (Na2SO4) ac­ the installation of 14 dry sorbent injection systems at power stations in
cording to the following reaction schemes: the USA (US EPA, 2019). Fig. 1 below provides a schematic overview of
the typical injection locations (Carpenter, 2012).
Na2 CO3 + SO2 →Na2 SO3 + CO2 (5)
To illustrate the performance capability of sodium-based sorbents, a
full-scale dry sorbent injection test campaign was successfully con­
1
Na2 CO3 + SO2 + O2 →Na2 SO4 + CO2 (6) ducted in 2011 on a commercial scale power generation unit (Unit 3 -
2
150 MW) of the Battle River Station of ATCO Power, using both trona
Na2 CO3 + SO3 →Na2 SO4 + CO2 (7) and pre-milled NaHCO3 as absorbents (EPRI, 2011). Sorbent injection
was facilitated via 12 lances into two ducts downstream of the secondary
Recycled and/or make-up NaHSO3 and Na2SO3 react with SO3 in the air heater. A SO2 removal efficiency of 45.6 % and 57.6 % was achieved
flue gas according to: for trona (injection rate of 3.4 ton/h and unit load of 110 MW) and
NaHSO3 + SO3 →NaHSO4 + SO2 (8) NaHCO3 (injection rate of 1.9 ton/h and unit load of 70 MW) respec­
tively. At full load and an injection rate of 4.3 ton/h a SO2 removal ef­
Na2 SO3 + 2SO3 + H2 O→2NaHSO4 + SO2 (9) ficiency of 50.7 % was achieved for NaHCO3 (EPRI, 2011). Sorbent
injection systems are best suited for coal-fired power stations, using low
Wilhelm (2004) reported that reactions R8 and R9 will be given by
to medium sulphur coals, with capacities less than 300 MWe and mod­
reactions R10 and R11, when the molar ratio of Na:SO3 exceeds 1.
erate SO2 removal requirements (Carpenter, 2012; US EPA, 2019; Zhou
2NaHSO3 + SO3 →Na2 SO4 + 2SO2 + H2 O (10) et al., 2010). For South Africa’s power utility company, Eskom, DSI
therefore provides a suitable option for power stations that have derated
Na2 SO3 + SO3 →Na2 SO4 + SO2 (11) capacities and are close to reach their end of useful life. Exceptions on
the rule do exist as is evident from the installation and demonstration of
Zhu (2010) reported that side reactions between NaHCO3 and NO
DSI systems on boilers with generation capacities of up to 1300 MWe
takes place (reaction R12), when a small portion of NO is converted to
(Campobenedetto and Silva, 2011).
NO2, to subsequently produce NaNO3 (thus reducing NOx emissions).
Dry sorbent injection processes are categorised according to the
The removal of NaNO3 is given by Reaction R12 and R13 (Johnson et al.,
following distinct injection locations: (1) Furnace Sorbent Injection
2005)
(FSI), (2) Economiser Sorbent Injection (ESI), (3) Duct Sorbent Injection
2NaHCO3 + SO2 + NO + O2 →Na2 SO4 + NO2 + H2 O + 2CO2 (12) (DSI) and (4) hybrid systems (mostly a combination of the furnace and
duct injection systems or may involve humidification of the flue gas).
1
2NaHCO3 + 2NO2 + O2 →2NaNO3 + H2 O + 2CO2 (13) These processes employ the pneumatic transportation of the powdered
2 sorbent, via injection lances, into the designated injection locations,
Some of the measures suggested by Zhu (2010) for avoiding brown whilst the solid by-products (i.e., sulphites and sulphates) are collected
stack plume, includes adding urea to NaHCO3, high temperature with the fly ash using the appropriate particulate matter control device

171
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

Fig. 1. Possible sorbent injection locations (adapted from Carpenter, 2012).

(ESP or FFP). Fabric filters have been found (Carpenter, 2012) to sorbent used, the system is suitable for the removal of not only SO2 and
generally achieve higher SO2 removal efficiencies than ESPs due to a SO3 but also HCl and HF. The DSI technology has also been shown to
prolonged reaction time because of the additional flow resistance pro­ enhance mercury removal and is currently also in commercial use in
vided by the FFP filter cake. Sahu (2013) reported that various re­ several power stations across the US (Carpenter, 2012).
searchers and vendors investigated trona (unmilled Solvay T200 (D50 = Low temperatures (typically between 120 and 180◦ C) within the
30 µm)) injection at the economizer inlet, air heater inlet, and the ESP ducting (below the melting point of sodium carbonate, sodium sulphite
inlet locations, over temperatures of 374 ◦ C, 288 ◦ C and 110 ◦ C, and sodium sulphate) makes the technology suitable to employ sodium-
respectively. based absorbents for FGD (e.g., nahcolite, trona, sodium carbonate etc.).
Chemical reactions previously discussed mentioned (thermal
Furnace sorbent injection (FSI). Furnace sorbent injection involves the decomposition and SOx removal) are also applicable in the case where
direct injection of the sorbent into the top part of the boiler/furnace sodium-based absorbents are used during DSI. In summary the advan­
(where the operational temperatures typically range between 750 and tages of sodium-based duct sorbent injection, more specifically, include:
1230ºC). FSI is the simplest and most cost-effective, commercial FGD
technology to install. Operating costs can, however, be high due to low ■ Sodium-based absorbents are more reactive (70 to 90 % removal
or insufficient sorbent utilization (Carpenter, 2012). It should be noted efficiency) than their calcium-based counterparts (50 to 75 %
that the addition of solid absorbents via FSI can, however, reduce boiler removal efficiency) and do not require any flue gas humidification
efficiency especially if injection is facilitated directly into the combus­ (Carpenter, 2012; VanDerWerff, 2011).
tion zone. Limestone, dolomite and dolomite (CaCO3.MgCO3) are mostly ■ Both NaHCO3 and trona have been shown to reduce NOx emissions
employed for FSI, whilst literature on injection of sodium-based absor­ by around 10–25 % without an SCR (Maziuk, 2007).
bents remains scarce and can be attributed mainly to high operational ■ Sodium-based absorbents decrease fly ash resistivity which in turn
temperatures in the furnace. At these temperatures (generally higher may also result in the improvement of ESP performance (Carpenter,
than the normal boiling point of sodium-based absorbents) 2012).
sodium-based absorbents are expected to melt and cause slagging or ■ Sodium-based DSI also offers a multipollutant reduction co-benefit
fouling within the furnace. These melted absorbents have been also re­ through the enhancement of mercury, SO3, HF and HCl removal
ported to be poisonous to Selective Catalytic Reduction (SCR) catalysts (Carpenter, 2012).
(Blythe, 2004).
Drawbacks of sodium-based duct sorbent injection, more specif­
Economizer sorbent injection (ESI). ESI involves the injection of sorbent ically, include:
into the economizer of a coal-fired boiler at operational temperatures
ranging between 300 and 650ºC (Carpenter, 2012). An ESI study per­ ■ Higher running costs due to the higher raw material costs as
formed on a 90 MW boiler (combusting bituminous coal with 0.7 wt. % compared to lime and limestone (Carpenter, 2012).
S), using trona (normalized stoichiometric ratio of 2.5) as sorbent ■ The recovered fly ash/ Na2SO4 mixture has been reported to have a
resulted in a 50 % SO2 removal efficiency. Furthermore, reducing the d50 little economic value (high sodium content makes it unsuitable for
particle size of the trona sorbent from 30 µm to 13.7 µm improved the cement/concrete use) (Carpenter, 2012).
SO2 removal efficiency to 74 % for a trona normalized stoichiometric ■ Na2SO4 is water soluble which makes waste disposal difficult and
ratio of 3.5. A computational fluid dynamics (CFD) study conducted by also poses a significant environmental risk (Sahu, 2013).
Cremer et al. (2008) to optimize the injection location and distribution ■ Trona has been reported to be more difficult to handle than hydrated
of trona concluded that a SO2 removal efficiency of 80 % and 90 % could lime due to its small particle size (approximately 28 µm average
be achieved for a normalized trona stoichiometric ratio of 2.5 and 3.5, particle size), its cohesiveness, affinity for moisture, and its tem­
respectively. For calcium-based ESI, it has been shown that the injection perature sensitivities (Ritzenthaler, 2007; Ritzenthaler et al., 2007).
of water in the ducting between the air heater and dust collection device ■ The formation of liquid sodium bisulphate (NaHSO4) at tempera­
could increase the SO2 removal efficiency to up to 80 %. The increased tures greater than 185ºC, which can deposit on the air heater and
efficiency could mainly be attributed to the hydration of any unreacted duct, causing build-up and plugging. Soot blowing might therefore
CaO. This approach may present a means to also improve the SO2 be required on a regular basis, which in turn may increase opera­
removal efficiency of sodium-based sorbents although the theory stands tional expenditure (Kong and Wood, 2010).
to be tested (Srivastava and Jozewicz, 2001). Despite its advantages ESI ■ Dry sodium-based sorbent injection chemical reactions are also
is not currently in commercial use which may warrant further known to reduce NO to NO2 (Sahu, 2013; Smith et al., 1997) which in
investigation. turn can result in an increase in NO2 emissions burden (brown plume
from the stack).
Duct sorbent injection (DSI). DSI involves the injection of the sorbent
directly into the ducting between different unit operations within the 3.4.3. Comparative assessment between direct sorbent injection processes
coal-fired power generation process (refer to Fig. 1). Depending on the A comparative assessment between the SO2 and SO3 emissions
reduction performance of ESI and DSI technology is provided in Table 1.

