Biomass Review - 2ndrevision - Clean
Biomass Review - 2ndrevision - Clean
Source: A Review.
Article:
Tomlin, AS orcid.org/0000-0001-6621-9492 (2021) Air Quality and Climate Impacts of
Biomass Use as an Energy Source: A Review. Energy & Fuels. acs.energyfuels.1c01523.
ISSN 0887-0624
https://doi.org/10.1021/acs.energyfuels.1c01523
Reuse
Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless
indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by
national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of
the full text version. This is indicated by the licence information on the White Rose Research Online record
for the item.
Takedown
If you consider content in White Rose Research Online to be in breach of UK law, please notify us by
emailing [email protected] including the URL of the record and the reason for the withdrawal request.
[email protected]
https://eprints.whiterose.ac.uk/
The Air Quality and Climate Impacts of Biomass Use as an Energy Source: A
Review
Alison S. Tomlin*
School of Chemical and Process Engineering
University of Leeds
Leeds LS29JT, [email protected]
ABSTRACT
Commitments made to meet the goals of the Paris Agreement, and the need to develop low carbon
sources of power and heat, are likely to lead to the continued or even increased use of biomass as
an energy source. Biomass is already replacing coal within large scale electricity generation systems.
At smaller scales it is utilised within combined heat and power systems, as well domestically for the
provision of heat. Unabated, biomass combustion leads to a range of pollutant emissions, including
CO, CO2, methane, black and organic carbon, which are of relevance from a climate perspective, as
well as particulate matter, nitrogen oxides, volatile organic compounds, dioxins, furans and
precursors of ozone and secondary organic aerosols, which impact on air quality (AQ) and human
health. Epidemiological data suggests that biomass smoke contributes to at least 40 000 premature
deaths per year in Europe, as well as negatively affecting respiratory and cardiovascular health.
Hence, if biomass is to be used as a future energy source, its AQ impacts need to be addressed.
Here, a short review is provided of the potential climate and AQ impacts of the direct use of solid
biomass as a fuel for the provision of power and heat. The review provides a brief summary of the
chemical and physical characteristics of emissions from biomass utilisation for energy provision. The
impacts of biomass burning on ambient AQ and health (in particular for fine particulates) are
addressed in both outdoor and indoor environments, and for industrialised and developing
countries. Issues of appliance scale, and how these influence available mitigation options, emissions
factors and regulatory limits for biomass combustors, are discussed. Final perspectives are offered,
from the point of view of the need to develop “win-win” strategies for the future utilisation of
biomass which minimise both climate AND air quality impacts.
TABLE OF CONTENTS
1. INTRODUCTION
2. POTENTIAL CLIMATE IMPACTS FROM BIOMASS BURNING
2.1 Climate Impacts of Black Carbon and Organic Carbon
2.2 Gaseous Emissions of Relevance to Climate Impact
2.3 Negative Emissions Technologies (NETs)
2.4 The Use of Forestry Products
2.5 Dedicated Energy Crops
2.6 Peat
3. AIR QUALITY STANDARDS AND ATTAINMENT WITHIN EUROPE
4. PHYSICAL AND CHEMICAL CHARACTERISTICS OF EMISSIONS FROM BIOMASS COMBUSTION
5. IMPACTS OF BIOMASS BURNING ON AIR QUALITY IN DEVELOPED COUNTRIES
6. HEALTH IMPACTS OF PM FROM BIOMASS BURNING
7. EMISSIONS FACTORS AND MITIGATION STRATEGIES FOR BIOMASS COMBUSTORS: ISSUES OF
SCALE AND BIOMASS COMPOSITION
7.1 Fuel Pre-Treatment Methods for Biomass Resources
7.2 Emissions Factors and Abatement Methods for Larger Scale Devices
7.3 Emissions Factors and Abatement Methods for Medium Scale Plant and CHP
7.4 Emissions Factors for Domestic Scale Devices
7.5 Influence of Fuel Quality
7.6 Abatement Methods for Small Scale Devices
7.7 PM Composition Based Emissions Factors
7.8 VOC Emissions Factors
8. ISSUES OF FUEL SWITCHING
9. IMPACTS ON INDOOR AIR QUALITY AND HEALTH IN DEVELOPING COUNTRIES
9.1 Indoor Concentrations
9.2 Mitigation Measures in Developing Countries
10. FINAL DISCUSSION AND FUTURE DIRECTIONS
1. INTRODUCTION
The commitment of countries around the world to meeting the goals of the Paris Agreement1
necessitates the rapid uptake of low carbon energy systems for the provision of electricity, heat and
mobility. The large-scale deployment of low carbon technologies such as solar, wind and hydro
power has led to reductions in the carbon intensity of the electricity sector in many countries, with
the greenhouse gas (GHG) emission intensity of electricity generation within the EU for example,
falling from over 500 g CO2eq/kWh in 1990 to around 275 CO2eq/kWh in 2019.2 However, in sectors
that are harder to decarbonise, such as heat and transportation, progress has been slower. The
transport sector unfortunately provides some alarm bells in terms of strategies that were previously
chosen with the aim to reduce GHG emissions, that have impacted on other sustainability goals
including air quality (AQ). Within the EU, for a period of time, diesel vehicles were favoured over
their gasoline counterparts due to lower CO2 emissions per km. However, the increased uptake of
diesel vehicles has perpetuated high concentrations of nitrogen oxides (NOx) and airborne
particulate matter (PM) in European cities, with known consequences on human health. It is
important to acknowledge therefore, that climate change is not the only environmental challenge
that results from emissions to the atmosphere. The World Health Organisation (WHO) estimates that
around 7 million deaths occur globally each year as a result of exposure to air pollution, including
both outdoor and indoor exposure.3 Although it is somewhat difficult to separate the effects of
individual pollutants on morbidity and mortality from epidemiological data,4 it is known that a
number of different pollutants are involved, including nitrogen dioxide (NO2), fine particulates
(PM2.5), sulphates, ozone and dioxins. There are some complex issues at play, since technologies
proposed to reduce CO2 emissions do not necessarily provide the best way of tackling local air
pollution, or even of reducing other compounds of relevance to radiative forcing such as nitrous
oxide, methane and atmospheric aerosol. It is therefore important to ask whether proposed
strategies for reducing CO2 have the desired effect on atmospheric greenhouse gases (GHGs) and
aerosols and consequently the total radiative forcing of the atmosphere, as well as what the
implications might be for local AQ. Low carbon options should only be chosen as part of future
energy strategies if they do not lead to a further deterioration of AQ, and better still if they lead to
AQ improvements.
Passenger transport is likely to move away from internal combustion engines, with the electrification
of passenger cars rising in many European countries, led by Norway which surpassed 50% of new car
sales being electric in 2020.5 At least part of the impact of diesel exhaust emissions on air urban
quality is likely to decline as a result. Globally, however, despite the increasing penetration of
renewables and hydro, electricity production is still dominated by the combustion of fossil fuels,
with the International Energy Agency (IEA) estimating that 38% of that production came from coal
combustion in 2018. As countries seek to address commitments made through the Paris agreement,
a move away from coal for electricity production will be necessary, with the replacement of coal by
biomass fuels being one potential option within mixed energy strategies.6 In 2017, renewables made
up around 14 % of global energy supply, with biomass forming 70 % of the renewables total. 7
Globally, biomass supplied close to 600 TWh of electricity in 2019, more than double the amount
supplied in 2009.8 Usage of biomass for power generation, is dominated by Europe and Asia as
illustrated in Figure 1a. A report from the International Institute for Applied Systems Analysis
estimates that, between 2005 and 2030, the use of biomass for energy production in the EU is likely
to double.9 Globally, as part of the Sustainable Development Scenario, 2000-2030, the use of
biomass for power is predicted to almost double from current rates by 2030.8
(a) (b)
Figure 1. Global biomass usage figures for 2017 within different continents (a) power generation
from biomass (b) domestic supply of biomass.7
There are also significant challenges in providing low carbon heat, which is required in both
industrial and domestic sectors. Industrially, natural gas is currently used for heat provision in many
different processes including steel and glass production, chemical and mineral industries. Whilst
liquid biofuels are being explored as a potential part of the future energy mix for such industries,
solid biomass is unlikely to form a major component. The same may not be true however, for the
provision of domestic heat and for combined heat and power (CHP). Across Europe for example,
biomass is used in small scale residential (<50 kilowatt thermal (kWth)) heating appliances and in
medium scale (<20 MWth) district heating/industrial systems as well as in large centralised CHP
plants.10 The current number of residential biomass combustion systems within the EU is estimated
to be ∼65 million direct heating appliances (fireplaces, stoves and cookers) and ∼8 million indirect
heating appliances (boilers), firing mainly using wood logs, but also wood chips and fuel pellets.11
Figure 2 shows that such devices, particularly domestic stoves, have been estimated to make a
substantial contribution to primary fine particulate emissions within the EU, and were predicted to
have grown in terms of their relative importance compared to other sources such as traffic, from
25% of total primary sources in 2000, to 38% in 2020.12
(a)
(b)
Figure 2. Sector contributions to primary PM2.5 emissions in the EU15, predicted for (a) 2000 and (b)
2020. Reproduced from Amann, M.; Bertok, I.; Cofala, J.; Gyarfas F.; Heyes, C.; Klimont, Z.; Schöpp,
W.; Winiwarter, W., M. A Baseline Scenarios for the Clean Air for Europe (CAFE) Programme, 2005,
Submitted to the European Commission Directorate General for Environment, Directorate C –
Environment and Health. Copyright [2005] with permission from IIASA.12
CHP offers efficiency benefits over conventional power generation since the process captures and
utilises heat that is a by-product of the electricity generation process. By generating heat and power
simultaneously, CHP can reduce carbon emissions by up to 30% compared to the separate means of
conventional generation via a boiler and power station, with an overall efficiency of > 80%. CHP is
therefore likely to form part of strategies to decarbonise both electricity and heat sectors with low
carbon fuels such as biomass likely to form part of the mix, potentially replacing natural gas.