172
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

Table 1 sorbent injection schedule (continuous and batch), (8) injection at

Comparison of sorbent injection systems. temperatures up to 427 ◦ C, (9) pre-decomposition of the nahcolite as
Injection Sorbent Product SO2 removal SO3 removal well as (10) initial SO2 concentration. Yeh et al. (1982) investigated the
System efficiency ( %) efficiency ( %) effects of baghouse temperature and cleaning cycle time, sorbent par­
ESI Trona Na2SO3 50 - 80 - ticle size, sorbent/sulphur ratio, location of sorbent injection point, and
Na2SO4 - sorbent injection schedule (batch or continuous). The same authors
DSI Na2CO3 Na2SO3 - 90 - 98 (Solution (Muzio and Arand, 1981; Yeh et al., 1982) reported that, except for
NaHCO3 Na2SO4 - Injection) baghouse cleaning cycle rate, each parameter had a significant effect on
Trona Na2SO3 70 - 90 ≥ 90
Na2SO4 70 − 80 ≥ 90
SO2 removal efficiency.
Na2SO3
Na2SO4 4.1. Sorbent type, particle size and stoichiometric ratio
Na2SO3 Na2SO4 - 90 - 98 (Solution
NaHSO4 - Injection)
Amongst all available sodium-based absorbents (e.g., thermonatrite,
NaHCO3 Na2SO4 - 91 - 98 (Solution
NaHSO4 - Injection) trona, nahcolite, natron, dawsonite, gaylussite, shortite, burkeite and
hanksite), the most used and studied sodium-based absorbents are
commercial sodium bicarbonate, trona and nahcolite (Kostick, 1994).
FSI has been excluded from this comparison as sodium-based absorbents Consequently, this section of the paper focuses solely on the effect of
have been reported to be incompatible with the typical high tempera­ these absorbents on SO2 removal efficiency. Muzio & Arand (1981) and
tures experienced in pulverized fuel boilers. Table 1 was adapted from Yeh et al. (1982) compared the SO2 removal efficiencies of nahcolite and
the table provided by Carpenter (2012) and is based on the collective trona whereas Smith et al. (1997) compared the SO2 removal rates be­
work of Benson et al. (2003), Blythe (2004), Commission (2006), tween 11 µm diameter sodium bicarbonate and 10 µm sodium sesqui­
Cremer et al. (2008), Srivastava & Jozewicz (2001) and VanDerWerff carbonate (trona) at flue gas temperatures 204 – 210 ◦ C; SO2
(2011). SO2 and SO3 removal efficiencies quoted in the table above are concentrations between 450 and 500 ppm; and 2.5 s in-duct residence
typical ranges reported in literature and will be site and operations time. Muzio & Arand (1981) reported that the behaviour of nahcolite
specific. As an example, a power generation facility equipped with a was similar to that of commercial NaHCO3, despite the fact that the
fabric filter plant will generally achieve a higher (5–10 %) SOx removal latter had a relatively lower SO2 removal efficiency. Smith et al. (1997)
efficiency than one with ESPs as the filter cake on the fabric filter in­ reported that both trona and NaHCO3 were able to achieve 70 % SO2
creases the surface area and contact time for the SO2/SO3 to react with removal with NaHCO3 exhibiting a higher utilization of sodium when
the sorbent. An even distribution of the sorbent across the boiler/duct compared to trona. As such, sodium bicarbonate can achieve 70 % SO2
and an adequate residence time at the proper temperature are critical for removal at a lower molar ratio of sodium to SO2 than trona, (i.e., 2Na/S)
high SOx removal efficiencies. of 1.9 and 0.9, for trona and sodium bicarbonate, respectively.
A higher SO2 reduction (up to 90 %) can be achieved when injecting Furthermore, the investigation of Smith et al. (1997) indicated that
dry sodium-based absorbents into the flue gas duct. Although ESI can NaHCO3 fines (mean particle diameter of 11.2 µm) had the highest SO2
achieve 50–80 % SO2 reduction, it is not currently in commercial use removal efficiency. Their findings show that NaHCO3 exhibits a signif­
(Carpenter, 2012). In most cases, SO3 control efficiencies of 70 % to 90 icantly larger SO2 removal efficiency than trona within the NSR
% are required to appreciably lower, or eliminate, plume opacity due to (normalized stoichiometric ratio) range of 0.5 to 2.5 (e.g., for NaHCO3 a
the increased SO3 concentration from the oxidation of SO2 to SO3 in the NSR of 1 is required to achieve 50 % SO2 removal efficiency, whilst for
SCR unit. DSI is more efficient at removing both furnace- and trona a SO2 removal efficiency of only 34 % is achieved at the same NSR
SCR-generated SO3, capturing over 90 % with sodium-based absorbents level). These observations are also in agreement with what was observed
(Carpenter, 2012). Sodium bisulphite, sulphite and carbonate can also by other investigators (Carson, 1980; Keener and Davis, 1984; Muzio
be injected as solutions as part of the duct spray dry process to remove and Sonnichsen, 1984; Shah et al., 1978; Stern 1978). An inverse linear
90–98 % of the SO3. Flue gas humidification or conditioning with SO3 relationship between SO2 removal and mean particle diameter was
coupled with the injection of sodium-based absorbents can possibly also observed by Yeh et al. (1982) i.e., SO2 removal increased with the
increase fly-ash reactivity (reduction of resistivity) leading to a subse­ decrease in mean particle diameter.
quent increase in ESP performance. The same authors (Smith et al., 1997; Carson, 1980; Keener and
However, as stated above, trona is more difficult to handle than Davis, 1984; Muzio and Sonnichsen, 1984; Shah et al., 1978; Stern 1978)
hydrated lime (Carpenter, 2012). Caution is expressed when employing reported SO2 removal efficiencies of up to 90 % for absorbents with
humidification as part of sodium-based direct sorbent injection as the mean particle diameters of less than 65 µm. Muzio & Arand (1981) re­
formation of liquid NaHSO4 can result in unacceptable fouling within ported that decreasing the nahcolite particle size improved the SO2
parts of the duct and downstream equipment. Another concern with the removal efficiency, particularly at cycle times shorter than 60 min (50 %
use of both sodium and calcium-based absorbents is the effect collected removal for 37 µm nahcolite compared to 30 % for 74 µm nahcolite over
spent sorbent on the market value and saleability of the fly ash. a 60-minute cycle time at a NSR of 0.8). The work of Keener & Davis
Pre-collection of the fly ash may resolve this concern. Finally, another (1984) also confirmed the dependence of the SO2 removal rate of
major drawback with the carbonate-type absorbents is the generation of Na2CO3 on particle size (in the range of 20 to 200 μm) for temperatures
CO2 from the reactions with SO2 and SO3, which inevitably increases less than 232◦ C. Accordingly, for a reaction period of 600 s, SO2 removal
total CO2 emissions from the coal-fired power plant (Carpenter, 2012). rates of close to 50 %, 35 % and 20 % was respectively achieved for 20,
90 and 200 μm particles. In contrast, however, it was found that the
4. Factors affecting the efficiency of dry sodium-based FGD NaHCO3 sulphation reaction was less dependent on particle size. Keener
processes & Davis (1984) ascribed this to the probable enhancement of sorbent
particle characteristics because of thermal decomposition which
The SO2 removal efficiency of sodium-based absorbents is dependent occurred during the reaction.
on several factors. Muzio and Arand (1981) quantified the effects of the Stoichiometric ratio is another important factor that affects the SO2
following process parameters on the dry SO2 removal process: (1) sor­ removal efficiency of sodium-based absorbents. A study by Yeh et al.
bent type (i.e., nahcolite, trona, commercial sodium bicarbonate), (2) (1982) reported an SO2 removal of 90 % for a Na2/S ratio of 1.3 while
normalized stoichiometric ratio (NSR), (3) sorbent particle size, (4) using NaHCO3 of particle size 32 µm in a baghouse at 204.44 ◦ C when
baghouse temperature, (5) air-to-cloth ratio, (6) cleaning cycle time, (7) coal of 1.6 % S was combusted. NaHCO3 utilization was estimated at 70

173
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

%. To study the effect of stoichiometric ratios, some authors (e.g., Muzio 4.3. Injection location
and Arand, 1981) employed the phenomenon known as normalized
stoichiometric ratio (NSR), which is described as the measure of the From the field tests conducted by various vendors and researchers,
amount of sodium injected relative to the sulphur in the flue gas as Sahu (2013) reported that one set of tests focused on the effect of in­
measured by a continuous SO2 monitor. A NSR of 1 therefore implies jection position on the SO2 removal efficiency of trona. The positions
that 2 mol of Na per mole of SO2, which is the stoichiometric require­ that were considered include the economizer inlet, the air heater inlet,
ment to form sodium sulphate, Na2SO4. For nahcolite a SO2 removal and the ESP inlet. The reported, respective average temperatures
efficiency of 67 % can be achieved when a NSR of 1.0 is used (Muzio and recorded in those positions were 374 ◦ C, 288 ◦ C and 110 ◦ C respectively.
Arand, 1981). For trona, the same authors reported a SO2 removal ef­ These tests were primarily carried out using an as-received, unmilled
ficiency of 40 % at NSR of 1.0, and a SO2 removal efficiency of 70 % at Solvay T200 material (D50 of 30 µm). From the experiments it was
NSR of 3.0. concluded that the highest SO2 removal efficiency was realized during
A comparison between SO2 removal efficiencies and NSR during injection at the economizer inlet (ranging between 48 and 55 % for NSRs
sodium injection trails for commercial demonstration (22 MW and 100 between 2.3 and 2.5), followed by the air heater (ranging between 42
MW) showed close agreement with the work of Muzio & Arand (1981). and 43 % for NSRs between 2.1 and 2.5) and then the ESP inlet (ranging
From the demonstration-scale results it was concluded that sodium bi­ between 30 and 35 % for NSRs between 2.6 and 2.7)
carbonate (or nahcolite) is the most effective, with a SO2 removal effi­ According to Yeh et al. (1982) no significant effect of injection
ciency of 80 % at a NSR of 1, while trona could only achieve a 67 % location (baghouse inlet versus duct injection upstream of air heater
removal efficiency at a similar NSR. In addition, Na2CO3 was found to be inlet) was observed when using sorbent with a mean particle size of 32
relatively unreactive (only between 5 and 15 % SO2 removal efficiency), µm. However, for sorbent particles with a mean particle size of 69 µm
although other investigations (Albin, 1986; Hooper and Bland, 1985; SO2 removal efficiencies were reported to decrease from 87 % when
Muzio and Often, 1987) have shown higher removal efficiencies. Similar injecting upstream, to 64 % when injecting at the baghouse inlet.
observations with respect to the effect of stoichiometric ratio was made Similarly for 180 µm particles, a decrease from 64 % to 43 % was re­
by Pilat & Wilder (2007) and Walawska et al. (2014). ported respectively. The authors attributed these observations to parti­
cle separation and settling, i.e., the coarser and heavier particles
separate from the flue gas and settle out before chemical reactions can
4.2. Sorbent injection schedule progress further. To better compare the effect of particle size and in­
jection location, the authors maintained the baghouse temperature at
Yeh et al. (1982) studied the effect of sorbent injection schedule on 204.4 ◦ C, whilst using a Na2/S of 1.5 and particle sizes of 32, 69 and 180
both SO2 removal efficiency and sorbent utilization. During the inves­ µm. Muzio & Arand (1981) also made consistent reports whilst
tigation two experimental sets were employed, namely: (1) a compari­ comparing the SO2 removal efficiency on the duct and on the inlet of the
son of continuous injection to a 1 min on / 1 min off injection regime and baghouse. In their studies using trona at a NSR of 2, they reported SO2
(2) a comparison of continuous injection to a 3 min on / 1 min off in­ removal efficiencies of 70 to 78 % for duct injection as compared to the
jection regime. The findings from this study are summarised in Table 2, 50 to 62 % observed at the inlet to the baghouse.
clearly illustrating the higher SO2 removal efficiencies and sorbent uti­
lization obtained for both experimental sets. 4.4. Operating temperature
Muzio and Arand (1981) tested the performance of nahcolite on SO2
removal efficiency by comparing batch and continuous injection The effect of baghouse temperature using NaHCO3 on SO2 removal
schedules into the baghouse inlet. Their injection schedule results were efficiency was investigated by Yeh et al. (1982) and their findings
classified as follows: (I) continuous injection, steady-state level, indicated that the SO2 removal efficiency increases with an increase in
maximum removal, (II) continuous injection, 90-minute average temperature and reaches a plateau around 204.44 ◦ C. Consistent reports
removal, (III) bag precoated (batch), 90-minute average removal and were also made by Muzio & Arand (1981), who reported that for trona,
(IV) continuous injection for 54 % of 90-minute cycle, 90-minute nahcolite and synthetic Na2CO3, the SO2 removal efficiency was
average removal. These tests were conducted under NSR of 0.9, enhanced by higher temperatures. For nahcolite, it was observed that a
200-mesh (74 µm) nahcolite and an injection temperature of 132.2 ◦ C. It maximum SO2 removal efficiency is achieved around 288 ◦ C. Further­
was evident from their results that during continuous injection, the SO2 more, Smith et al. (1997) reported that the humidification of the flue gas
level at the baghouse outlet reaches a steady-state level (around 40 min) to a 15.6 ◦ C saturation temperature increased SO2 removal by up to 20 %
and maintains this level for the remainder of the test period. In contrast, when injecting trona (at constant Na2/S ratios). More recently, Lu et al.
the SO2 concentration rapidly went through a minimum for the batch (2022) investigated the dry desulphurization performance of NaHCO3
injection experiment and steadily rises again as the nahcolite, which was (d10 = 3.2 μm; d50 = 13.2 μm; d90 = 33.5 μm) in a fixed bed experimental
precoated on the bags, is consumed. Their findings illustrated the im­ assembly and concluded that SO2 removal efficiency remained relatively
provements made in SO2 removal efficiency by either precoating the constant (between 95 and 100 %) over the temperature range of 100 to
bags or injecting at a rate such that the total amount of sorbent is 190◦ C. Hereafter, it reduced significantly (to close to 60 % SO2 removal
injected during the first 54 % of the cycle. In both cases, SO2 removal efficiency at 220◦ C) with further temperature increases. Although the
increases from about 50 % for continuous injection to 66 % for the batch authors do not offer an explanation to the observed SO2 removal trend it
and semi-batch feeding of the nahcolite over a 90-minute test period. is believed that it can be ascribed to a change in sorbent physical
characteristics (specific surface area) as a result of the thermal decom­
Table 2 position of NaHCO3. As an example, Walawska et al. (2014) have shown
Effect of sorbent injection schedule (taken from Yeh et al., 1982). that the BET (Brunauer, Emmett and Teller) surface area of NaHCO3
increases, depending on the mechanical activation method employed,
Sorbent Injection Na2/S SO2 Removal Sorbent Utilization
Schedule Ratio Efficiency ( %) ( %) by a factor between 3 and 10 when thermally activated at 300◦ C. A more
in-depth investigation with respect to the relationship between SO2
Experimental set 1
Continuous 1.30 77.6 59.7
removal efficiency, sorbent physical properties and temperature could
1 min on: 1 min off 1.30 58.6 45.1 therefore provide useful for future comparative studies between
Experimental set 2 different sodium-base absorbents.
Continuous 0.95 65.1 68.5
3 min on: 1 min off 0.98 48.0 49.0

174
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

4.5. Sulphur content in coal changes (less than 5 %) for a 1 s sorbent particle duct residence time at a
gas velocity of 15 m/s, particle diameter of 5–10 mm, and temperatures
As coal is heterogenous in nature it can contain different concen­ in the 120–210.8 ◦ C range.
trations of organic sulphur and inorganic sulphur. The proportions of
these two main forms of sulphur may also vary from coal seam to coal 4.8. Baghouse air-to-cloth ratio
seam; and even if the total sulphur is the same the total amount of
sulphur released as SOx can vary significantly under identical combus­ The baghouse air-to-cloth ratio is defined as the actual volumetric
tion conditions. Yeh et al. (1982) investigated SO2 removal with nah­ flow rate through the baghouse divided by the area of the bags. This is
colite while combusting three different coals: Kentucky coal (1.0 wt. % effectively the velocity of the gases through the bags. Thus, the air-to-
S), Pittsburgh coal (1.6 wt. % S) and West Virginia coal (3.1 wt. % S). cloth ratio determines the relative residence time as the gases pass
The baghouse temperature was maintained between 204 and 216 ◦ C through the filter cake (Muzio and Arand, 1981). Muzio and Arand
during their investigation. From their investigation it was concluded (1981) varied the air-to-cloth ratio by lowering the firing rate which
that a SO2 removal efficiency of 90 % and sorbent utilization of 82 % was simulates low load operation of a utility system; and by removing half
achieved for the Pittsburgh Coal, whilst the Kentucky coal showed a the bags from the baghouse. Increasing the air-to-cloth ratio from 2 to 4
similar trend. However, the reported SO2 removal efficiencies achieved by removing the bags had essentially no effect on the SO2 removal
when burning the West Virginia coal were lower than those obtained for process while using nahcolite injection into the baghouse.
the other two coals. The effect of sulphur content was further investi­ The same authors reported that increasing the air-to-cloth ratio and
gated whilst employing trona as a sorbent. From these experiments it (thus decreasing the residence time of the gas as it passed through the
was observed that the difference in SO2 removal efficiencies of Pitts­ filter cake containing the sodium) could be expected to reduce the
burgh coal and that of West Virginia coal was more pronounced. effectiveness of the SO2 removal process. Results of these tests at NSRs of
0.93 and 2.33 indicated that there was no significant difference between
4.6. Inlet SO2 concentration the SO2 removal efficiencies at air-to-cloth ratios of 3 or 4 either for the
removal based on the steady-state SO2 level or the integrated removal
Pilat & Wilder (2007) studied the effect of the inlet SO2 concentra­ for a 60-minute cycle time. At a NSR of 0.93, the SO2 removal based on
tion on the removal efficiency at flue gas temperatures ranging between the steady-state level was somewhat lower at the high air-to-cloth ratio,
204 and 210 ◦ C with a 2.5 s in-duct residence time. During the study five although it appears to be within the margin of error of the data.
test series were executed using 11 mm MMD NaHCO3 particles whilst
spiking the flue gas with pure SO2 (concentrations ranging between 300
and 2600 ppm). From the results it was deduced that the inlet SO2 4.9. ESP operation
concentration had very little effect on the SO2 removal efficiency of the
small 11 mm MMD NaHCO3 particles. Pilat & Wilder (2007) also studied the effect of ESP corona voltage on
Pilat & Wilder (2007) attributed their observations to the fact that the SO2 removal efficiency, whilst performing NaHCO3 injection at
similar reaction rate limiting steps were dominant throughout this 204◦ C with an absorbent particle size of 11 μm and SO2 inlet concen­
300–2600 ppm concentration range and that no gas phase diffusion tration of 350 ppm. Results from this campaign are summarized in
limitations were present. Their report was consistent with the work of Table 3, where a decrease in ESP corona voltage can be observed to
Davis et al. (1978), who concluded that the inlet SO2 concentration had result in a slight increase in the SO2 removal efficiency. According to the
no effect on its removal efficiency. However, a study examining the authors, lower Corona voltages probably increased the gas-particle
removal efficiency of a bench-scale thin-bed of nahcolite performed by contact time in the ESP due to lower dust collection efficiencies.
Stern (1978) suggested a dependence on SO2 concentration. Stern’s At a higher stoichiometric ratio and upstream residence time(i.e., 1.5
bench scale studies concluded that SO2 removal efficiencies increased and 5 s), if was found that both the upstream and downstream SO2-
with increasing inlet SO2 concentrations. A theoretical model predicting removal efficiencies increased by a factor of between 2 and 3 (i.e., 25 to
a higher SO2 removal efficiency with higher inlet SO2 in the 200–3000 75 % and 40 to 80 % respectively). The relative increase in SO2 removal
ppm SO2 concentration range at 127◦ C was also presented by Wu et al. across the ESP appeared to depend on the extent of SO2 removal (SO2
(2004). reaction with the sorbent) upstream of the ESP.

4.7. Residence time in the duct 5. Developments related to other sodium-based absorbents

As previously discussed, the dry absorbents can be injected either at Kumar & Jana (2021) recently published a review paper in which
the inlet to the baghouse (downstream of the heat exchanger) or into the they provided a critical review of advances made in absorbents and
high temperature duct. The residence time of the combustion products techniques for both wet and dry FGD. In the review a comprehensive
and dry sorbent in suspension in the duct from the point of injection to account of absorbent performance is provided for not only the more
the heat exchanger is normally determined to be in the range of 2 s. In an commonly known and used absorbents (i.e., limestone and lime) but
actual utility boiler, residence times between the convective section and
the air preheater are on the order of 0.5 s, and temperatures are Table 3
approximately 427 ◦ C. Effect of ESP corona voltage on SO2 removal efficiency (adapted from Pilat and
The residence times through the air heater are in the order of 0.5 s, Wilder, 2007).
with temperatures close to 149 ◦ C. The residence time in the duct be­ Duct SO2 inlet Stoichiometric Corona SO2 removal
tween the air heater and the particulate collection device is on the order residence concentration ratio voltage efficiency ( %)
time (s) (ppm) (kV/cm) ESP ESP
of a few seconds (Muzio and Arand, 1981). Pilat & Wilder (2007) studied
Inlet Outlet
the effect of in-duct residence time on SO2 removal efficiency for flue gas
2.5 350 1.0 2.6 25 40
temperatures ranging between 204 and 210 ◦ C and SO2 inlet concen­
2.5 350 1.0 1.3 – 45
trations ranging between 350 and 380 ppm. Significant SO2 removal 2.5 350 1.0 0.0 – 45
efficiencies (between 35 and 70 %) were achieved within a 3 s duct 5.0 350 1.5 2.6 75 80
residence time at a temperature of 204.4 ◦ C. The findings of Pilat and 5.0 350 1.0 2.6 – 52
Wilder (2007) were, however, not in agreement with the model pre­ 5.0 350 1.0 1.3 – 60
5.0 350 1.0 0.0 63
dictions of Wu et al. (2004) who reported very small SO2 concentration
–

175
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

also sodium-based absorbents. A summary of other (apart from Na2CO3, found to increase with an increase in L/G ratio.
NaHCO3, nahcolite and trona) sodium-based absorbents investigated for Contrary to the use of inorganic sodium-based absorbents, Sun et al.
application in FGD, as adapted from the work performed by Kumar & (2010) investigated the SO2 removal performance of an organometallic
Jana (2021), is summarized in Table 4. A few additional and noteworthy sodium-based absorbent (i.e., sodium humate) at varying absorbent
sodium-based absorbent studies have, however, been found by the au­ concentrations, pH, temperature, gas flowrate, O2 concentration as well
thors of this review paper and reference to these publications (**) are as SO2 and NO2 inlet concentrations. The investigators were able to
therefore subsequently included in Table 4 to supplement the review achieve SO2 removal efficiencies of more than 98 % and prescribed an
provided by Kumar & Jana (2021) absorbent concentration of 0.06 mg/L and a pH of 4.5 for optimum SO2
The work of Chang and Rochelle (1981) revealed that the SO2 ab­ removal performance. A similar study was also performed in 2013 by
sorption rate increased with increasing NaCl concentration and reaching Zhao & Wu (2013) using sodium humate extracted from sludge collected
a maximum value around 0.4 M NaCl. Furthermore, at constant HCl from the thickening tank of a wastewater treatment plant. Comparable
concentration and temperature no significant change could be observed SO2 removal efficiencies were observed whilst the authors also make a
in the mass transfer rate for a change in SO2 concentration. Experiments compelling case for employing or repurposing readily available, indus­
performed in a sieve tray column, maintaining an acidic environment, trial or commercial waste materials as a cost-effective means of
indicated removal efficiencies of 36 – 72 % for NOx and 88 – 100 % for combatting the emissions of harmful pollutants. To the best of the au­
SO2 respectively (Chien and Chu, 2000). The removal efficiency was also thor’s knowledge, no publications currently exist where either a waste,
sodium-bearing stream or organometallic sodium-based absorbent, such
as sodium humate, is used for either semi-dry or dry FGD application. A
Table 4 further addition to the focus area of non-conventional sodium-based
Summary of other sodium-based sorbent research performed (adapted from
absorbents is the comparative, sulphation performance study conducted
Kumar and Kana, 2021).
by Rhamani et al. (2015) wherein the effect of SO2 inlet concentration,
Type of Composition Experimental Type of Reference process temperature and liquid/gas ratio on SO2 removal efficiency of
absorbent of SO2 parameters contactor
α-amino salts (sodium glycinate and sodium lysinate) was investigated.
NaCl, H2O and Pure and 25 ◦ C, Stirred tank Chang & Based on the experimental results, the amino acid solutions showed
HCl dilute 1 atm reactor Rochelle
good SO2 removal efficiencies with sodium lysinate being the superior.
0.0002 – 0.98 (300 rpm) (1981)
bar A completely different approach to FGD was followed by Park et al.
NaOH with 5 × Pure 25 ± 2 ◦ C, Sphere and Vazquez (2017). The authors investigated both NO and SO2 removal efficiencies
10− 3 wt. % 1 atm cylinder et al. in a wet ESP by making use of NaClO2 as absorbent. From the study it
sodium column (1988) was determined that NO and SO2 removal rates of 34.8 – 72.9 mmol/m3.
lauryl reactor
s and 36.6 – 84.7 mmol/m3.s could be respectively achieved by varying
sulphate
(SLS) the molar concentration of the absorbent. Another novel approach w.r.t.
NaOH (0 – 0.9 Dilute 25 – 50 ◦ C Packed Schultes FGD was the utilization of a polypropylene hollow fibre membrane
M) column (1998) reactor by Joo et al. (2020) and Park et al. (2021) to study the SO2
NaOH & Dilute and 50 ◦ C, Stirred tank Chu et al.
removal efficiency using different absorbents (refer to Table 4). A 99.9
KmnO4 mixture 1 atm reactor (2001)
with NO
% removal efficiency was achieved by using lower CaO concentrations
NaClO with 300 – 800 65 ◦ C, pH 7, Sieve tray Chien & as compared to NaOH. No significant membrane deterioration was
NaOH ppm L/G ratio: column Chu (2000) observed even after multiple weeks of operation (Joo et al., 2020). The
5 – 10 L/m3 study of Park et al. (2021) is also only one amongst a very few studies
NaClO2 SO2: 0 – 1500 25 – 65 ◦ C Bench-scale Chien et al.
where pollutant removal efficiencies were evaluated in the presence of
ppm pH 4 – 10 spraying (2003)
NO: 100 – column two or more gaseous pollutants. From the results it was evident that in
800 ppm the presence of both CO2 and SO2, SO2 removal performance decreases
NaClO with 2000 mg/m3 50 ◦ C, Bubbling Zhao et al. due to fouling (“corrosion”) of the membrane contactors’ potting ma­
NaClO2 1 atm, pH 5.5 reactor (2010)
terial. However, the authors believe that this problem might be over­
(molar ratio
of 4)
come by using a potting material with high chemical stability.
**C9H8Na2O4 SO2: 1100 - 25 – 85 ◦ C Bubbling Sun et al.
(Sodium 2700 ppm 1 atm reactor (2010) 6. Further considerations for sodium-based dry sorbent
humate) NO2: 340 0.01 – 12 g/ injection
ppm ml
pH 4 - 10
**C9H8Na2O4 2000 - 3000 25 - 65 ◦ C Bubbling Zhao & Hu When selecting the most appropriate FGD process, one needs to
sludge ppm 1 atm reactor (2013) consider several factors including: electricity generation capacity of the
(Sodium 0.12 m3/h associated power plant, fuel type, fuel-to-sulphur ratio, availability, and
humate) (gas)
price of the absorbent to be used, absorbent type, potential market for
pH 3 - 11
**Na2HPO4, 1000 - 1800 25 - 60 ◦ C Gas-liquid Rahmanim
FGD by-products, waste disposal and management, as well as environ­
α-amino acid ppm 1 atm contactor et al. mental factors (e.g., waste disposal and wastewater treatment). From
salts (sodium L/G ratio: (Packed Bed) (2015) the preceding discussions and based on available SO2 removal efficiency
glycinate & 0.005 – 0.01 data, dry, sodium-based sorbent injection might be considered a viable
sodium
option for FGD, especially for a semi-arid country like South Africa. The
lysinate)
NaClO2 300 – 600 27 W power Wet Park et al. sections to follow seek to highlight some further operational and tech­
ppmv supply electrostatic (2017) nical aspects to consider for semi-dry and or dry, sodium-based FGD.
precipitator
NaOH, Na2SO3 2000 ppm 50 ◦ C, Hollow fibre Joo et al.
6.1. Potential impact on particulate matter emissions
& CaO 1 atm membrane (2020)
reactor
**NaOH, SO2: 371 – Ambient T & Hollow fibre Park et al. The dry sorbent injection process has the potential to increase the
NaHCO3, 3500 ppm P membrane (2021) particulate matter (PM) burden from ESPs due to an increased load on
Na2CO3, CO2: 15 vol. L/G ratio: reactor especially ESP ash handling systems (i.e., hoppers, fans, blowers, ducts,
Na2SO3 % 12.5 – 37.5
etc.). The subsequent increase in PM emissions can be attributed to