Williams et al., based on a study of several pathways to meet carbon reduction targets in the UK
Climate Change Act, suggest that to a certain extent, particulate emissions reductions resulting from
the electrification of transport could be balanced by significant increases in biomass burning in
pathway scenarios involving CHP and domestic biomass use. 13 These issues therefore pose the
question as to whether the use of biomass as an energy source has the potential to become the
“new diesel” in terms of being promoted as a lower carbon source of power and heat compared to
fossil fuels, but with potentially unwanted side effects in terms of increased air pollution exposure,
especially to fine particulates and NO2.
Globally, biomass is extensively used for both cooking and heating with domestic supplies of
biomass being significant across all continents apart from Oceania as shown in Figure 1b. The
regional profile for domestic usage of biomass is unsurprisingly quite different from that for power
generation, with the Asian and African continents dominating domestic supply, due to usage of
firewood and charcoal fuels for cooking and heating. Within the developing world the current energy
landscape and future perspectives are quite different from those in industrialised countries. It was
estimated in 2000 that approximately half of the global population relied on coal and biomass for
domestic heating and cooking, with significant impacts on AQ and health.14 Here then, improving
combustion technologies for the utilisation of biomass fuels in order to lower emissions, has the
potential to provide significant benefits to both health-related air pollution problems as well as
climate change.
The review will focus on the AQ, and to a lesser extent, the health and climate impacts of the
utilisation of solid biomass fuels for the provision of heat and electricity both in indoor and outdoor
settings. Solid biomass dominates the global energy use of biofuels, with 86% of use in the form of
primary solid biofuels in 2017.7 Whilst solid biomass can be processed to form a wide range of
products including gas and liquid fuels, the intention here is to focus on the impact of its use in solid
form i.e. where the fuel fed directly to the combustion device is a solid, although this may include
processed solid fuels and internal gasification strategies. The nature of the short review necessitates
a selective rather than a comprehensive use of references. Examples of both policy and technically
driven mitigation strategies will be given. For policy driven measures in developed countries, the
focus will be more on the European region, since not all countries can be covered in such a short
review and the challenges faced by Europe are representative of other industrialised nations. Final
discussions will offer future directions for the effective use of solid biomass resources within the
context of reducing AQ and GHG impacts in both industrialised and developing regions.
The combustion of biomass can result in emissions of relevance to both air quality and climate, and
it is important to consider both impacts in an integrated way when designing strategies for the
utilisation of biomass. Hence in this section, the potential climate impacts of biomass utilisation as
an energy source are covered from the perspective of both emissions to atmosphere, and
assumptions about the carbon neutrality of biomass resources.
2.1 Climate Impacts of Black Carbon and Organic Carbon
Black carbon (BC) plays an important role in the Earth's climate system because it absorbs solar
radiation, influences cloud processes, and through deposition can alter the melting of snow and ice
cover. Bond et al. provide an estimate for the direct radiative forcing of atmospheric BC in the
industrial‐era (1750 to 2005) due to absorption of +0.71 Wm−2 with 90% uncertainty bounds of
(+0.08, +1.27) Wm−2.15 In addition to direct forcing effects, BC aerosol can influence cloud
characteristics and can also influence the surface albedo through deposition onto snow and ice.
When cloud and cryosphere adjustments are taken into account this increases radiative forcing
effects to +1.1 Wm−2 with 90% uncertainty bounds of +0.17 to +2.1 Wm−2. This makes BC the second
most important contributer to changes in radiative forcing since 1750, with only CO2 being higher at
1.68 Wm−2 (1750 to 2011).16 Since BC has a short atmospheric lifetime, several authors have
suggested that reducing emissions of BC would provide a rapid payback in terms of reducing global
warming.17,18 Hence reducing BC emissions from combustion provides a win-win scenario of leading
to improvements both in climate forcing and in human health outcomes by improving AQ. Organic
carbon (OC) on the other hand is thought be a climate cooler with estimates of forcing in the
industrial era of -0.29 (-0.47 - -0.08) Wm−2.16 BC and OC are both major pollutants from biomass
burning and their relative emissions will influence subsequent climate impacts. It is therefore vital
that emissions profiles of both BC and OC from biomass combustors are well understood and this
point will be returned to in section 7.
Globally, N2O emissions are dominated by agricultural sources rather than combustion ones,
although this is of relevance to the use of biomass as an energy source where nitrogen based
fertilisers are used to enhance yields. Full life cycle analyses (LCAs) are required to assess the overall
GHG emissions of biomass fuel utilisation from production, to transportation and final end use
within different energy sectors. The review of such lifecycle studies is outside the scope of this work,
and although there are a growing number of reviews of LCAs for liquid biofuel use within the
transport sector (so called Well-to-Wheels analyses),19,20 there are fewer addressing the use of solid
biomass fuels in the electricity and heat sectors. A recent example however, compares the life cycle
impacts of bioenergy power plants fuelled by three types of residual biomass (forest residues, cereal
straws and cattle slurry) with fossil based systems using the European power generation mix as well
as coal and natural gas power plants.21 Systematic reviews of bioenergy LCAs were also presented by
Muench and Guenther22 and Varun et al.23 focussing on the electricity and heat sectors. A specific
study for the heat sector was carried by Giuntoli et al. who estimated that emissions of N2O are
responsible for 27%, 13%, 6% of the total GHG impact of the use of logs within a domestic stove,
wood chips used within a district heating plant, and pellets used within a domestic stove
respectively.24 Emissions of methane from both biogenic and fossil sources had a more minor
influence: 9%, 8% and 7% for the log, chip and pellet pathways respectively. GHG emissions from
biomass utilisation cannot therefore be ignored. Such LCAs are critical to the assessment of policies
being developed to mitigate climate change, but of course the robustness of the conclusions drawn
will depend heavily on the accuracy of the inputs to the models, including emissions factors for each
part of the pathway such as crop fertilisation, fuel transportation, fuel processing, combustion, etc.
The methodologies used for estimating GHG emissions factors for a range of solid and gaseous
bioenergy pathways is discussed in Giuntoli et al. within the context of assessing compliance with
the EU’s Renewable Energy Directive (RED) (2009/28/EC) and the Fuel Quality Directive (FQD)
(2009/30/EC).25 The report presents emissions factors for N2O, methane and overall CO2 eq for a
range of production and utilisation pathways including those for pellet boilers, pellet and wood chip
based CHP plant, and straw based boilers, and for different biomass sources, fuel transportation,
processing and fertilisation strategies.
Biomass is a very heterogeneous material and the impacts of its utilisation on both climate and AQ
depend very much on how it is sourced and treated prior to its combustion.26 In this section, factors
potentially affecting the carbon neutrality of selected biomass fuels are briefly covered.
Even within the targets set by the Paris Agreement, some non-zero sources of GHGs are likely to be
unavoidable. This has led to the concept of negative emissions strategies or technologies (NETs),
where the net result of the process should be to remove CO2 from the atmosphere. NETs are part of
all pathways to net zero CO2 by 2050 that were modelled for the 2018 Intergovernmental Panel on
Climate Change (IPCC) report on achieving only 1.5 °C of warming by 2100. The slower GHG
emissions are reduced, the larger such technologies feature in the proposed pathways.27 In pathway
P4, where a resource- and energy-intensive scenario is assumed, bioenergy, carbon capture and
storage (BECCS) technologies feature very heavily by the middle of the 21st Century, sequestering 20
GtCO2/yr. Within BECCS, the carbon is assumed to be fully absorbed during biomass growth, and is
captured pre- or post-combustion and stored underground indefinitely. If only the combustion
process is considered, then producing a kWh of electricity from wood causes significantly more CO2
than from using coal. The assumption of carbon neutrality stems from the assumption that the
carbon released when biomass is burned, is sourced from forests that are sustainably managed such
that it will be recaptured by a new stock of trees in the future.28 In the domestic sector, the use of
wood has been marketed as renewable, since it is at least possible that the source of wood may be
renewed or replanted, thus over time sequestering the CO2 that is released on its combustion. That
time however, may be decades,29,30 when we in fact require very short time scale emission reduction
strategies in order to avoid a 1.5 °C overshoot. The carbon reduction benefits of bioenergy
compared to the use of fossil fuels are thus time‐dependent29 and supply to both the domestic
market and large scale power producers leads to complex CO2 accounting issues.