176
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

increased quantities of unreacted sorbent and FGD reaction products (e. sodium-based absorbents, are scarce in literature. The thermal decom­
g., sorbent with SO2, HF, HCl etc.) yielded during the FGD process. Both position of sodium bicarbonate/trona/nahcolite will vary along the
the unreacted sorbent and the resulting products do not only increase different injection locations, according to local temperatures. Further­
the loading, but will also alter the physicochemical properties of the more, lower temperatures (60 – 70 ◦ C) have been reported to favour CO2
solids to be handled by the ESP. The capacity of the ESP with respect to capture (Liang et al., 2004) by sodium-based absorbents, whilst tem­
the particulate loading, electrical resistivity of the particles, particles peratures above 120 ◦ C have been reported to favour decomposition.
size distribution, sulphuric acid mist concentration etc., will need to be
thoroughly considered to ensure optimum ESP performance. Electrical 6.4. Cost implications
resistivity dictates the ease or difficulty by which ash particles are
charged during ESP operation. According to Sahu (2013), sodium-based The capital cost and energy consumption of direct sorbent injection
absorbents have the added advantage of reducing ash resistivity and systems have been found to be considerably lower in comparison to
increase overall ESP performance. Flue gas conditioning with SO3 semi-dry and wet FGD processes due to their simplicity, lower water
(which converts to sulphuric acid mist in the presence of water vapour) consumption and relative ease to handle and manage dry products from
exhibits the same effect and according to Blythe (2004), sodium-based the process. Capital / installation costs have been quoted to range be­
sorbent has the affinity for preferentially absorbing sulphuric acid tween 8 and 12 US$/kW for SO2 control and 4 to 8 US$/kW for SO3/
mist. Upgrading and retrofitting existing ESPs to accommodate for these H2SO4 control (Campobenedetto and Silva, 2011).
combined FGD solutions can, however, result in tens of millions (USD) in Higher values (around 20 US$/kW) have, however, been quoted by
additional expenditure (Sahu 2013). For power generation facilities Staudt et al. (2011) for a basic injection system with storage silo.
equipped with fabric filter plants (FFP), an increase in particulate matter Operational costs will be plant and location specific and are influenced
burden through the flue gas stack can also occur due to more frequent by various factors including mining and transportation of raw material
bag failures (e.g., mechanical degradation) from a higher particulate (e.g., trona), production of raw material (e.g., sodium carbonate), ma­
loading which induces greater wear on the FFP bag filters (Carpenter, terial preparation (milling) and storage, PM control equipment upgrades
2012). / modifications (e.g., ESP) and landfill upgrade. Operational and capital
expenditures will be higher in cases where additional storage and ma­
6.2. Potential leachability of FGD by-products terials handling are necessary. However, this is still lower than the
capital cost for wet or dry scrubbers (approximately 400 US$/kW)
An environmental concern may arise from the leaching of Na2SO3 (Campobenedetto and Silva, 2011). A comparative / indicative cost
and or Na2SO4 (products of NaHCO3 or trona flue gas desulphurization) estimate for different SO2 abatement retrofit options, including for dry
during and after landfill disposal. This can be attributed to the higher sorbent injection with sodium-based absorbent, was recently conducted
solubility of these products as compared to the limestone / lime ana­ by Jacobs (2021) for the Vales Point power station (2 × 600 MW) in
logues (i.e., CaSO3 and CaSO4), The leachability of these sodium sul­ New-South Wales, Australia. The study confirmed the lower capital
phite/sulphate products depends on several factors which include (1) expenditure and higher operational costs for dry sorbent injection (in
moisture conditioning and handling, (2) water consumption during this case furnace sorbent injection) as can be seen in Table 5.
landfill dust control, (3) compaction of the wetted FGD products as well A large fraction of the operating cost can be attributed to the price of
as (4) pelletization (Madsen, 2017). The aforementioned factors have the sorbent used in FGD. In general, the sodium-based absorbents are
been reported to have a substantial impact on the hydraulic conductivity more expensive than the calcium-based absorbents. Srinivasn (2004)
and leaching behaviour (Madsen, 2017). On evaluating the impact of assessed and compared the raw material cost of natural occurring ab­
trona injection and its effect on the resulting ash product, Wang and Su sorbents for the US market (i.e., limestone, hydrated lime, trona and
(2011) concluded the following: NaHCO3) and found the cost of trona to be significantly lower
(approximately a factor 2) than NaHCO3. This aspect will however need
■ The injection of trona altered the physicochemical properties of the to be assessed on a country specific basis. Furthermore, sorbent utili­
fly ash, e.g., the specific surface area, particle morphology and zation in dry based FGD processes is normally lower than in wet based
microstructure. It also resulted in an increase in sodium, sulphur and processes, therefore resulting in higher control costs in terms of cost per
carbonate concentration. tonne of SO2 reduced. The use of real time, continuous measurement of
■ An increase in fly ash solubility, pH, and leachability of anionic el­ SO2 and SO3 in the power plant stacks could lower the sorbent injection
ements including fluoride, sulphate, chloride, and trace oxyanions of rate, with consequent cost savings.
concern especially As and Se was observed.
■ Compared to the conventional fly ash, trona ash leached significantly 6.5. Potential abrasiveness of absorbents
more As and Se in all conditions, including varying leaching time,
pH, storage time conditions. Multiple factors may contribute to the Of all the natural occurring sodium-based absorbents (e.g., sodium
enhanced As and Se leaching from trona ash, including more alkaline bicarbonate, trona and nahcolite), trona has been found to be the most
pH, greater ash solubility, presence of high concentrations of abrasive (VanDerWerff, 2011). The abrasive nature of trona can be
competing anions (such as sulphate and carbonate), and a greater Se attributed to its silica (SiO2) content, a factor that must be considered
(VI) fraction in trona ash.
■ The increased solubility of the produced fly ash will most probably Table 5
increase the leachability of As and Se into groundwater which Comparative estimate of SO2 abatement options for Vales Point Power Station
therefore necessitates the need for lining of landfills. (adapted from Jacobs, 2021).
Technology &. Furnace Sorbent Semi- Wet Seawater
6.3. Potential impact on CO2 emissions Techno-economic Injection Dry Scrubber Scrubber
aspect (SDA)
The chemical reactions involved in semi-dry and dry sodium-based Sorbent Sodium-based Lime Limestone Seawater
sorbent FGD processes were discussed previously or can be found else­ CAPEX (US$/kW) 41 – 58* 117 – 233 – 426 175 – 321
where (Carpenter, 2012; Pandey et al., 2005). Noticeably, is the fact that 361
all these chemical reactions produce CO2, the greenhouse gas. Studies OPEX (US$/kW-y) 18 – 23* 6–9 5–9 4–6

that focus on the emissions of CO2 (or more specifically multi-pollutant *

Based on sodium-based absorbent price of 371 US$ per tonne and 70 %
emissions control performance of absorbents), while using collection efficiency of bag filter plant.

177
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

during the design of the process of the pneumatic injection system to sodium-based FGD.
ensure operational and mechanical integrity throughout the operational
lifecycle of the facility. Premature and frequent failures may result in 6.7. Improving sorbent performance with additives and activation
higher operational costs. technologies

6.6. Valorisation potential of FGD by-products and recovery of spent In recent years, several research and technology development studies
sorbents have been performed to improve absorbent performance but also to
lower operational costs of semi-dry and dry FGD systems. Some studies
Desulphurization by-products derived from the use of sodium-based have focused specifically on the addition of additives whilst others have
sorbents in FGD have numerous applications ranging from metallurgy considered alternative absorbent activation strategies to enhance the
(mineral processing), wood processing, glass manufacture, commercial SO2 removal efficiency during semi-dry and or dry FGD. França et al.
and household laundry detergents and chemical industry among others. (2020) investigated the effect of seven additives (magnesium hydroxide,
One of the most notable uses of Na2SO4 is the Kraft Process (sulphate ammonium nitrate, ammonium acetate, ammonium phosphate, sodium
process) for wood pulp manufacturing to produce paper products and hydroxide, citric acid, and urea) on the SO2 removal efficiency of Ca
building supplies (Bajpai, 2018). This process utilizes heat reduction (OH)2 during semi-dry FGD. In their experiments, two concentrations (2
process to convert Na2SO4 to Na2SO3 which breaks the bond in wood and 4 wt. %) of additives were applied to Ca(OH)2 slurries of 10 and 20
cellulose making it malleable and able to be extruded (Bajpai, 2018). wt. % respectively. From the results it was concluded that the ammo­
Na2SO4 is also one of the main raw materials in glass manufacturing nium additives (especially ammonium nitrate) increased the SO2
process as a fining agent chiefly because it provides the necessary alkali removal efficiency by as much as 13.5 % when compared to the highest
base, prevents sum formation, and removes air bubbles from molten SO2 removal efficiency when only using Ca(OH)2. Very few research
glass during glass refining process (Global Markets Insights, 2021). papers do, however, consider the effect of additives (e.g., urea and
The application of Na2SO4 in mineral processing has been tested as ammonia) on the SO2 removal efficiency of sodium-based sorbents in
an additive in pyrometallurgy to obtain ferronickel concentrates from either semi-dry or dry application (Bland, 1990; Darmastaedter, 1990;
nickel laterite ores (Jiang et al., 2013). This has been effectively tested Hooper, 1988). Unfortunately, a thorough review of more recent liter­
by several researchers showing significant improvement in nickel ature (past 10 years) produced no meaningful results which therefore
extraction with high recovery from laterite ores (Jiang et al., 2013; Li creates the opportunity for further research in the field of sodium-based
et al., 2012a; Li et al., 2012b). Other applications of Na2SO4 include FGD additives and their effects.
textile production industry (as a catalyst for colour fixation and ab­ For dry scrubbing purposes, NaHCO3 is generally pulverized into fine
sorption of dyes on fabrics), in households and commercial detergent particles (~ 20 μm) to increase the absorbent specific surface area and
and soap manufacturing process (as a pH neutralizer and to maintain the subsequently enhance SO2 removal efficiency. However, the storage and
viscosity of the finished product), and in pharmaceutical industry (as transport of this ultrafine reagent is characterized by several operational
laxative and food additive) (Global Markets Insights, 2021). On the issues (e.g., bridging and or blocking) which requires the costly addition
other hand, Na2SO3 can be used as a depressant in the processing of of a dispersing agent to prevent occurrence. As a means of investigating
complex ores by selective floatation in media with high ion content. Due possible solutions to overcome these issues Hwang et al. (2021) evalu­
to its oxidising and reducing properties, Na2SO3 has been used as a re­ ated the pollutant removal efficiency of porous NaHCO3 produced via
agent to depress pyrite in copper sulfides (Molaei et al., 2018), copper thermal vacuum treatment from coarse NaHCO3. Although tested for
depressant in molybdenite ores (Suyantara et al., 2021) and arsenopy­ HCl removal from flue gas, the optimum conditions for enhanced spe­
rite from gold ores (Kydros et al., 1993). cific surface area and effective HCl removal was determined to be 180
For most sorbent injection processes, FGD products and spent ◦
C, 30 kPa and a residence time of 1 h. Absorbent activation using me­
absorbent is normally collected with the fly ash fraction in the power chanical and thermal techniques was also investigated by Walawska
station’s main dust collection device (i.e., ESP or FFP). Consequently, for et al. (2014). In their study, NaHCO3 and Na2CO3 was subjected to
sodium-based FGD this solid waste mixture has little economic value (e. various grinding techniques (fluid bed opposed jet mill, fine impact mill
g., cement manufacturing) if the sodium content is too high and may and electromagnetic mill) followed by thermal activation and subse­
also result in higher disposal costs due to the higher solubility of sodium- quent evaluation of the SO2 removal efficiency. A comparison between
based FGD products (Carpenter, 2012). In addition, significant amounts the different milling techniques revealed that electromagnetic grinding
of unreacted sorbent might be lost which may subsequently result in produced an absorbent with the highest specific surface area and SO2
higher operational costs if not recovered or regenerated. Collecting fly removal efficiency.
ash before the sorbent injection location presents one practical solution A 19.8 % improvement on SO2 removal efficiency was observed
but would require a secondary dust collection device to recover the between the SO2 removal efficiencies of the finest milled NaHCO3 (6.7
spent sorbent and FGD by-product. This could increase capital expen­ μm) and ungrounded absorbent. Furthermore, thermal activation at
diture due to the installation of additional unit operations. Salmenoja 300◦ C showed that absorbent surface area could be increased by at least
et al. (2020) has, however, proposed a process of directly recovering a factor 13 (from 0.6 to 8.6 m2/g). Although the preceding studies have
Na2SO4 in fly ash (patented for a recovery boiler of a chemical pulp mill) been more fundamental in nature, technology optimization efforts and
which can be considered as a possible alternative for sorbent and FGD commercial trials have also proven to be useful in improving dry FGD.
recovery from a fly ash mixture. Several other methods for the recovery One such an example is the SORBMIX technology offered by Mobotec
and regeneration of sodium compounds from dry FGD wastes have also which, in contrast to a conventional injection lance protruding into the
been reviewed and proposed by EPRI (1981) and Detourney & Coustry flue gas duct for mixing, uses a boosted air jet with high velocity to
(2010) and include amongst others the Aqueous Carbonate Process effectively mix the dry FGD sorbent into the flue gas (Liu, 2015). The
(ACP), carbonation, evaporative crystallization, acidification with technology also has the following added advantages: more operational
H2SO4, dissolution accompanied by electrodialysis and the Ampot pro­ flexibility by being able to handle various load ranges and reduced
cess. The presence of similar key industries (e.g., glass manufacturing, maintenance and operational issues when compared to conventional
paper and pulp, textiles, mining and minerals beneficiation) within the lances. An industrial scale SORBMIX trial (~ 774 kg/kWh inlet SO2)
South African economic sector, combined with the potential of the conducted over a 12-hour period, at various stoichiometric ratios (Ca/S
aforementioned FGD by-product recovery and regeneration schemes, of between 2.5 and 7.5) indicated, on average, an 8 to 10 % improve­
therefore creates the opportunity for potential markets to be explored ment in SO2 reduction made by SORBMIX as compared to the standard
for utilizing and recovery of waste products from semi-dry and dry, injection lance setup.

178
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

6.8. Optimization of abatement performance using modelling techniques location and the number of injection points as a means of improving
absorbent utilization and SO2 sulphation efficiency. A comprehensive
Over the past two decades several papers have focused their atten­ review of the application of CFD for semi-dry flue gas desulphurization
tion more on the use of both fundamental and advanced modelling has also been recently provided by Lerotholi et al. (2022). Although
techniques as a means of not only gaining insight into the mechanistic computationally intensive, these studies offer a cost-effective means to
processes of desulphurization but also to optimize desulphurization ef­ benchmark and screen FGD technologies to be installed in new coal-fired
ficiency through more reliable and accurate FGD system design and power stations or retrofitted to older facilities.
simulation. Kinetic studies of the semi-dry and or dry sulphation reac­
tion of sodium-based sorbents are relatively rare in literature with only a 7. Research gaps and significance for the south african power
few publications cited (Bibalani and Ebrahim, 2022; Keener and Khang, industry
1993; Kimura and Smith, 1987; Wu et al., 2004). One example is the
comprehensive study performed by Bibalani & Ebrahim (2022) on The generation of additional taxable CO2 as well as the increase in
Na2CO3 at various operating temperatures and SO2 concentrations (i.e., the life-cycle cost for waste disposal (gypsum by-product disposal) are
100 – 250◦ C and 0.13 – 1.12 vol. % respectively). Modelling studies of amongst the factors which could result in the sustained net negative
this nature normally consider the order of reaction to be either fractional operational expenditure of wet FGD units currently installed in South
(e.g., 0.5) or 1st order, whilst the preferred kinetic models employed Africa’s power industry (Bagus et al., 2018; Strickroth et al., 2020;
include the random pore model, shrinking core model, modified grain Gruenewaldt, 2013; Campbell, 2015; Vosloo, 2018; Koralegedara et al.,
model or the volume reaction model. In their study, the authors devel­ 2019). Another threat to the successful implementation and sustainable
oped and evaluated a sophisticated random pore model, which accounts operation of wet FGD technology is the scarcity, quality, and trans­
for the whole pore size distribution (the reader is referred directly to the portation cost of limestone in South Africa (Strickroth et al., 2020). The
paper for more details on the governing equations). This is a substantial Kusile Power Station is currently supplied with limestone from the
improvement on the modelling approaches of Keener and Khang (1993) Idwala Industrial Holdings (Pty) Ltd quarry in Danielskuil in the
and Kimura & Smith (1987) who failed to consider absorbent specific Northern Cape (more than 700 km away from the Kusile Power Station).
surface area as well as diffusional resistances during the sulphation re­ The challenge is, therefore, to evaluate and select economically viable
action. For the experimental temperature range investigated, Bibalani & and sustainable SO2 abatement technologies for further roll-out to South
Ebrahim (2022) determined the activation energy to be 22.5 kJ.mol–1, Africa’s ageing coal-fired power industry. It is also not advisable that a
the diffusion coefficient (DP) of SO2 through the product layer to range “one size fits all” approach is followed, but rather a life cycle and
between 12×10–19 and 15×10–18 m2.s–1 and the surface rate constant risk-based cost benefit approach be applied taking into consideration the
(kS) to range between 8.8 × 10–6 and 9.1 × 10–5 m.s–1. A comparative expected remaining operational life and emissions apportionment of
overview of rate constants and applied kinetic models for various other each power station in the fleet. The preceding paragraphs of this paper
absorbents are also included in the works of Bibalani & Ebrahim (2022). have listed several advantages (e.g., higher cost of sodium as compared
To further modelling efforts, Wu et al. (2004) developed and eval­ to lime/limestone, but higher utilization) of utilizing sodium-based ab­
uated a mathematical model to simulate dry sorbent (NaHCO3 in this sorbents over the more commonly used calcium-based absorbents. A
case) duct injection across a fabric filter. The developed model employs number of available processing technologies (several small-scale and a
parallel reaction kinetics and assumes that the process is composed of an few full-scale commercial applications in practice) also makes
initial stage where absorbent interacts with SO2 in the duct followed by sodium-based FGD an attractive option to the utility industry (Srinivasn,
sorbent collection and further reaction on the fabric filter. Various pilot 2004). Literature and research performed on the emissions reduction
and demonstration scale operational results were used to validate and performance and economic feasibility of implementing sodium-based
assess the model against typical operational parameters (temperature, FGD for the power generation industry in South Africa is relatively
particle size, residence time, NSR, SO2 concentration and decomposition scarce and almost non-existing. Based on the findings made during this
time). From the model validation and sensitivity analysis it was review the main areas of further research therefore include:
concluded that the SO2 removal efficiency within the duct is negligibly
small for most operational conditions, whilst significant SO2 abatement ■ An evaluation of the general availability, market (potential sup­
occurs across the fabric filter cake (Wu et al., 2004). The most notable pliers), location, and associated costs of both natural occurring and
findings made during this modelling investigation was the effect of synthetic sodium-based absorbents (e.g., trona, nahcolite, NaHCO3)
particle size and flue gas / filtration temperature on SO2 removal effi­ in South Africa. As an example, it is known that Botash SA (Pty) Ltd
ciency. Accordingly, 10 μm absorbent particles showed a monotonic (a partner / subsidiary of Idwala Industrial Holdings (Pty) Ltd) is an
decrease in removal efficiency (from approximately 69 to 35 %) with an industrial supplier of both NaHCO3 (derived from nahcolite) and
increase in flue gas temperature, whilst for the other particle sizes (20, soda ash (Na2CO3). Botash SA operates as an agent for Botswana Ash
50 and 70 μm) the removal efficiency first increased then decreased. The SA which is the supplier of choice for natural sodium products in
flue gas temperature associated with each local maxima for SO2 removal Africa. The Botash SA operational facility is conveniently located in
also decreased almost linearly with an increase in flue gas temperature. Alrode, Alberton where its Natalspruit depot comprises of both a
For the fabric filter bag modelling analysis (at a given particle size), it 24,000 Mt bulk storage (3 silos) facility and a 2000 m2 warehouse
was shown that although the optimum filtration temperature (local facility (Botash SA, 2021).
maxima around 127◦ C) for SO2 removal does not change significantly ■ As sodium-based absorbents are generally known to be more
with a change in SO2 concentration, the maximum SO2 removal effi­ expensive than their calcium-based counterparts, further effort is
ciency does however increase with an increase in SO2 concentration (Wu also required to investigate the feasibility of repurposing sodium-
et al., 2004). bearing waste streams (e.g., industrial brines, caustic scrubber
Other more advanced modelling approaches have also started to gain streams, etc.) as low-cost or cost-effective alternatives. This approach
popularity, with a few studies employing response surface methodolo­ also resonates well with the current global effort to stimulate circular
gies (RSM), artificial neural networks (ANN) and computational fluid economies as a means of eliminating waste and pollution through the
dynamics (CFD) (Heiszwolf and Nyssen, 2016; Makomere et al., 2023). circulation of raw materials and products (United Nations Industrial
The purpose of these advanced modelling studies, although historically Development Organization (UNIDO), 2023).
more focused on calcium-based sorbents, is not only to evaluate the ■ Assess the physiochemical characteristics (e.g., specific surface area,
statistical significance of several operational parameters but also to chemical composition / purity, particle size distribution, etc.) of
understand two-phase flow phenomena inside the duct and optimize the locally sourced natural occurring, waste bearing and or synthetic