A valid question is whether the potential for reliance on NETs in current plans is potentially locking
us into energy and resource intensive lifestyles both now and in the near future, based on the
premise of speculative technologies further down the line.26 The scale of biomass resource
requirement assumed in the more BECCs intensive pathways put forward by the IPCC, is typically
one to two times the area of India, which raises questions about land availability, biodiversity loss,
competition with food production, and competing demands for bioenergy resources from other
sectors such as transport, heating, and industrial processes. 26
Globally, the estimate for wood pellet use in 2018 is > 50 million tons per annum, with Europe in
particular forming the largest consumer region. 31 The total consumption of wood pellets in the EU
was estimated to be 20.3 million metric tons in 2015 with 70 % of this supply sourced from the US.32
By country, the UK was the top consumer of wood pellets in the EU region in 2018, with 8 million
metric tons used. This has occurred as a result of the large scale switching of power generation from
coal to biomass at the UK’s largest power station, Drax (4 GW), causing the UK to become the
world’s largest importer of wood pellets over a period of just 5 years.33 The supply is predominantly
sourced from the US, Canada and Latvia. Drax is mandated under law to include lifecycle emissions
from the cultivation, harvest, processing and transport of wood pellets, and claims that most come
from waste products such as thinnings, forest and sawmill residues. The methodology potentially
however, ignores large changes in forest carbon stocks and indirect effects caused by increased
demand for wood products. The approach was subsequently challenged by a counter factual model
which asks the question, what will happen to this waste wood if it is not pelletised and exported to a
distant power station? The report found that a wide range of scenarios could result, from wood
burning having the potential to reduce atmospheric CO2, all the way to it being worse than burning
coal.33 The differences arise from the vastly differing CO2 emissions that can arise when in reality,
using these seemingly waste products, could actually be diverting them from other end uses. The
distinction for forest residues is between fine woody materials, which may be better burnt in a
power station instead of being left on the forest floor to rot and release methane, compared to
larger diameter wood that could potentially store its carbon on the forest floor for many years.29,33
Giuntoli et al. suggest a high sensitivity of the mitigation potential of using forest residues to decay
rates of material on the forest floor, concluding that power generation from forest logging residues
is an effective mitigation solution compared to the current EU mix, only when decay rates are above
5.2% a-1.21 A perhaps more controversial area is the harvest rates that are used to supply the
increased demand for pellets in forests that are typically naturally regenerating. According to the
counterfactual model, if forests are harvested every 60 years instead of every 70 years, then the
amount of carbon locked in the forest is reduced, and the CO2 emissions from the combustion of the
resulting thinnings exceeds that of coal usage. The general principle for the sustainable use of wood
should be that stocks (and their embedded carbon) increase more rapidly due to the development of
a market that uses the wood, than they would have in its absence.28 If forests are cultivated on
marginal lands this may well be the case, but where biomass is derived from waste products of
industries primarily concerned with producing other wood products such as timber or wood pulp for
paper, or when land with high carbon stocks is converted to low productivity bioenergy plantations,
this cannot be taken for granted.28,29 Whether stock regeneration management practices are
adjusted to suit the waste wood market also muddies the waters as highlighted above. Drax on the
other hand argues that “forests are growing, not shrinking, in all of the US states that Drax sources
from” and hence that carbon stocks are increasing as a result of this new market.33 Biodiversity is
also a key factor that must be considered if we are to seek win-win solutions, as discussed in a
recent report on the use of biomass for power in the EU by the European Commission Joint Research
Centre (JRC).34 The JRC report also highlights accounting anomalies in reported usage of woody
biomass across the EU compared to sources. The issues around carbon accounting are clearly
complex, which makes trading in waste wood resources difficult to manage from a sustainability
accountability and enforcement perspective. What looks like an NET on paper may or may not play
out that way in the real world. The use of detailed LCAs and counterfactual models will be required
to adequately assess the overall GHG savings from different feedstocks and utilisation strategies.
Dedicated energy crops are grown specifically for their utilisation in energy production in ways that
do not displace food production. They include short rotation wood crops such as willow, and
perennial grasses such as miscanthus and reed canary. In principle they avoid the time-scale issues
that forestry based biomass presents, in that they tend to grow and be cropped on annual cycles.
When directly compared with alternative land use options their effective CO2 per area of land can be
low, as long as indirect emissions of GHGs such as N2O from fertiliser use are avoided. However, the
carbon impacts from indirect land use change when biomass crops replace other crops are uncertain
and potentially could be quite large requiring further consideration. There are also potentially
substantial implications for biodiversity which are not covered here but were, for example,
addressed in a recent review for the South Eastern US region.35 They do not currently form a large
part of the power production or the domestic heating market but could have potential for
application in CHP or boiler applications. In Europe for example, solid biomass energy crops covered
about 50,000 – 60,000 ha of land in 2007 which is significantly lower than the 2.5 million ha used for
energy crops grown for the production of transportation biofuels such as cereals and rape. 36 A range
of other second and third generation biomass feedstocks exist which include straw, bagasse,
jatropha, waste vegetable oil, municipal solid waste and algae. Many of these may be processed to
liquid biofuels rather than having large global usage as a direct source of energy and thus are not the
focus of this review. Excellent reviews on the production of liquid biofuels from 1st to 3rd generation
sources can be found in Nigam and Singh37 and Fivga et al.38
2.6 Peat
Peat is a low calorific fuel that is still used in both domestic heat and electricity generation sectors. It
is sometimes considered a “slowly renewable energy” and thus is mentioned briefly here. It is,
however, classified along with solid fossil fuels rather than biomass fuels by the IPCC since its
emissions are considered to be comparable to coal. Peatlands in their natural state act as a sink for
atmospheric CO2, and the carbon content of the world’s peatlands represents around a third of all
terrestrial soil organic carbon, despite covering only 3% of the earth’s surface.39 However, whilst
pristine peatlands are generally a net carbon sink, they can also be a significant source of methane
and there are significant uncertainties in estimating the carbon storage in peatlands under current
and future climate scenarios.40 Disturbed peatlands (e.g. through drainage for forestry, agriculture
and peat removal) can alter the balance of CO2 vs. CH4 emissions but it is suggested that subsequent
to the harvesting of peat for energy on a peatland, the remaining cutaway persists as an
atmospheric source of CO2 as the residual peat continues to decompose. The extraction of pristine
peatlands for energy (as has been the case in Ireland) rather than drained peatlands, leads to the
highest net GHG emissions from energy production; greater than the use of coal per unit of energy
produced. A life cycle impact assessment of peat use for energy generation in Ireland, including
processes such as peatland drainage and industrial extraction, transportation, combustion, and
subsequent use of the cut-away land41 found that the combustion phase contributed over 95 % of
the total life cycle GHG emissions. To a certain extent, the emissions could be mitigated by the
restoration of the cutaway areas, through for example, afforestation of cutaway peatlands.42
However, there can also be considerable losses of soil carbon from residual peat that continues to
decompose.41,42 Time-scales again are critical, since the return of the land to a carbon sequestration
state may take many years. A further controversial aspect of peatlands in tropical regions of
Southeast Asia, is that they are being cleared to create land for industrial palm oil and pulpwood
plantations, a bi-product of which could be used as forestry wastes for biomass energy production.
The IPCC estimates that tropical peatland cultivation generates the equivalent of 440 million tonnes
of CO2 per year.43
Overall, the literature suggests that appropriate methods of carbon accounting using a common set
of assumptions and system boundaries32 will be a vital part of using biomass effectively, both as a
renewable energy resource and within NET strategies. In particular, issues such as land-use change,
definition of waste products, harvest rates, timescales of forest residue decay and carbon
sequestration as well as soil based carbon should be considered.32 Counterfactual models have been
suggested as a means of developing more realistic net GHG changes for particular utilisation
strategies.
The AQ impacts of biomass combustion must be considered when defining effective strategies for its
future use. This must include assessing the current state of AQ in a particular country or region and
whether the future use of biomass will impact on that region’s ability to meet AQ standards. Many
nations have ambient standards for several of the pollutants that can result from biomass
combustion. Whilst it is not possible in such a short review to consider the AQ standards in all
countries, Europe is covered here as an example of an industrialised region with a well-established
framework for monitoring AQ and the attainment of ambient concentration standards. WHO
standards are also addressed since often they provide more stringent targets than national limits,
and should in principle underpin long term health based targets for all countries.
Nitrogen oxides (NO + NO2) are acidic gases and ozone precursors and can affect human health and
vegetation. NO2 is thought to have both acute and chronic effects on airways and lung function,
particularly in people with asthma. As a result, both hourly and annual standards have been set
within Europe for NO2 to protect human health and vegetation within the Directive on Ambient Air
Quality and Cleaner Air for Europe (2008/50/EC). An hourly standard of 200 μgm-³ should not be
exceeded more than 18 times per calendar year and an annual standard of 40 μgm-3 has been set for
the protection of human health, consistent with WHO guidelines. The EU standard for ozone is 120
μgm-³ as a maximum daily 8 hour mean which is slightly higher than that set by the WHO of 100
μgm-³.
Air quality standards for particulates within Europe are based on controlling concentrations of PM10
to below 50 μgm-3 for a running 24-hour average, with 35 annual exceedences permitted per year.
There is also an annual standard of 40 μgm-3. The term PM10 refers to the mass of particles that pass
through a size-selective orifice with a 50% collection efficiency cut-off at 10 μm aerodynamic
diameter. This metric was chosen for the development of air quality and emissions standards as it
best represents those particles most likely to reach the lung acinus.44 An annual limit value of 25
μgm-3 has also been set for PM2.5 although this is significantly higher than the annual mean target set
by the WHO of 10 μgm-3. In terms of impact, a distinction is made between PM10 which are termed
“thoracic” particles, PM2.5 termed “respirable” particles that can penetrate into the gas-exchange
region of the lung, and ultrafine particles which are smaller than 100 nm. Exposure reduction targets
have also been set within Europe for PM2.5 based on a percentage reduction from the average
exposure indicator (AEI) in 2010, although no standards have yet been set for exposure to ultrafine
particles.
A variety of legislation and technological developments have led to reductions in emissions of major
pollutants and therefore improvements in AQ in Europe. NOx emissions decreased by 44% and PM10
by 24 % between 1990 and 2011. As a result, the percentage of urban citizens exposed to pollutant
levels above standards fell between 2000 and 2018. However, poor AQ remains a problem and in
2018 the percentage of citizens exposed to levels above EU limit values was estimated to be 34 % for
O3, 15 % for PM10, 3.8 % for PM2.5, 3.6 % for NO2 and 15-20 % for Benzo-a-Pyrene (BaP).45 The
percentage for PM10 has in fact increased in recent years. Hence there remains a challenge for all
pollutants, even with respect to EU Directive values. WHO recommendations are more stringent and
estimates of exceedances of these run to 73.6 % of the population of Europe for PM2.5 and 98.6 % for
O3. Europe is an example of a region with a well-established legal framework which aims to reduce
the exposure of its population to harmful pollutants, but which is a long way off reaching the goal of
complete compliance with the WHO’s AQ guidelines. This reaffirms the point that any policies
chosen for the reduction of GHGs need to consider AQ impacts very carefully.