179
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

sodium-based absorbents (inorganic and organometallic) in South Semi-dry scrubbing systems (sodium-based or calcium-based),
Africa and benchmark against published information on sodium- namely spray dry scrubbers (SDSs) and circulating dry scrubbers
based absorbents used for FGD studies internationally. A detailed (CDSs), consume about 60 % less water than wet scrubbers. Their ad­
understanding of each absorbents physiochemical comparison could vantages include lower investment costs (than a similar sized wet
provide useful to draw meaningful conclusions with respect to the scrubber), limited production of wastewater, dry by-products, lower
absorbent’s desulphurization performance under different operating parasitic power consumption, a smaller footprint (which may be easier
conditions. for retrofit applications), and additionally capture over 95 % of the SO3,
■ Evaluate the effect of different operational variables (temperature, HCl, HF and oxidised mercury. It is important to highlight that general
SO2 concentration, particle size, NSR, etc.) on the SO2 removal ef­ literature and review publications on semi-dry processes that employs
ficiency of the locally acquired sodium-based absorbents during sodium-based absorbents are scarce.
semi-dry and dry FGD tests and benchmark against published results Processes that commonly utilizes sodium-based absorbents are sor­
on sodium-based FGD. Here the application of statistical sound bent injection processes. They have zero water usage or a minimal
parametric and optimization studies (e.g., RSM) could play an amount in cases where the flue gas is humidified to improve perfor­
important role in evaluating multi-variate effects during desulpha­ mance. They have been reported to be simple to install and operate, easy
tion using sodium-based sorbents (extracted from waste, natural to retrofit with their small space requirements, and produce no waste­
occurring and synthetic). Of particular importance is the inclusion of water. In addition, they also remove, to some extent, HCl, HF and
multipollutant and additive studies to assess the individual pollutant mercury species present in the flue gas. By-products are dry and there­
removal efficiencies in typical, simulated flue gas mixtures (mixtures fore easier to handle and manage as compared to the wet by-products
of SO2, CO2, Hg, etc.). This, to the authors knowledge, is a significant from wet scrubbers. Their capital cost and energy consumption of the
gap in the current body of knowledge. sorbent injection systems have also been reported to be significantly
■ Kinetic studies of the semi-dry and or dry sulphation reaction of lower than the semi-dry and wet FGD processes. Parasitic power con­
sodium-based sorbents are relatively rare in literature and therefore, sumption has been reported to be around 0.2 % of electric capacity.
presents an ideal opportunity for further contribution to the body of The drawbacks of dry sorbent inject includes higher operating costs
knowledge. A thorough kinetic and modelling study on the SO2 due to the high price and consumption of the sorbent. The bulk of the
removal efficiency of various locally sourced, sodium-based absor­ operating cost is the price of the sorbent. Trona, though, is more difficult
bents is imperative to gain further insight in the design and selection to handle than hydrated lime due to its small particle size (28 µm), its
of suitable semi-dry and dry FGD technologies to be employed in the cohesiveness, and affinity for moisture. In addition, care must be taken
South African power generation sector. These fundamental studies to avoid the formation of liquid NaHSO3 which can foul the downstream
can be further supplemented by more advanced modelling ap­ equipment and ductwork. Solid by-products are of less economic value
proaches such as CFD. and pose danger of leaching when landfilled. The sorbent injection
■ Perform a techno-economic study to evaluate the economic feasi­ systems are best suited for use in small (300 MWe) power plants that
bility (and to compare to limestone wet FGD) of retrofitting existing utilise low to medium sulphur coals, and where only a moderate SO2
power stations with sodium-based FGD technologies. A similar removal efficiency is required.
approach can be followed as was done for the technology selection Poor sorbent utilisation and SO2 removal efficiency (as compared to
studies performed by Bagus et al. (2018) and Harris et al. (2014). A wet FGD technologies) of the absorbents are also major setbacks,
more recent study which focused on a first order technoeconomic although this can be improved through the use of additives and sorbent
comparison between the wet FGD and Sulfacid® SO2 abatement activation strategies. The efficiency of sodium-based FGD process has
technologies can also provide useful information in this regard been reported to be influenced by various factors, e.g., physicochemical
(Strickroth et al., 2020). The proposed techno-economic study properties of the absorbents, sodium to sulphur stoichiometric ratio,
should not only be limited solely to sodium-based absorbents and sorbent injection schedule, injection location, operating temperature,
technologies but should also take into consideration hybrid and sulphur content in coal, inlet SO2 concentration, sorbent resident time in
regenerative technologies (e.g., dual-alkali process). Although not the duct, baghouse air-to-cloth ratio, ESP operation, etc. To conclude,
done for the utility sector, a publication on the operation of a high removal efficiencies of SO2, SO3, HCl and HF with trona, nahcolite
concentrated model dual-alkali scrubber plant at the Lonmin smelter and sodium bicarbonate have been demonstrated at many power plants
has provided useful insight on the rationale of technology selection over the years. Its low capital cost makes dry sorbent injection even
and the plant performance thus far (Bezuidenhout et al., 2012). more attractive in today’s difficult economic environment. Further
research work is also recommended to strengthen especially the South
8. Conclusions African knowledge base with respect to the sodium-based sorbent
market, the physiochemical properties and SO2 reduction potential of
SOx emission control technologies with varying water demands are locally acquired sodium-based absorbents (both in diluted SO2 but also
available globally for use in response to stringent emissions regulations. simulated flue gas streams), reaction rate modelling studies as well as its
Low-water to no-water technologies are attractive to arid and drought techno-economic feasibility for the South African utility industry.
vulnerable regions/countries like South Africa. There is, therefore, a Additionally, it is recommended to consider the reusability of the
need for FGD systems offering low water consumption, high SO2 and SO3 generated waste products from sodium-based dry FGD.
removal efficiencies, high reliability, low capital, operating and main­
tenance costs, low auxiliary power consumption and marketable FGD Funding
by-products. A review of publications on technologies that employ
sodium-based absorbents are scarce, thus, necessitating a need to eval­ The authors declare that no funds, grants, or other support were
uate this aspect in this paper. received during the preparation of this manuscript.
Wet scrubbers (sodium-based or calcium-based), result in up to 10 –
15 % of the evaporative water losses of water-cooled coal-fired power Author contributions
plants. In facilities that employ dry/air cooling, the water usage due to
wet scrubber can reach in the order of 40 to 70 %. Sodium-based wet D. Mchabe conceived the original idea for the paper (on request from
scrubber technologies available in literature includes soda ash wet Eskom and as part of its EPPEI Task Force program) and prepared the
scrubber, dual-alkali scrubber, Wellman-Lord process, Airborne process first draft (i.e., overview of processes, reaction chemistry and effect of
etc. process parameters). B.B. Hattingh and L. Koech further developed the

180
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

manuscript with the support from H.L. Rutto and H.W.J.P. Neomagus to Carson, J.R., 1980. Removal of Sulphur Dioxide and Nitric Oxide from a Flue Gas Stream
by Two Sodium Alkalis of Various sizes. Dissertation. University of Tennessee –
include aspects related to other non-common Na-based sorbents, reac­
Knoxville.
tion modelling, absorbent benefits and disadvantages, cost consider­ Chang, C.S., Rochelle, G.T., 1981. SO2 absorption into aqueous solutions. AIChE J. 27,
ations, gaps for further research, etc. All authors reviewed and approved 292–298. https://doi.org/10.1002/aic.690270217.
the final manuscript before submission to the journal. Chien, T.W., Chu, H., 2000. Removal of SO2 and NO from flue gas by wet scrubbing using
an aqueous NaClO2 solution. J. Hazard Mater. 80, 43–57. https://doi.org/10.1016/
S0304-3894(00)00274-0.
Data and availability Chien, T.W., Chu, H., Hsueh, H.T., 2003. Kinetic study on absorption of SO2 and NOx
with acidic NaClO2 solutions using the spraying column. Int. J. Environ. Eng. 129,
967–974. https://doi.org/10.1061/(ASCE)0733-9372(2003)129:11(967).
Data sharing is not applicable. The authors confirm that the data Chu, H., Chien, T.W., Li, S.Y., 2001. Simultaneous absorption of SO2 and NO from flue
analysed in this study were a re-analysis of existing data, which are gas with KMnO4/NaOH solutions. Sci. Total Environ. 275, 127–135. https://doi.org/
10.1016/S0048-9697(00)00860-3.
openly available at locations cited in the reference section. Commission, E., 2006. Technical report. Seville, Spain, European Commission. Joint
Research Centre, Institute for Prospective Technological Studies, European IPPC
Bureau, p. 618.
Declaration of competing interest Cordoba, P., 2015. Status of flue gas desulphurisation (FGD) systems from coal-’ fired
power plants: overview of the physic-chemical control processes of wet limestone
FGDs. Fuel 144, 274–286. https://doi.org/10.1016/j.fuel.2014.12.065.
The authors declare that they have no known competing financial Cremer, M.A., Wang, D.H., Senior, C.L., 2008. Testing and model-based optimization of
interests or personal relationships that could have appeared to influence SO2 removal with trona in coal fired utility boilers. In: ‘Proceedings of the 7th Power
Plant Air Pollutant Control “Mega” Symposium’, 2, pp. 837–850.
the work reported in this paper. Dahiya, S., Anhäuser, A., Farrow, A., Thieriot, H., Kumar, A., Myllyvirta, L., 2020. Global
SO2 Emission Hotspot Database. Centre for Research on Energy and Clean Air and
Greenpeace India, Delhi, p. 48.
Acknowledgements
Dantuluri, S.R., 1988. Dissertation. University of Tennessee Knoxville.
Darmastaedter, E., 1990. Sodium Bicarbonate injection: A small Plant SO2/NOx Option.
The authors wish to thank the Eskom Power Plant Engineering Power Engineering.
Institute (EPPEI) for their financial support and assistance in establish­ Davis, W., Keener, T., Lavely, L., 1978. Research on the removal of SO2 by additive
injection techniques on a stoker-fired boiler. In: ‘71st. Annual Meeting of the Air
ing a longstanding collaborative relationship between the North-West Pollution Control Association’. June 25-30, 1978.
University and the Vaal University of Technology (VUT) within the Detourney, J.P., Coustry, F., 2010. Process for producing sodium bicarbonate for flue gas
EPPEI Emissions Control Specialisation Centre (EC SC). desulphurization. (Patent: US 2010/0290967 A1). https://patents.google.com/pat
ent/US20100290967A1/en. Date of access: 12 February 2023.
Department of Forestry, Fisheries & Environment (DFFE), 2004. National environmental
References management: air quality act, No. 39, Pretoria, government gazette. https://www.
gov.za/documents/national-environment-management-air-quality-act.
Department of Forestry, Fisheries & Environment (DFFE), 2019. Listed Activities and
Albin, D.W., 1986. Full scale demonstration of dry sodium injection flue gas
Associated Minimum Emissions Standards Identified in Terms of Section 21 of the
desulfurization at city of colorado springs ray D. nixon power plant. In: Proceedings
National Environmental Management: Air Quality act, No. 39, Pretoria, Government
of the Joint EPRI/EPA Symposium on Dry SO2 and Simultaneous SO2/N0x Control
Gazette. https://www.gov.za/documents/national-environmental-management-air-
Technologies, Volume 2, EPRI CS4966, December 1986.
quality-act-list-activities-associated-minimum.
Bagus, M., van Wyk, L., Chang, D., 2018. Medupi flue gas desulphurisation -technology
Donahue, B., Oxley, J., Buie, C., 1980. Technical report. Construction Engineering
selection report. Technical report. Rev. 2. Johannesburg.
Research Lab (Army) Champaign IL. https://apps.dtic.mil/sti/citations/
Bajpai, P., 2018. Brief description of the pulp and papermaking process. Biotechnology
ADA092132.
for Pulp and Paper Processing. Springer, pp. 9–26. https://doi.org/10.1007/978-
Dzhonova-Atanasova, D., Razkazova-Velkova, E., Ljutzkanov, L., Kolev, N., Kolev, D.,
981-10-7853-8_2.
2013. Energy efficient SO2 removal from flue gases using the method of Wellman-
Balland, J.dependentP., Thome-Kozmiensky, K.J., Thiel, S., Thome-Kozmiensky, E.,
Lord. J. Chem. Technol. Metall. 48 (5), 457–464.
Winter, F., 2017. Ready-to-use sodium bicarbonate in flue gas treatment –
Electric Power Research Institute (EPRI), 1981. Recovery, utilization, and disposal of
experience feedback. In: Juchelkova, D. (Ed.), Ready-to-use sodium bicarbonate in
solid by-products generated by dry flue gas desulphurization systems: state of the art
flue gas treatment – experience feedback. Waste Management. Vol. 7. Waste-to-Ergy
and research needs. Technical Report No.: EPRI CS-1765, Project 1260-16. 164p.
449–462. https://www.vivis.de/2017/09/waste-management-volume-7/405/.
https://www.osti.gov/biblio/6508471.
Benson, L.B., Smith, K.J., Roden, R.A., Loch, E., Potts, J., 2003. Control of sulphur
Electric Power Research Institute (EPRI), 2011. Dry sorbent injection for control of SO2
dioxide and sulphur trioxide using by-product of a magnesium-enhanced lime FGD
and acid gas hazardous air pollutions: full-scale testing at Battle River Station Unit 3.
system, in ‘Institute of Clean Air Companies. Forum Fam Plan West Hemisph 3,
Palo Alto, CA, Technical Report No.: 1024747. https://www.epri.com/research/pr
14–15.
oducts/1024747.
Bezuidenhout, G.A., Davis, J., van Beek, B., Eksteen, J.J., 2012. Operation of a
Eskom, 2018. Eskom Report No: 474–10175.
concentrated mode dual-alkali scrubber plant at the Lonmin smelter. J. South Afr.
Eskom, 2019. Eskom Report No: 348-9953598, p. 27.
Inst. Min. Metall. 112, 657–665.
Garg, A., Kapshe, M., Shukla, P., Ghosh, D., 2002. Large point source (LPS) emissions
Bibalani, I.O., Ebrahim, H.A., 2022. Comparison of sulfur dioxide removal reactions
from India: regional and sectoral analysis. Atmos. Environ. 36, 213–224. https://doi.
kinetics by Na2CO3 and other different sorbents from coal-fired power plants. Chem.
org/10.1016/S1352-2310(01)00439-3.
Biochem. Eng. Q. 36 (3), 195–205. https://doi.org/10.15255/CABEQ.2022.2069.
França, Ĭ.W.L., Cartaxo, S.J.M., Batos-Neto, M., Gonçalves, L.R.B., Fernandes, F.A.N.,
Bland, V.V., Martin, C.E., 1990. Full-scale Demonstration of Additives for NO2 Reduction
2020. Effect of additives to improve calcium-based sorbents in semi-dry flue gas
With Dry Sodium FGD, Technical Report No.: EPRI GS-6852. Electric Power
desulphurization. Emiss. Control. Sci. Technol. 6, 105–112. https://doi.org/
Research Institute, Palo Alto, CA. June 1990.
10.1007/s40825-020-00156-0.
Bloomberg (2022) Flue Gas Desulfurization (FGD) Market size Worth $ 24.7 Billion,
Global Market Insights, 2021. Sodium Sulphate Market Size and Share | Industry
Globally, By 2028 At 4.3% CAGR: Verified Market Research®. 24 January. URL:
Statistics 2021 - 2027 (Industry Report No. GMI1750), Bulk & Speciality Chemicals.
https://www.bloomberg.com/press-releases/2022-01-24/flue-gas-desulfurization-
Global Market Insights Inc., DelawareUSA, 19975. https://www.gminsights.com/ind
fgd-market-size-worth-24-7-billion-globally-by-2028-at-4-3-cagr-verified-market-re
ustry-analysis/sodium-sulphate-market.
search.
Gruenewaldt, R.G., 2013. The co-disposal of gypsum with ash at Kusile power station: air
Blythe, G.M., 2004. Furnace injection of alkaline absorbents for sulfuric acid removal.
quality assessment. https://www.eskom.co.za/OurCompany/SustainableDevel
Technical report. URS Corporation 9400 Amberglen, p. 78729. Boulevard Austin,
opment/EnvironmentalImpactAssessments/Kusile60Yr/Documents/H-Kusile10yr
Texas.
AshDisposalAQA.
Botash, S.A., 2021. Company website of botswana Ash South Africa (Pty) Ltd. URL: https:
Gubb, A., 2009. National Environmental management: Air quality Act (39 of 2004).
//www.botash.co.za/.
Wildlife and Environmental Society of South Africa. http://www.enviropaedia.com/
Campbell, C. 2015. Waste assessment of ash and flue gas desulphurisation wastes for the
topic/default.Php.
medupi power station, Rev. 2. Eskom, Midrand. http://www.eskom.co.za/OurCom
Harris, D., Hart, A., Odendaal, D., 2014. Medupi flue gas desulphurisation -technology
pany/SustainableDevelopment/EnvironmentalImpactAssessments/medupi/DEIR%
selection study report, Technical report, Rev. 1. Johannesburg.
20Appendicies/Appendix%20G-2_Waste%20Assessment%20Report.pdf.
Heiszwolf, J.J., Nyssen, O., 2016. Improving Dry Sorbent Injection With Computational
Campobenedetto, E.J., Silva, A.A., 2011. Low-cost multi-pollutant control solution
Fluid dynamics: A cost Effective Emission Control technology, Emission Control.
demonstrations. In: ‘In: 104th Air & Waste Management Association Annual
NPT Process Technology, p. 3. https://www.sorbacal.com/sites/default/files
Conference and Exhibition, Orlando, FL, USA, 21-24 Jun 2011, pp. 766–774.
/downloadcenter/heiszwolf_nyssen_-_2016_-_improving_dry_sorbent_injection_with_
Carpenter, A.M., 2012. Low water FGD technologies, Vol December (2012) CCC/210
computational_fluid_dynamics.pdf. Date of access: 2 February 2023.
ISBN 978-92-9029-530-3.
Carpenter, A.M., 2013. Advances in multi-pollutant control. IEA Clean Coal Centre 11.
CCC/227 ISBN 978-92-9029-547-1.