Emissions from the combustion of biomass within both large scale power generation processes and
residential and commercial heating appliances will include a complex mixture of gases and particles.
Gaseous emissions will contain differing levels of carbon monoxide (CO) and carbon dioxide (CO2)
depending on the completeness of combustion, NOx from both fuel, and to a lesser extent thermal
sources, a wide range of volatile organic compounds (VOCs), sulphur oxides (SOx) and a range of
trace species potentially including heavy metals and polyaromatic hydrocarbons (PAHs), furans and
dioxins; as well as organic and inorganic aerosol particulates.46 A schematic diagram summarising the
potential variety of emissions is shown in Figure 3.
Wood consists primarily of the polymers cellulose, hemicellulose and lignin. Other biomass fuels
such as grasses, and wheat stubble also contain these polymers, although in different relative
proportions. Small amounts of low-molecular-weight organic compounds such as resins, waxes,
sugars and inorganic salts are also present in wood. During combustion, pyrolysis occurs producing a
variety of smaller molecules. Biomass combustion is typically inefficient, and as a result, a variety of
partially oxidised organic chemicals are generated in particulate emissions from biomass. PM
emissions are perhaps of the greatest concern since, particularly for domestic use of biomass, they
are difficult to control for several reasons that will be discussed later in the review. Of the PM
produced during wood combustion in an enclosed residential wood burner, 96% was suggested to
be in the PM10 size fraction and 93% in PM2.5 by Houck and Tiegs.47 A study of PM emissions from a
fixed grate domestic stove using a range of raw and torrefied biomass fuels, found over 96% of PM
mass to be in the submicron (PM1) range for all fuels tested.48 This puts the majority of emissions
into the respirable fraction of PM and is also of significance in terms of meeting AQ and exposure
standards for PM2.5.
Figure 3. Schematic diagram illustrating the wide variety of emissions that can result from the
combustion of biomass within both residential and commercial appliances.
The chemical characteristics and PM size distribution of the emissions are likely to vary depending
on the type of appliance, burn rates, fuel moisture content, and the type of biomass used. The
chemical components prevalent within each size fraction are also highly variable and may affect
their toxicological impacts.49 Fine et al. suggested that as much as 74% of the PM2.5 emitted from
domestic fireplaces burning wood was OC and 1-18% was elemental carbon (essentially BC).50,51
Depending on the wood species burned, salts and minerals present within PM could include Na, Mg,
K, Ca, Zn, ammonium, sulphates, chlorides, carbonates and nitrates.52 Potassium was found to be the
most abundant elemental species contributing up to 1.5 wt % of fine particle mass.51 Its presence in
biomass fuels has been found to lead to severe deposit formation, thus causing potential corrosion
problems.53 The organic compounds present in wood smoke will also vary depending on combustion
conditions and wood characteristics such as lignin and resin content. McDonald et al., in a study of
emissions from a range of wood burning stoves and fireplaces, suggested that most of the aliphatic
and olefin hydrocarbon VOC content is made up of C1-C6 compounds.54 Dienes, especially 1,3-
butadiene and 1,3-cyclopentadiene, were also found to be enriched in wood combustion emissions
deriving from the decomposition of terpenoids.55 Significant concentrations of dicarboxylic and
diterpenoid acids were found in wood smoke in a study by Rogge et al.56, particularly from the
burning of pine woods. A comprehensive review of the emissions of various aromatic compounds
and PAH from a range of different wood-types is given in Fine et al. for various species of wood
grown in the Northeastern United States.51 They found that the pyrolysis product of cellulose
(levoglucosan) was by far the most abundant component of the fine particle organic emissions (3%-
12%). The high emission rates of levoglucosan and its uniqueness to biomass combustion make it an
important candidate as a marker for biomass combustion in general, that can be used in ambient air
pollution studies for source apportionment of PM (see section 5). Other compounds include a
variety of aldehydes, phenols, alcohols, ketones, carboxylic acids, methane, ethane, ethanol,
formaldehyde, acetaldehyde, acrolein, cresol, formic acid and acetic acid.52,54,57 Dioxins
(polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs)) are also
present in emissions from wood burning.58,59 Furans, formed from the decomposition of wood
cellulose upon heating, were found in significant quantities from softwood and hardwood emissions
by McDonald et al.54 A comprehensive review of PCDDs and PCDF emissions from biomass burning is
provided in Lavric et al.58
Mixing between fuel and oxygen within the combustion chamber has a strong role to play in the
level of emissions of carbonaceous pollutants, since the formation of soot and PAH may be
enhanced in fuel rich zones where incomplete combustion occurs. The conclusion of an extensive
review of Williams et al.46 was that increased biomass use in small units such as for space heating or
cooking could lead to ineffective mixing and potential problems arising from emissions of
particulates, particularly carbonaceous smoke. They suggest that biomass burned in larger
combustion units leads to better mixing thus reducing carbonaceous emissions, which can be
additionally removed from the flue gases using a range of different flue gas control options. They do
however conclude that there remains a potential health problem from submicron inorganic particles
from devices at all scales.
Secondary pollutants formed from gaseous emissions of VOCs from biomass combustion are also of
concern. Tropospheric O3 is formed in the atmosphere from chemical interactions of VOCs, NOx, and
a radical source such as OH, which is formed from the photolysis of O3, aldehydes, hydroperoxides,
or nitrous acid (HONO). Biomass burning is a potential source of VOCs, NOx, and HONO and this may
contribute to tropospheric ozone formation. However, VOC species are emitted at varying relative
ratios depending on the fuel type and burn conditions making it difficult to predict the overall impact
of emissions from biomass combustion on O3 formation.60 In addition, primary emissions of organic
species can go on to form Secondary Organic Aerosol (SOA) which is particulate mass formed in the
atmosphere from a complex series of reactions of a large number of organic species resulting in the
formation of low volatility and/or high solubility oxidation products that partition to the particle
phase. SOA formation from biomass emissions is also highly variable.60,61
It is clear that to fully understand the impacts of biomass burning, a wide range of compounds needs
to be considered from the point of view of emissions factors for both gas and aerosol phase species,
and their impact on atmospheric concentrations of pollutants with potential health risks. These
issues will be covered in sections 7 and 5 respectively.
As suggested in the previous section, emissions arising from the combustion of biomass are likely to
be a significant contributor to atmospheric concentrations of, in particular, NO2 and PM2.5, with
domestic burning of wood, pellets, briquettes and peat of particular concern. Identifying the
contribution of biomass burning to atmospheric PM10 and PM2.5 can be approached using source
apportionment methods based on measurements at a receptor site (receptor-oriented models), but
it is not a straightforward task. Such methods are commonly based on the use of chemical tracers.
While there are no absolutely specific chemical tracers for the burning of wood, levoglucosan and
water soluble fractions of organic carbon, potassium and rubidium are potential candidates, with the
ratio of the concentrations of different tracers used in some studies and principal component
analysis (PCA) of multiple tracers used in others.51,62,63 The ratio between levoglucosan and
mannosan concentrations has also been shown to indicate a preference for softwood or hardwood
particulates within measured samples.64 The ratio of levoglucosan to wood smoke mass and the
ratio of levoglucosan to potassium concentrations within atmospheric PM can be highly variable,
according to the type of wood being burnt and the combustion conditions, which makes the
conversion process more difficult.65,66 A third method for estimating wood smoke concentrations is
based on the use of multi-wavelength aethalometers67 and depends upon the assumption that
elemental carbon arises from only road traffic and wood smoke, and that the two sources of black
carbon may be differentiated from their Angstrom coefficients. 66 Many studies use more than one
approach in order to reduce measurement based uncertainties.
Measurements of the contribution of wood smoke to overall PM concentration are highly variable
and depend on season, location and the conversion factors used. Critical reviews of the suitability of
specific organic markers and biomass emissions factors are given in Vicente and Alves68 and
Puxbaum et al.65 A range of values for the conversion from levoglucosan to woodsmoke
concentrations have been suggested within the literature from 6.0–23, with lower factors more
typical for softwood fires in stoves and fire places, and higher factors typical for hardwood types
such as beech.51,65,69 Bari et al., for example, suggest a higher value of 22.87 for hard woods and a
lower value of 10.7 for soft woods.69 Schmidl et al. on the other hand, suggest relatively similar
values for a range common woods growing in mid-European Alpine regions, apart from beech which
had a lower value below 5.70 The use of a conversion factor allows estimates for the contribution of
biomass burning to PM2.5 and PM10 to be made possible, but also means they are subject to
uncertainties, which makes the comparison of data between different sites rather tricky. Fuller et al.
compared aethalometer and levoglucosan estimates of mean PM across London from wood burning
and found them to be in good agreement, with estimates of winter contributions to PM10 of 1.8 μg
m−3 (levoglucosan) and 2.0 μg m−3 (aethalometer), representing between 7% and 9% of mean PM10
across the London transect.71
Some authors have suggested that within urban areas the contribution of wood smoke is generally
lower than at rural sites, and Caseiro and Oliveira suggested that the lack of difference between
urban roadside and urban background levoglucosan concentrations points towards the regional
nature of woodsmoke pollution.64 They estimated that wood burning contributed to about 3.1% of
the winter PM10 mass in Oporto, and to 3.7% in Copenhagen. Using multi-wavelength aethalometers
within an Alpine valley in Switzerland, Sandradewi et al. estimated that 51% of the BC was due to
wood burning and 49% due to traffic emissions, suggesting a high contribution of woodsmoke to PM
in rural mountain locations.67
Vicente and Alves provide a wide ranging review of available studies of outdoor AQ data68 related to
biomass burning in developed countries. They do not specifically suggest contributions to PM
concentrations, rather providing a summary of levoglucosan concentrations from different studies.
However, using an average conversion factor of 7.35 as suggested in Puxbaum et al.,65 allows the
range of PM2.5 concentrations from biomass burning to be estimated. For winter concentrations
within European cities these range from 2.8 μg m−3 in Porto up to 7 μg m−3 in Milan. The highest
measured urban concentrations presented in the review are around 20 μg m−3 in Fresno and Libby in
the United States and in Geeveston, Tasmania. These concentrations are twice the PM2.5 exposure
levels recommended by the WHO of 10 μg m−3.
Karagulian et al. compiled a source apportionment database for PM in global cities with domestic
fuel burning emerging as the main contributor to PM mass concentrations in Africa (34%), and in
Central and Eastern Europe (32%), and still a highly significant source in the Americas (25%),
Northwestern Europe (22%), the Southern China region (21%), South Eastern Asia (19%), and India
(16%).72 This is consistent with the review of Vicente and Alves within Europe,68 where accounting
for seasonal variation, the percentage contributions of biomass burning to atmospheric PM mass
concentrations could be as low as 1 % in the summer, but as high as 50% in winter samples for some
regions. Bari et al. suggested that during winter months, as high as 59% of ambient PM10 pollution
could be attributed to residential wood-fired heating in Stuttgart, Germany.69 It is clear that elevated
fine particulate levels due to biomass burning are a significant problem in developed countries and
affect both rural and urban areas.
The reasons for burning biomass domestically can vary, with it being the only source of heating in
some homes, to being an occasional secondary source of heat in others. Kalogridis et al., used an
absorption based method to assess the contribution of biomass burning to winter BC in Athens. 73
They found that 30% of winter time BC came from biomass combustion in 2013-2014 with a clear
diurnal pattern, showing highest concentrations during the night-time, supporting the theory of local
domestic heating being the main source. Fuller et al. suggest that an increase in the contribution of
biomass burning to PM concentrations over weekends in London indicates that the reasons for
burning wood are mainly discretionary.71 On the other hand, Saffari et al., reported that AQ in the
Greek city of Thessaloniki worsened during the recent economic crisis due to residents burning more
wood/biomass rather than using the more expensive options of oil and gas.74 As a result they
reported a 30% increase in PM2.5 concentrations associated with residential wood combustion in
2012 and 2013 which was concurrent with 30–40% lower concentrations of V and Ni, associated
with the combustion of residential fuel oil and industrial activity. In Denmark, a doubling of wood
stoves and boilers over a ten-year period was partially attributed to increased fossil fuel costs.75
However, a recent survey by the UK Department for Environment, Food and Rural Affairs (DEFRA)
provides a quantification of different solid fuel use in UK domestic settings from April 2018-March
2019, based on data collected by the market research group Kantar.76 Emissions inventories
estimate that domestic wood burning in the UK contributed 38% of primary emissions of PM2.5 in
2019, with emissions from this source having more than doubled since 2003.77 This is consistent with
the 40% suggested for Denmark.75 The Kantar study found that just 8% of people in the UK burned
solid fuels indoors, with two-thirds of them living in urban areas. One third had open fires as
opposed to modern stoves, and 96% had alternative sources of heating such as gas or electricity.
This suggests that, at least in the UK, discretionary burning as suggested by Fuller et al. 71 is
prevalent.
Sample location can be an issue in terms of obtaining an accurate picture of the contribution of
biomass burning to PM mass in urban areas. Typically, long term sample networks have been
designed on the basis that traffic emissions are likely to dominate, and are often therefore at
roadside locations or at urban background sites. Fewer sample sites tend to be located in residential
areas away from major roads, which could lead to a poor representation of domestic sources of
emissions within available data sets. In a hybrid approach, Larson et al. used a combination of fixed
and mobile nephelometer monitoring, coupled with a spatial buffering procedure to estimate the
spatial patterns of woodsmoke in and around Vancouver during 19 cold, clear winter evenings from
November, 2004 to March, 2005.78 Measurements at the fixed monitoring site were correlated with
the measured average levoglucosan concentrations at the same locations producing an R2 of 0.66,
thus to a certain extent validating the mobile nephelometer data. The use of this type of hybrid
approach allows for better spatial coverage of the city, allowing the identification of regions with
relatively high woodsmoke levels within an urban airshed. Such “hotspots” were found in residential
neighbourhoods, and thus in areas that would not normally be monitored using spatially fixed sites.
This is illustrated in Figure 4, reproduced from ref.,78 and showing large regions of high woodsmoke
concentrations which are at substantial distances from the PM2.5 regulatory monitoring sites. In
some residential regions, the two-week average PM2.5 concentrations well exceeded the 10 μgm-3
recommended by the WHO.78 The study suggests that were the WHO limit values for PM2.5 to be
adopted by different countries, that monitoring strategies suitable for assessing attainment may
need to be developed.
Figure 4. Adjusted light scattering values from measurements in the Greater Vancouver Regional
District (GVRD). The locations of the regulatory PM2.5 samplers, the fixed filter-based sampling sites,
and the two major GVRD mobile sampling routes are also shown. The dark shaded regions show the
highest woodsmoke levels as predicted by the “best” model. Reproduced from Larson, T.; Su, J.;
Baribeau, A.-M.; Buzzelli, M.; Setton, E.; Brauer, M. A Spatial Model of Urban Winter Woodsmoke
Concentrations, Environ. Sci. Technol, 2007, 41 (7), 2429–2436. Copyright [2007] American Chemical
Society.
Clearly, woodsmoke concentrations can vary over quite short distances. Allen and Rector conducted
a rural study in the Adirondacks of NY.79 They found that increases in concentrations were mainly
driven by local sources, as demonstrated by the lack of correlation of the paired sites used in the
study. One location 40 metres from a residence with a wood stove, experienced repeated episodes
of elevated PM2.5 concentrations, with a maximum 3-hour average of 150 µgm–3, and a maximum 24-
hour rolling average of 64 µgm–3, despite being in compliance with the United States
Environmental Protection Agency (US EPA) National Ambient Air Quality Standards (NAAQS) for
PM2.5. This highlights both the localised and episodic nature of elevated PM from domestic wood
burning.
Another important feature to note about available studies is that those based on chemical tracers
tend to be short term in nature, perhaps being gathered during a specific measurement campaign.
The use of chemical tracers and the analytical burden that they pose, makes it difficult to make
spatially detailed analyses of the long term trends in PM from biomass burning. A recent report
suggested that increased coverage of aethalometer networks could enhance our ability to track long
term changes in PM from biomass burning, since they can be run in a more automatic way and in
general have reasonable correlations with chemical tracer methods. 80 Several examples of long term
trend analysis of atmospheric particulates using aethalometers exist in the literature.81–83 Such
studies have facilitated the separation of the effects of localised emissions from wood burning from
particulate episodes which occurred due to long range transport.81,83 They have also provided data
on the diurnal and seasonal variability in ratios of BC to OC, which is important for understanding the
optical properties of carbonaceous aerosols, thus potentially reducing uncertainties within climate
models.84,85 From a policy perspective, availability of long term trend data is incredibly useful, since
the impacts from wood burning on ambient AQ are likely to be less important for breaches of
specific concentration targets, but will have greater effects with respect to PM2.5 exposure reduction
targets within for example, EU AQ directives. Long term aethalometer data can be used to
determine the time dependence of source contributions to carbonaceous PM. 86 Monitoring long
term trends that can be linked to sources, is necessary to tackle those that could impact most on
exposure reduction. So far, the WHO annual mean guidelines for PM2.5 have not yet been adopted as
AQ targets within Europe or the US. If they are in the future, then better spatial coverage of PM2.5
measurements will be required to assess the attainment of these targets, which will be challenging if
there are continued increases in domestic wood burning.
Estimating the contributions of biomass burning to atmospheric gaseous pollutants such as NO2 has
to be tackled in a different way since there is no means of identifying the source of individual NO2
molecules from atmospheric sampling. However, receptor modelling can be carried out by looking at
both receptor and source profiles for multiple species within a Positive Matrix Factorisation (PMF)
based approach.87 In such an approach, source types are identified by comparing them to measured
profiles providing fingerprints, and source weightings are used to determine how much each source
contributed to a measured atmospheric concentration sample. Liu et al. for example, applied a PMF
approach to estimate source contributions to a range of pollutants, including NO2, in the Chinese
Mega City of Tianjin.87 The biomass factor was identified by higher loadings of K, OC and BC and by
means of this footprint, the contribution of biomass burning to atmospheric concentrations of a
range of pollutants including NO2 could be determined. The performance and uncertainty of PMF
and other receptor models for source apportionment was reviewed by Belis et al.88 who concluded
that the accuracy of receptor models is sufficient for the needs of air quality management. PMF
approaches can be combined with trajectory modelling based approaches (source-oriented models)
in order to identify the geographical source regions that contribute to pollutant concentrations at
the selected measurement sites.89 This type of approach is more commonly used to evaluate the
long range transport of emissions from agricultural residue burning and outdoor fires rather than
more localised sources such as domestic wood burning.90,91
A limited number of studies have examined the effects of stove usage on levels of indoor air
pollution under real world usage in developed countries.92–95 Semmens et al. studied PM2.5 and
particle number concentrations and infiltration in 96 homes in Northwestern US and Alaska that
were using older model wood stoves as the primary source of heating.93 During two forty-eight hour
sampling periods they found daily average PM2.5 concentrations of around 29 μgm-3 suggesting
significant infiltration of PM with average indoor PM2.5 concentrations exceeding WHO ambient AQ
guidelines and approaching the US EPA 24-h standard equivalent (25 μgm-3 annual average).