181
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

Hooper, R.G., Bland, V., 1985. Combined SO2 and NOx removal from flue gas by injection Maziuk, J., 2007. Dry injection of trona sorbent to mitigate SOx, NOx and Hg emissions.
of sodium sorbents. In: Proceedings of the 3rd EPRI Conference on Fabric Filter In: ‘In: International conference on air quality VI, proceedings, Arlington, VA, USA,
Technology for Coal-Fired Power Plants, November 1985. 24-27 Sep 2007. University of North Dakota, Energy and Environment Research
Hooper, R., 1988. Full scale dry sodium injection demonstration meets SO2 BACT. Power Centre, Grand Forks, ND, USA, p. 9, 2007.
Eng. June https://www.osti.gov/etdeweb/biblio/7182901. Mchabe, D., 2020. PhD thesis. North-West University (South Africa.
Hwang, I.dependentH., Matsuo, T., Matusto, T., Tojo, Y., Sameshima, R., 2021. Dry Miller, I., 1979. Dry scrubbing looms large in SO2 clean-up plans. https://www.osti.
scrubbing of municipal solid waste incineration flue gas using porous sodium gov/biblio/6753481.
carbonate produced via vacuum thermal treatment of sodium bicarbonate. J. Mater. Mocek, K., Lippert, E., Erdos, E., 1983. Reactivity of the solid sodium carbon-¨ate towards
Cycles Waste. Manag. 23, 1609–1616. https://doi.org/10.1007/s10163-021-01241- the gaseous hydrogen chloride and the sulphur dioxide. Collect. Czech Chem.
4. Commun. 48, 3500–3507. https://doi.org/10.1135/cccc19833500.
Jacobs, 2021. Sulphur oxide (SOx) emissions and reduction options for vales point power Molaei, N., Hoseinian, F.S., Rezai, B., 2018. A study on the effect of active pyrite on
station, memorandum, Project no.: IS389000, 11 p. URL: https://hdp-au-prod- flotation of porphyry copper ores. Physicochem. Probl. Miner. Process. 54 (3),
app-nswepa-yoursay-files.s3.ap-southeast-2.amazonaws.com/4016/3460/3627/202 922–933. https://doi.org/10.5277/ppmp1894.
1.10.8_-_Delta_-_4_Additional_Information_-_Jacobs_Vales_Point_SOx_PRS_Memo.pdF. Mortson, M., Telesz, R.W., 2001. Flue gas desulfurization using recycled sodium
Jiang, M., Sun, T., Liu, Z., Kou, J., Liu, N., Zhang, S., 2013. Mechanism of sodium sulfate bicarbonate, in ‘The US EPA/DOE/EPRI combined power plant air pollutant control.
in promoting selective reduction of nickel laterite ore during reduction roasting the mega symposium’, Citeseer. https://www.yumpu.com/en/document/view/
process. Int. J. Miner. Process 123, 32–38. https://doi.org/10.1016/j. 11838366/flue-gas-desulfurization-using-recycled-sodium-bicarbonate.
minpro.2013.04.005. Mortson, M., Xia, Q., 2006. Advanced pollution control - the airborne process™ and its
Johnson, D.W., Ehrnschwender, M.S., Seidman, L., 2005. Airborne process™” benefits to China. In: ‘International Forum on Innovation and Environment, Beijing,
advancement of multi-pollutant emissions control technology and by-product China, 23-24 Oct 2006., Airborne.
utilization. In: ‘Electric Power Conference, Chicago’. Muzio, L., Arand, J., 1981. Bench-scale Study of the Dry Removal of SO2 With Nahcolite
Joo, S.H., Bhatti, U.H., Park, H.J., Jeong, D.H., Baek, I.H., Nam, S.C., Lee, K.B., 2020. and trona. EPRI CS-1744. Electric Power Research Institute, Palo Alto California.
Selective removal of SO2 from coal-fired flue gas by alkaline solvents using a Muzio, L.J., Often, G.R., 1987. Assessment of dry sorbent emission control technologies
membrane contactor. Chem. Eng. Process Process Intensification. 147, 107772 part I. fundamental processes. JAPCA 37, 642–654. https://doi.org/10.1080/
https://doi.org/10.1016/j.cep.2019.107772. 08940630.1987.10466251.
Keener, T.C., Davis, W.T., 1984. Study of the reaction of SO2 with NaHCO3 and Na2CO3. Muzio, L., Sonnichsen, T., 1984. Dry SO2 Particulate Removal for Coal-Fired Boilers.
JAPCA 34, 651–654. https://doi.org/10.1080/00022470.1984.10465793. Electric Power Research Institute.
Koech, L., Rutto, H., Lerotholi, L., Everson, R.C., Neomagus, H., Branken, D., Often, G., McElroy, M., Muzio, L., 1987. Assessment of dry sorbent emission control
Moganelwa, A., 2021. Spray drying absorption for desulphurization: a review of technologies part I. Appl., JAPCA. 37, 968–980. https://doi.org/10.1080/
recent developments. Clean Technol., Environ., Policy 23, 1665–1686. https://doi. 08940630.1987.10466288.
org/10.1007/s10098-021-02066-3. Pandey, R.A., Biswas, R., Chakrabarti, T., Devotta, S., 2005. Flue gas desulfurization:
Keener, T.C., Khang, S.J., 1993. Kinetics of the sodium bicarbonate—sulfur dioxide physicochemical and biotechnological approaches. Crit. Rev. Environ. Sci. Technol.
reaction. Chem. Eng. Sci. 48 (16), 2859–2865. 35, 571–622. https://doi.org/10.1080/10643380500326374.
Kimura, S., Smith, J.M., 1987. Kinetics of the sodium carbonate–sulfur dioxide reaction. Park, H.W., Park, D.W., 2017. Removal kinetics for gaseous NO and SO2 by an aqueous
AIChE J. 33, 1522–1532. https://doi.org/10.1002/aic.690330912. NaClO2 solution mist in a wet electrostatic precipitator. Environ. Technol. 38,
Kong, Y., Balland, J.P., 2011. Effective removal of HCl and SO2 with dry injection of 835–843. https://doi.org/10.1080/09593330.2016.1213770.
sodium bicarbonate or trona. In: ‘North American Waste-to-Energy Conference’, Park, H.S., Kang, D., Kang, J.H., Kim, K., Kim, J., Song, H., 2021. Selective sulphur
54570, pp. 195–199. https://doi.org/10.1115/NAWTEC19-5408. dioxide absorption from simulated flue gas using various aqueous alkali solutions in
Kong, Y., Wood, M., 2011. Dry injection of sodium absorbents for air pollution control. a polypropylene hollow fibre membrane contactor: removal efficiency and use of
Period Am. Acad. Environ. Eng. 47, 20–23. sulphur dioxide. MDPI: Int. J. Environ. Res. Public Health 18, 597–612. https://doi.
Kong, Y., Wood, M.D., 2010. Dry injection of trona for SO3 control. Power 154, 114. org/10.3390/ijerph18020597.
–114. Pflughoeft-Hassett, D., Ladwid, K., Hassett, D., Dockter, B., Heebink, L., Eylands, K.,
Koralegedara, N.H., Pinto, P.X., Dionysiou, D.D., Al-Abed, S.R., 2019. Recent advances in Hoffarth, J., 2009. Characteristics and performance of fly ash from sodium sorbent
flue gas desulfurization gypsum processes and applications a review. J. Environ. scrubbing of SO3 emissions from coal-based power plants. In: ‘Proceedings of the
Manage. 251, 109572 https://doi.org/10.1016/j.jenvman.2019.109572. 2009 World of Coal Ash Conference, Lexington, KY.
Kostick, D., 1994. Soda ash in industrial minerals and rocks. Society for Mining, Pilat, M.J., Wilder, J.M., 2007. Pilot scale SO2 control by dry sodium bicarbonate
Metallurgy and Exploration, 6th ed., pp. 929–955 injection and an electrostatic precipitator. Environ. Prog. 26, 263–270. https://doi.
Kumar, L., Jana, S.K., 2021. Advances in absorbents and techniques used in wet and dry org/10.1002/ep.10212.
FGD: a critical review. Rev. Chem. Eng. 37 https://doi.org/10.1515/revce-2020- Poullikkas, A., 2015. Review of design, operating, and financial considerations in flue gas
0029. desulfurization systems. Energy Technol. Policy 2, 92–103. https://doi.org/
Kydros, K.A., Angelidis, T.N., Matis, K.A., 1993. Selective flotation of an auriferous bulk 10.1080/23317000.2015.1064794.
pyrite-arsenopyrite concentrate in presence of sodium sulphoxy-salts. Miner. Eng. 6, Pretorius, I., Piketh, S., Burger, R., Neomagus, H., 2015. A perspective on south african
1257–1264. https://doi.org/10.1016/0892-6875(93)90103-T. coal fired power station emissions. J. Energy South Afr. 26, 27–40.
Lerotholi, L., Everson, R.C., Koech, L., Neomagus, H.W.J.P., Rutto, H., Branken, D., Rahmani, F., Mowla, D., Karimi, G., Golkhar, A., Rahmatmand, B., 2015. SO2 removal
Hattingh, B.B., Sukdeo, P., 2022. Semi‑dry flue gas desulphurization in spray towers: from simulated flue gas using various aqueous solutions: absorption equilibria and
a critical review of applicable models for computational fluid dynamics analysis. operational data in a packed column. Sep. Purif. Technol. 153, 162–169. https://doi.
Clean Technol. Environ. Policy 24, 2011–2060. https://doi.org/10.1007/s10098- org/10.1016/j.seppur.2014.10.028.
022-02308-y. Ritzenthaler, D., 2007. SO3 control: AEP Pioneers and Refines Trona Injection Process
Li, G., Shi, T., Rao, M., Jiang, T., Zhang, Y., 2012a. Beneficiation of nickeliferous laterite For SO3 Mitigation. Coal Power, pp. 1–6. In: https://www.powermag.com/SO3-con
by reduction roasting in the presence of sodium sulfate. Miner. Eng. 32, 19–26. trol-aep-pioneers-and-refines-trona-injection-process-for-SO3-mitigation/.
https://doi.org/10.1016/j.mineng.2012.03.012. Ritzenthaler, D., Hume, J., Moretti, A., Tonn, D., 2007. Evolution of AEP’s SO3
Li, G.H., Rao, M.J., Jiang, T., Shi, T.M., Huang, Q.Q., 2012b. Reduction roasting- mitigation. In: ‘International conference on air quality VI, proceedings, Arlington,
magnetic separation mechanisms of nickelferous laterite ore in presence of sodium VA, USA, 24-27 Sep 2007.
salts. Chin. J. Nonferrous Met. 22, 274–278. Ross, K., 2012. Eskom air quality strategy, 32–1143, Technical report, Eskom.
Liu, G., 2015. An Innovative mixing method to lower the cost of operating DSIand ACI Sahu, R., 2013. Technical report. Southern Alliance for Clean Energy.
systems. Power Eng. https://www.power-eng.com/emissions/air-pollution-control Salmenoja, K., Metsamuuronen, K., Joutsimo, M., Reilama, I., 2020. Method for treating
-equipment-services/an-innovative-mixing-method-to-lower-the-cost-of-operating fly ash of a recovery boiler. (Patent: US 2020/0181841 A1). https://patents.google.
-dsi-and-aci-systems/. com/patent/US20200181841A1/en.
Liang, Y., Harrison, D., Gupta, R., Green, D., McMichael, W., 2004. Carbon dioxide Sasaoka, E., Sada, N., Uddin, M., 1998. Preparation of macroporous lime from natural
capture using dry sodium-based absorbents. Energy Fuels 18, 569–575. https://doi. lime by swelling method with acetic acid for high-temperature desulfurization. Ind.
org/10.1021/ef030158f. Eng. Chem. Res. 37, 3943–3949. https://doi.org/10.1021/ie980137m.
Lu, J., Fu, Y., Wang, J., Chen, H., 2022. Study on the desulfurization performance of Satriana, M., 1981. New Developments in Flue Gas Desulfurization Technology. Noyes
calcium-based desulfurizer and NaHCO3 desulfurizer. Environ. Sci. Pollut. Res. Data Corporation, United States. https://www.osti.gov/biblio/5434138.
Preprint 19. https://doi.org/10.21203/rs.3.rs-1716413/v1. Schnelle Jr, K.B., Dunn, R.F., Ternes, M.E., 2015. Air Pollution Control Technology
Lisnic, R., Jinga, S., 2018. Study on current state and future trends of flue gas Handbook, 2nd ed. CRC Press. https://doi.org/10.1201/b19286.
desulphurization technologies: a review. Rev Romana Mater/Rom. J. Mater. 48, Schultes, M., 1998. Absorption of sulphur dioxide with sodium hydroxide solution in
83–90. packed columns. Chem. Eng. Technol. 21, 201–209. https://doi.org/10.1002/(SICI)
Madsen, JNHRDRGC, 2017. Effect of moisture conditioning and handling on leaching 1521-4125(199802)21:2<201::AID− CEAT201>3.0.CO;2-W.
and physical properties of sodium bicarbonate flue gas desulfurization materials. In: Shah, N., Teixeira, D., Muzio, L., 1978. Bench-scale evaluation of Dry alkalis for
‘2017 World of Coal Ash (WOCA) Conference in Lexington, KY -May 9-11,2017’. removing SO2 from boiler flue gases. In: ‘symposium on the transfer and utilization
Maina, P., 2012. Improvement of lime reactivity towards desulfurization by hydration of particulate control technology, Denver, CO’.
agents. Chem. Sci. Trans. 2 (1), 147–159. https://doi.org/10.7598/cst2013.233. Siagi, Z., Mbarawa, M., Mohamed, A., Lee, K., Dahlan, I., 2007. The effects of limestone
Makomere, R., Rutto, H., Koech, L., 2023. The assessment of response surface type on the sulphur capture of slaked lime. Fuel 86, 2660–2666. https://doi.org/
methodology (RSM) and artificial neural network (ANN) modelling in dry flue gas 10.1016/j.fuel.2007.03.034.
desulfurization at low temperatures. J. Environ. Sci. Health A 12. https://doi.org/
10.1080/10934529.2023.2174334.

182
D. Mchabe et al. South African Journal of Chemical Engineering 48 (2024) 167–183

Smith, R., Shiomoto, G., Muzio, L., Hills, L., Hunt, C., 1997. Integrated dry NO/SO2 Vehlow, J., 2015. Air pollution control systems in WtE units: an overview. Waste
emission control system sodium-based dry sorbent injection test report. DOE Manage. 37, 58–74. https://doi.org/10.1016/j.wasman.2014.05.025.
Contract Number DEFCC22: 91PC90550, NTIS DE97054372INZ. Vernon, J.L., Jones, T., 1993. Sulphur and coal. IEA Clean Coal Centre Technical. Report
Smith, S.J., van Aardenne, J., Klimont, Z., Andres, R.J., Volke, A., Arias, S.D., 2011. No. IEACR-57. 62 pp.
Anthropogenic sulphur dioxide emissions: 1850–2005. ACP 11, 1101–1116. https:// Vosloo, M., 2018. Application for variations to the existing waste management license
doi.org/10.5194/acp-11-1101-2011. (12/9/11/L50/5/R1) for the medupi power station ash disposal facility, limpopo
Srinivasn, R., 2004. PhD thesis. University of Cincinnati. province application for variation of the exiting waste management license (WML)
Srivastava, R.K., 2000. Controlling SO2 Emissions: A review of Technologies. US for the Medupi power station ash disposal facility. Eskom, Midrand. http://www.
Environmental Protection Agency, p. 113. Report no.:EPA/600/R-00/093. eskom.co.za/OurCompany/SustainableDevelopment/EnvironmentalImpactAsse
Srivastava, R.K., Jozewicz, W., 2001. Flue Gas Desulphurization: the state of the art. J Air ssments/MedupiADF_WML/Documents/12949-46-Rep-001-WM%20Variation%
Waste Manag. Assoc. 51, 1676–1688. https://doi.org/10.1080/ 20Tech%20Report-Rev1.pdf.
10473289.2001.10464387. Walawska, B., Szymanek, A., Pajdak, A., Nowak, M., 2014. Flue gas desulfurization by
Staudt, J.E., Bradley, M.J., Associates, 2011. Control technologies to reduce conventional mechanically and thermally activated sodium bicarbonate. Pol. J. Chem. Tech. 16,
and hazardous air pollutants from coal-fired power plants. In: Northeast States for 56–62. https://doi.org/10.2478/pjct-2014-0051.
Coordinated Air Use Management (NESCAUM), Boston, USA, p. 32. URL: www.ne Wang, J., Su, T., 2011. Leaching Behavior of Coal Combustion Products and the
scaum.org. Environmental Implication in Road Construction [April 2011] (No. NUTC R214). United
Stern, F.R., 1978. Bench-scale Study of Sulphur and Nitrogen Oxides Absorption by States. Dept. of Transportation. Research and Innovative Technology
Nahcolite and trona. Dissertation. University of North Dakota. Administration.
Strickroth, A., Schumacher, M., Hasse, G.W., Kgomo, I., 2020. Next-generation, Wilhelm, J.H., 2004. SBS injection fights off SO3. Power Eng. Int. 12 (12), 28–30. htt
affordable SO2 abatement for coal-fired power generation a comparison of ps://www.powerengineeringint.com/world-regions/north-america/sbs-injection
limestone-based wet flue gas desulphurization and Sulfacid® technologies for -fights-off-sosub3-sub/.
Medupi power station. J. South Afr. Inst. Min. Metall. 120, 581–590. Wu, C., Khang, S.dependentJ., Keener, T.C., Lee, S.dependentK., 2004. A model for dry
Sun, Z., Zhao, Y., Gao, H., Hu, G., 2010. Removal of SO2 from flue gas by sodium humate sodium bicarbonate duct injection flue gas desulfurization. Adv. Environ. Res. 8
solution. Energy Fuels 24, 1013–1019. https://doi.org/10.1021/ef901052r. (3–4), 655–666. https://doi.org/10.1016/S1093-0191(03)00038-8.
Suyantara, G.P.W., Hirajima, T., Miki, H., Sasaki, K., Kuroiwa, S., Aoki, Y., 2021. Effect of Yan, S., Wu, G., 2017. SO2 emissions in China – their network and hierarchical structures.
Na2SO3 on the floatability of chalcopyrite and enargite. Miner. Eng. 173, 107222 Sci. Rep. 7 (1) https://doi.org/10.1038/srep46216.
https://doi.org/10.1016/j.mineng.2021.107222. Yeh, J.T., Demski, R.J., Joubert, J.I., 1982. Control of SO2 emissions by dry sorbent
United Nations Industrial Development Organization (UNIDO), 2023. Circular Economy. injection. In: ‘ACS Symposium Series. American Chemical Society, Washington, DC.
URL: https://www.unido.org/sites/default/files/2017-07/Circular_Economy_UN https://doi.org/10.1021/bk-1982-0188.ch016, 1982.’, ACS Publications.
IDO_0.pdf. Zhao, Y., Guo, T.X., Chen, Z.Y., Du, Y.R., 2010. Simultaneous removal of SO2 and NO
United States Environmental Protection Agency (US EPA), 2003. Air pollution control using M/NaClO2 complex absorbent. J. Chem. Eng. 160, 42–47. https://doi.org/
technology fact sheet, EPA-452/F-03-034. https://www3.epa.gov/ttncatc1/dir1 10.1016/j.cej.2010.02.060.
/ffdg.pdf. Zhao, Y., Hu, G., 2013. Removal of sulphur dioxide from flue gas using the sludge sodium
United States Environmental Protection Agency (US EPA), 2019. Clean air markets humate. Sci. World J. 2013, 8. https://doi.org/10.1155/2013/573051 p.
program data. coal-fired characteristics and controls, p. 2019 url: https://www.epa. Yildirim, Ö., Kiss, A.A., Hüser, N., Leßmann, K., Kenig, E.Y., 2012. Reactive absorption in
gov/airmarkets/power-plant-data-highlights. chemical process industry: a review on current activities. Chem. Eng. J. 213,
United States Environmental Protection Agency (US EPA), 2021. EPA Air pollution 371–391.
control cost manual, Section 5: SO2 and Acid Gas Controls, EPA-452/B-02-001, . Zhang, Y., Wang, T., Yang, H., Zhang, H., Zhang, X., 2015. Experimental study on SO2
VanDerWerff, J., 2011. Technical report. Nol-Tec Systems, Inc. of Lino Lakes, MN. recovery using a sodium–zinc sorbent-based flue gas desulfurization technology.
VanNes, R., Somers, R., Weeks, R., Frank, T., Ramans, G., LaManita, C., Lunt, R., Chin. J. Chem. Eng. 23 (1), 241–246. https://doi.org/10.1016/j.cjche.2014.10.007.
Valencia, J., 1979. Technical Report. Louisville Gas and Electric Co., KYUSA https:// Zhu, Q., 2010. Non-calcium desulphurisation technologies. IEA Clean Coal Centre. ISBN
doi.org/10.2172/5379912. 978-92-9029-490-0.
Vázquez, G., Antorrena, G., Chenlo, F., Paleo, F., 1988. Absorption of SO2 by aqueous
NaOH solutions in the presence of a surfactant. Chem. Eng. Technol. 11 (1),
156–162. https://doi.org/10.1002/ceat.270110121.

183