Piccardo et al. found installation and operational issues impacted on indoor concentrations and that
following education, indoor exposures could be reduced.92 Allen et al. also examined the impact of
stove upgrades at 15 homes in northern British Columbia, Canada over a 6-day sampling period, but
they did not find a consistent relationship between stove technology upgrades and indoor AQ
improvements in homes where stoves were exchanged.95 A recent 30 day study within the UK city of
Sheffield, showed that even modern UK DEFRA-certified residential stoves can lead to short term
peaks in exposure to PM1 and PM2.5 within the indoor environment.93,94 These transient peaks in
PM2.5 could reach up to hourly indoor means of 47.60 μgm-3 and were linked to refuelling frequency
and length of burn period, indicating that opening of the stove door was a primary mechanism for
introducing PM into the room. These short term exposures to high concentrations may be significant
from a health perspective and the systematic review of Orellano et al. found a positive relationship
between short term PM exposure and cardiovascular, respiratory, and cerebrovascular mortality.96
It is clear from these studies that particulates from biomass burning are likely to impact on human
health, although separating out the health effects of particulates from biomass burning compared to
other pollutant sources within epidemiological studies, is even more difficult than obtaining spatially
and temporally detailed concentration profiles. As Naeher et al. point out, exposure to woodsmoke
is as old as humanity itself, but the notion that woodsmoke, being a natural substance, must be
benign to humans is far from the truth.57 Based on epidemiological data, Sigsgaard et al. estimate
that the contribution of biomass smoke to premature mortality in Europe amounts to at least 40 000
deaths per year as well as negatively affecting respiratory and, possibly, cardiovascular
health.10 However, they do highlight the difficulties in assessing overall population risk which relies
on the use of risk coefficients from large scale meta-analyses involving a range of particle
concentration metrics based on PM mass (PM10, PM2.5), particle number or absorption based BC
concentrations. Gu et al. suggest that estimated premature deaths are approximately ten times
greater when using a BC-specific risk coefficient compared to using PM2.5 based coefficients, and as a
result estimate 171,000 premature mortalities in China attributable on an annual basis due to
exposure to rural residential BC, mainly from the domestic combustion of solid fuels.97 However,
significant uncertainties remain in determining composition based risk factors.
As pointed out by Zelikoff et al., a range of health effects can result from different pollutants emitted
from biomass burning at levels measured within indoor environments of households using wood
fires and stoves.98 CO has been shown to produce carboxyhemoglobin and to increase the incidence
of angina among people with cardiac disease. NO2 can bind to haemoglobin, affecting several
enzyme systems, causing vascular membrane injury and edema, and may produce
bronchoconstriction in asthmatics at low levels. PAHs may be immunosuppressive and
carcinogenic.99 Aldehydes are associated with upper airway irritation, headaches and other
neurophysiologic dysfunctions, exacerbation of bronchial asthma, and possibly cancer. PM is
perhaps the component of most concern within emissions from biomass burning, since it is relatively
much higher than when combusting gaseous fuels for cooking and heating, and potentially has a
range of effects due to both particle size and chemical composition.
There have been few controlled human exposure studies to PM within woodsmoke but that of
Sällsten et al., found evidence of woodsmoke-associated systemic inflammatory effects in study
participants.100 Most of the available evidence on health effects in humans is epidemiological in
nature, particularly documenting respiratory effects in children. Zelikoff et al., summarises that
exposure of preschool children living in homes heated with wood burning stoves or open fireplaces
causes a range of effects including decreased pulmonary lung function in young asthmatics and
increased incidence and severity of acute bronchitis and acute respiratory infections. 98 Wider
epidemiological studies reviewed in Boman et al.101 and Naeher et al.57 have evidenced the link
between short-term exposures to PM from wood smoke and adverse health effects such as asthma,
respiratory symptoms, daily mortality, and lung function. Boman et al. in particular, compared the
relative risks of wood smoke with general estimations for ambient PM and adverse health effects,
and found that they were even stronger in studies where residential wood combustion was
considered a major source of PM. Evidence of associations with cardiac events are suggested to be
weaker than for impacts on respiratory symptoms.57 A possible weakness of wider epidemiological
studies, is the lack of information on personal exposures. Many epidemiological studies are based on
ambient concentration measurements of total PM2.5 or PM10, and as suggested above, these are
often from fixed monitoring sites which might not be representative of human exposure, particularly
when people move between different exposure environments during their daily routine and spend
large amounts of time in indoor environments.
The conclusions from epidemiological studies are however, supported by the outcomes of
toxicological studies. Many toxicological studies on animals have used acute single-exposure
regimes, since they are often carried out to support investigations of the potential impact of smoke
on fire fighters or smoke inhalation victims from fires. Such studies, as summarised by Naeher et
al.57, identified the irritant effects of woodsmoke through the generation of oxidative stress
producing an inflammatory response. In some cases, acute lung injury and pulmonary lesions were
found that are typically seen in smoke injured victims from fires. Of most relevance to ambient
exposures however, are more prolonged exposure studies at lower concentrations and these are
much scarcer within the literature. However, an acute exposure study by Fick et al. on rabbits
showed that woodsmoke had the potential to alter pulmonary immune defence mechanisms
potentially creating increased susceptibility to infectious lung disease.102
Lal et al. reported from a long term exposure study on rats, that pulmonary lesions were progressive
with repeated smoke exposures.103 Zelikoff et al., following several 1 h per day exposures of rats to
smoke concentrations more typical of concentrations found indoors during the operation of a poorly
vented fireplaces, found suppression of bacterial clearance from the lungs.98 This was particularly
related to particle concentrations, although they did not find evidence of lung pathology or lung cell
damage.98 This suggests that prolonged exposure to woodsmoke may compromise pulmonary
immune mechanisms that are important for protection against infectious lung pathogens. Longer
term studies on rats of up to 6 months exposure to smoke from an uncertified wood stove burning
mixed oak species found that effects on general indicators of toxicity, bacterial clearance, cardiac
function, and carcinogenic potential were mild. 104 They were however, similar to those found in
diesel exposure studies and included increased platelet number, reduced liver weight and increased
spleen weight.104 Animal studies have suggested that whilst lung tumour prevalence in mice exposed
to woodsmoke was higher than in the control group, woodsmoke is a weaker carcinogen compared
to smoke from coal burning which was coincident with lower concentrations of BaP in woodsmoke
compared to that from coal.105
A number of in vitro studies as summarised in Naeher et al. indicate the generation of reactive
oxygen species (ROS) as a result of exposure to woodsmoke, which are known to be linked to the
progression of inflammatory disorders and potentially pulmonary fibrosis.57 Several studies also
indicated that woodsmoke extracts were mutagenic in bacterial systems.106 Correlation analysis
carried out by Saffari et al., using PM2.5 samples obtained during a winter sampling campaign in
Thessaloniki, also indicated a strong association between ROS activity in rat alveolar macrophage
cells and the PM bound concentrations of levoglucosan, galactosan, and potassium, highlighting the
potential impact of wood smoke on PM-induced toxicity.74
Biomass resources available for energy production are diverse in their physical and chemical
characteristics with direct consequences on air pollutant emissions when combusted. Clean woody
biomass with relatively low amounts of moisture and ash would be favoured over potentially
variable lower grade fuels which may have higher N or ash fractions, high moisture content, be
larger in terms of fuel particle size or contain unwanted trace components such as heavy metals. 107
The pretreatment of solid raw biomass can be used to upgrade the properties of the fuels prior to
use in combustion systems. Methods include washing, drying, sieving, leaching, densification
technologies such as biomass chipping, pelleting and briquetting, torrefaction, and hydrothermal
carbonisation (HTC), or wet torrefaction. During torrefaction, the biomass is heated to 225-275˚C
driving off water and semi-volatiles which can be used as a fuel to drive the process. Torrefaction
also contributes to structural changes in the fuel through breaking hemicellulose, lignin, and
cellulose chains at elevated temperatures.108 The result is a high energy density fuel with up to 50%
mass loss, but with little total energy loss. Wet torrefaction can be used to process low-grade,
chipped, biomass feedstocks using pressurised steam, which after rapid decompression, results in
breaking apart biomass material, with the released moisture removing part of alkaline and chlorine
components from the fuel. Following pelletisation, the resulting fuel is hydrophobic, with high
energy density. Leaching with water can also reduce fuel K and chloride concentrations by up to 80%
and 90% respectively. Pretreatment methods can therefore help to diversify potential bioenergy
resources to lower grades of biomass, which may help to increase the use of waste derived products,
broadening the resource base, as well as providing better defined fuel specifications to assist with
the design and management of clean combustion devices. Excellent reviews on pretreatment
methods can be found in refs109,110 and the IEA Policy report.107
Pretreatment methods are not however, without impacts on AQ. A number of studies have shown
there to be the potential for VOC, CO and CO2 emissions from the storage of torrefied fuels, which
needs to be carefully managed from an AQ and safety perspective during transport and
storage.111,112 Composition, and thus potential toxicity of the VOC mix, has been shown to depend on
a variety of factors such as treatment methods and temperatures, moisture content and storage
time.112 Pelletisation is also not without impacts on AQ and could potentially be a major source of
airborne VOCs and PM unless adequate controls are implemented. Drax were recently served with a
$2.5 million fine for emitting three times the regulated limit for VOCs from a pellet plant in
Mississippi, indicating the difficulties in ensuring that effective control equipment has been
regulated and installed.113 A recent study suggests that in the South Eastern US, pellet plants are
significantly more likely to be placed within vulnerable and economically depressed communities114
and often have not been subjected to appropriate permit regimes.115
Pretreated fuels can be potentially used in all scales of biomass combustion. Pelletised and
briquetted fuels are more likely to be utilised in medium to large scale systems, or potentially even
domestic scale boilers where automatic fuel feeding systems are part of the operation principle. As
an example, the total consumption of wood pellets in the EU in 2015 has been estimated as 20.3
million metric tons, of which 36% was used for power generation and the rest to meet heating
demand.32 Pellet baskets also allow the conversion of fireplaces or wood-burning stoves for use with
pellets. For manually fuelled systems such as wood stoves and cook stoves, pretreated fuels can also
be manufactured into firelogs or briquettes to be used in place of traditional wood logs. These can
be made from a variety of waste starting materials including: paraffin wax and sawdust; woodchips;
agricultural residues such as olive stones, ground nutshells and seeds, bagasse, corn cob, palm fibre
and cotton plants; cardboard; and coffee grounds. The physical and chemical characteristics of such
manufactured logs will clearly differ from traditional wood logs and will influence their combustion
characteristics and emissions.116 There have been some limited studies on emissions profiles from
the use of such manufactured fuels that will be briefly covered in the following sub-sections.
7.2 Emissions Factors and Abatement Methods for Larger Scale Devices
The available technologies for the utilisation of solid biomass for energy and heat provision include
pyrolysis, gasification, and combustion. In future scenarios these could be coupled with CCS in order
to provide NETs as discussed in section 2.3. Among these, combustion is the most commonly used
technology for heat and power production. Gasification is also of interest for future applications
because of its potential to reduce emissions of relevance to AQ, particularly PM, in comparison to
conventional combustion. Biomass combustion systems are available from a few kW up to several
100 MW.
Unabated, biomass combustion systems are likely to exhibit relatively high emissions of NOx and
particulates in comparison to furnaces using natural gas or light fuel oil.117 As a result they can
contribute significantly to atmospheric PM, as well as to ozone, and NO2 formation in the ambient
air via secondary atmospheric processes. An LCA indicated that 38.6% of the environmental impact
of a modern automatic wood furnace is attributed to NOx and 36.5% to PM10.117 Hence effective
abatement technologies are required to reduce AQ impacts, although the technology readiness and
cost effectiveness of such technologies vary substantially for different systems. The scale of
appliance, as well as whether it is operated automatically or manually, have potentially the largest
impact on the level of emissions of the major pollutants for reasons explained in this section. Table 1
summarises the different scales of appliances with their operation principle, as well as the legislative
regime in Europe that specifies limits on their air pollutant emissions.
Figure 5. Comparison of European PM and NOx emissions limits for different scales of combustion
devices based on an equivalent energy basis (g/GJ) net heat input. The chosen test method within
the EDD is the ESP/Dilution tunnel method.119 LSH – Local Space Heater. Conversions are based on
an assumed stoichiometric dry flue gas volume of 253 m3/GJ net heat input adjusted for reference
oxygen content and a fuel CV for dry wood of 17.3 GJ/tonne (dry basis) based on the Digest of UK
energy statistics from 2014.120
The European regulated emissions levels for PM and NOx for different scales of combustion device
are summarised in Figure 5. For PM there is about a factor of 20 difference between the limits for
Local Space Heaters (LSH) controlled by the EDD, compared to large combustion plant which is
controlled under the IED, when considered on an equivalent energy input basis. On an equivalent
fuel input basis then, the utilisation of biomass within large scale units will lead to substantially
lower AQ impacts than its use at the domestic scale. The residence times and improved mixing
possible in larger combustion units allow the burnout of the carbonaceous part of the particulates
and hence prior to flue gas clean up particulates from larger units will mainly consist of submicron
particles of inorganic material118 formed as a result of fuel constituents such as N, K, Cl, Ca, Na, Mg,
P, S. Sippula et al. showed that PM1.0 mass concentration correlated linearly with ash content and
composition of the fuels studied, particularly for the major PM forming elements K, Na, Cl, S.121
Nussbaumer states that native wood is usually the preferred biomass source due to low contents of
ash and nitrogen, whereas herbaceous biomass such as straw, miscanthus, switch grass, etc., have
higher contents of inorganics such as N, S, K, Cl, leading to higher emissions of NOx and particulates,
increased ash, corrosion, and deposits.117 For large combustion plant, submicron inorganic particles
may still remain a challenge for pollution controls since particle control methods such as cyclones,
electrostatic precipitators and baghouse/fabric filters tend to have lower efficiencies for smaller
particles.118 Ceramic filters may have potential for removing smaller size fractions and have high
thermal durability, although in smaller devices could cause issues related to pressure drop.122,123
7.3 Emissions Factors and Abatement Methods for Medium Scale Plant and CHP
Medium combustion plant and large pellet boilers also have low limits on PM emissions controlled in
Europe by the MCPD and EDD, necessitating the use of advanced combustion systems and
potentially flue gas clean-up methods. For larger boilers, emissions can be partially controlled by the
design of the combustor, for example using two-stage combustion with injection of secondary air
prior to a hot combustion chamber. Electronic combustion control and forced ventilation is also
recommended in order to ensure optimum operation in practice, and to enable good mixing with
less dependence on ambient climate conditions such as air temperature and external air flows. 124
Smaller appliances are less likely to use automatic systems, and although under typical operations
they can achieve lower PM emissions than manual stoves, variations in operational conditions due to
variable loads may lead to poor combustion conditions and consequently higher PM emissions than
for automatic systems.123 For example, Poláčik et al. compared an automatic boiler fed by spruce
pellets with a manual stove, fed by beech logs, and found fine particle emissions (in the size range of
17–544 nm) of 173 mg kg−1 and 1043 mg kg−1 respectively during a similar test cycle.125 Lower grade
pellets may also have higher ash content and the review of Rabaçal and Costa concludes that PM
emissions from pellet boilers are strongly correlated with the ash content and ash composition.123 In
the UK, the Renewable Heat Incentive (RHI) scheme has supported the expansion of both
commercial and domestic biomass boilers by providing financial support at a set rate per unit of
renewable heat produced per kWh to the owner of the heating system, for a period of seven years.
RHI systems are subject to emissions limits as shown in Figure 5, and fuel must be sourced from an
approved supplier which should in theory provide some consistency in fuel quality. This is important
as several studies have shown significant variation in PM emissions and size fractions for different
pellet types with an example shown in Figure 6.126
Figure 6. PM emissions of the different types of pellets for all operating conditions from a domestic
pellet-fired boiler126. Reprinted from Fuel, 119, Garcia-Maraver, A.; Zamorano, M.; Fernandes, U.;
Rabaçal, M.; Costa, M., Relationship between Fuel Quality and Gaseous and Particulate Matter
Emissions in a Domestic Pellet-Fired Boiler, 141–152, Copyright (2014), with permission from
Elsevier.
NOx emissions from different biomass sources and combustion devices can also vary greatly,
depending on many factors including the N content of the fuel and combustion temperature.
However, the variation with scale of combustor is much less severe for NOx than for PM as
illustrated in the closer agreement between NOx emissions limits shown in Figure 5. Air staging and
fuel staging have been developed as primary measures for NOx reduction, particularly within
automatic systems, and can offer potentials of 50% - 80% reduction.117 If further mitigation is
required, post combustion technologies are available such selective non-catalytic reduction (SNCR)
and selective catalytic reduction (SCR) which rely on the reaction NO + NH 2 → N2 + H2O with urea or
ammonia injected as a source of NH2. At the smaller scale, downdraft boilers with inverse flow have
also been introduced which apply the two-stage combustion principle used in larger systems. 117 In a
review of abatement technologies for biomass systems, Nussbaumer concludes that effective
operation at optimum excess air and maximum efficiency is best achieved with advanced process
control requiring the application of sensors for λ (air–fuel equivalence ratio), temperature, and
unburnt substances across all device scales.117
CHP using biomass sources is one potential means of providing low carbon electricity and heat
through, for example, the further development of heat networks. However, the overall impact of
CHP on AQ may be complex. On the one hand, CHP provided using large scale boilers would provide
overall efficiencies greater than those of conventional power generation through combustion, and
could replace domestic scale heat provision by local boilers, potentially leading to lower overall
emissions of pollutants. However, this could be combined with local increases in key pollutants such
as NOx and PM, since CHP is also more likely to be located in regions of high population compared to
conventional power plant, due to the need to supply urban heat networks. The overall uncertainties
in providing projections in potential AQ benefits or dis-benefits of an expansion in CHP from biomass
are large, as pointed out by Apsimon and Oxley.127 Guidance planning from Environmental
Protection UK encourages the use of larger plant, for example in conjunction with the development
of heat networks, rather than a larger number of small plant, since abatement technologies become
more cost effective for larger plant sizes and are currently more highly developed for larger systems
than for smaller community and domestic scale boilers. The impact and cost implications of a range
of control measures for biomass boilers ranging from SCR of NOx, to cyclones and ceramic filters for
PM reduction, were discussed in a review of the AQ impacts biomass combustion by the UK Air
Quality Expert Group (AQEG). Whilst there are small CO2 penalties involved in the use of such
mitigation technologies and a range of potential costs involved, the potential for reductions in AQ
impacts may be significant as highlighted in Table 2. Relative costs per unit energy benefit from
efficiencies of scale for larger devices. High efficiency methods such as ceramic filters may merit the
costs involved for new installations in areas with existing poor AQ, suggesting more localised policies
on emissions limits may be useful, as part of planning or permit applications.124
Table 2. Effectiveness of post combustion emissions reduction strategies for solid biomass boilers
(adapted from AQEG)119
Gasification and pyrolysis techniques perhaps offer the best potential to lower particulate emissions
from biomass utilisation since they are based on converting solid fuels into more uniform gaseous
(or liquid) fuels, thus improving mixing and lowering the potential for particle formation. Currently
such technologies are largely used in plant > 2 MWth. Further developments are required at smaller
scales and a sustainable route for the utilisation of the remaining char material should be developed.
A recent study by Pröll and Zerobin128 compared several methods for biomass utilisation in a CHP
plant with a fuel power input of 66 MWLHV. These included fluidised bed and chemical looping
combustion connected to a steam cycle CHP plant, and rotary kiln slow pyrolysis of biomass for
biochar soil storage with direct combustion of volatiles supplying the steam cycle CHP plant: all with
and without CO2 capture. The purpose of this work was not to study AQ related pollutants but to
compare overall efficiencies and CO2 emissions per unit energy, against the least CO2-intensive fossil
fuel, namely natural gas. The study shows however, that low emission technologies do exist at this
scale and furthermore, the study makes recommendations for global climate change mitigation
strategies, concluding that both BECCS and the use of biochar for soil enhancement may be effective
NETs when coupled with CCS. The decision on which technological path to follow will depend on the
type of available biomass, energy markets, and ecosystem demands. Further studies covering wider
aspects of sustainability including AQ impacts would be useful.
In 2017, 43% of solid biomass energy in the EU28 was used in the residential sector.129 Small
domestic boilers and local space heaters (LSH) such as stoves and open fireplaces have significantly
higher regulated limits than for larger plant, driven by the lack of cost effective pollution control
methods at this scale (Figure 5). For LSH, the PM emissions limits within the Ecodesign Directive
(EDD, 2009/125/EC) are 20 times higher than those in the IED for large plant when compared on the
basis of equivalent net heat input119 (~7 g/GJ for large combustion plant vs. ~140 g/GJ for a modern,
clean domestic stove as shown in Figure 5). The short residence time and poor mixing of the fuel and
air in small units leads to significantly higher emission rates of particulates than for larger systems
where combustion technologies such as fuel pulverisation and fluidised beds may be utilised along
with flue gas control methods.46 From an AQ perspective therefore, small devices such as domestic
stoves and boilers are expected to pose a larger problem than large combustion plant.
Mitchell et al. compared NOx and PM emissions from the combustion of a range of biomass fuels
with coal and smokeless coal in a fixed bed 5.7 kW stove, that in this case was not designed to meet
the EDD limits, but does allow for a comparison between fuels.130 Cycle average NOx emissions
ranged from 85 gGJ-1 for torrefied wood up to 259 and 438 gGJ-1 for smokeless fuel and peat
briquettes respectively. Peat was the worst of the fuels in all respects. The fuel-N content of the
latter two fuels was shown to be significantly higher than for the biomass derived fuels, with fuel-N
shown to be more important than combustion temperature and residence time for NOx emissions in
this case (Figure 7). The phase of combustion was also shown to be of importance with flaming
combustion producing higher NOx emissions than smouldering combustion within the overall cycle.
Ozgen et al., recently provided an overview of NOx emissions from biomass combustion for domestic
heat production, including a useful summary of emissions factors found in the literature which are
shown to be highly variable. They also suggest a strong relationship between NOx emissions and
fuel-N content, although the conversion efficiency is suggested to exponentially decline with
increasing fuel-N.131
The Mitchell study showed that flaming combustion also produced significantly higher emissions of
PM than the smouldering phase, although the difference was reduced in the case of ‘smokeless’
fuels including torrefied wood.130 Emissions factors for PM2.5 ranged from 32 gGJ-1 for torrefied
briquettes up to 90 gGJ-1 for commercially available seasoned mixed hardwood and 210 gGJ-1 for
peat briquettes when using a heated filter based measurement method. Fuel type as well as the
morphology of the fuel are clearly critical, and relationships were found between average total
particulate emissions and volatile matter within the fuels. The surface area of the fuel particles
determines the rate of initial devolatilisation as well as the progress of the flame front and
combustion of the char formed.130 These influence the burning rate and consequently the
temperature in the combustion chamber impacting on emissions.
Figure 7. NOx emissions for various fuels plotted against fuel-N content for (a) flaming, and (b)
smouldering phases. Fuel specifications: 1. Commercially available seasoned mixed hardwood; 2. Air
dried hardwood (silver birch); 3. Torrefied spruce wood (bark-free) briquettes; 4. Peat briquettes; 5.
Bituminous coal ; 6. Biomass/coal blend; 7. Low smoke anthracite; 8. Smokeless fuel. Reproduced
under the terms of the Creative Commons CC-BY license from Mitchell at al.130
Figure 8. The routes to the formation of smoke from biomass or coal, or a blend. Reproduced under
the terms of the Creative Commons CC-BY license from Mitchell at al.130
Several studies have linked the footprint of pyrolysis products from biomass sources with
mechanisms for soot formation. Decomposition products, such as furfural and eugenol which are
formed from cellulose and lignin decomposition are highly sooting since they can contribute to the
main routes of soot formation, namely the HACA (hydrogen abstraction, carbon addition)
mechanism which dominates in the case of furfural, and via cyclopentadienyl formation and reaction
in the case of eugenol (Figure 8).48,132 Carbonaceous particles can agglomerate to form chains
together with inorganic aerosol such as KCl and fragments of char.130,132 A recent study by Du et al.
suggested that that emission factors of condensed organic compounds in a domestic stove increased
with increasing KCl content for KCl impregnated fuels due to increased heterogeneous condensation
of semi-volatile hydrocarbons on pre-existing ash particles.133 Elsasser et al. found that the organic
fraction of PM emitted during the ignition phase was the main contributor to overall emissions in
their logwood stove study.134 Torvela et al. found that impaired combustion conditions increased the
level of condensable organics within the PM in a 40 kW grate combustion unit with PM1 emissions of
up to 160 gGJ-1.135 Pettersson, also reported significantly higher emissions of total PM for a log
burning stove under conditions of poor air supply and high firing intensity. PM was dominated by
fine particles (200-320 nm) and emissions factors varied in the range 38−350 g/GJ-1. Trubetskaya et
al. recently studied PM emissions factors from the combustion of a range of fuel types in both a
conventional and an EDD compliant wood stove.108 They found that the use of secondary air could
reduce PM from wood logs by about two-thirds due to improved mixing of fresh air with preliminary
combustion products. Vincente and Alves68 reviewed a wide range of studies to compare derived PM
emission factors from biomass combustion based on different measurement techniques. The values
for fireplaces ranged from 161-1610 gGJ-1 and from 20 up to 1433 gGJ-1 from woodstoves, indicating
an improvement for stoves over open fires. Pellet boilers were again an improvement over domestic
stoves (6-377 gGJ-1).
All of these studies highlight the huge variability of emissions factors depending on fuel type/quality,
and appliance design and operational practice. This also suggests that real world emissions may
deviate from those obtained under operating conditions during test cycles obtained for compliance
and research purposes. Scott, for example,136 compared emissions from the same appliances under
“real world” conditions as simulated within a laboratory, and under real world conditions where the
wood burners were operated normally by the householder using their own firewood supply. Their
results suggested that real world emissions may be up to 4-5 times those achieved within legislated
test cycles and laboratory “real world” studies. Coulson et al. suggested that in situ studies which
take account of the “operator” variable are essential for understanding real world emissions of
manually controlled devices.137 They presented a review of such in situ studies for New Zealand
which highlighted the variability in emissions factors, and showed that only 22% of stoves met their
legislated limits in the real world. Wood moisture, flue temperature and oxygen supply were the
major factors affecting emissions rate variability. Fachinger et al., conducted a real world
comparison study of a residential pellet stove and a wood stove using 11 hardwood species. They
suggested that real world emissions factors from pellet stoves can be expected to be closer to
laboratory-derived emission factors than those for wood stoves, due to lower dependence on user
operation.138 Typical ranges of emissions factors from average to best to worst as reported by
Nussbaumer are shown in Figure 9 for smaller devices such as pellet boilers, pellet stoves, log wood
boilers, and wood stoves. It is clear that domestic stoves present the highest emissions factors, often
reaching several hundred g/GJ and far in exceedance of the EDD limits. Such stoves have long
lifespans and are likely to remain in operation for many years unless schemes incentivising
replacement are offered.
Figure 9. Average emission factors (best, typical, worst) for pellet boilers, pellet stoves, log wood
boilers, and wood stoves. All data indicated as solid particles except for wood stoves, where data
including condensables (SPC) or measured in dilution tunnel are given in the last column.
Reproduced from Nussbaumer et al., Overview on Technologies for Biomass Combustion and
Emission Levels of Particulate Matter, Proceedings of the 16th European Biomass Conference &
Exhibition, Vol16, OA 9.2, Copyright (2008) with permission from ETA Florence Renewable
Energies.124
Nussbaumer suggests that wood stoves can be developed with designs that provide a hot
combustion chamber with good mixing. However, stoves are partially designed to provide visible
flames in order to achieve the “homely” feel desired by the customer. Operation without forced
ventilation therefore gives limited possibilities for improved combustion principles. He suggests that
designs should be developed that prevent improper operation of the stove such as over-fuelling and
the possibility to reduce the air inlet.124 Elsasser et al. for example, showed that an overloaded stove
emitted roughly 4 times the mass of PM compared to a stove operated according to the
manufacturer’s recommended conditions.134 The same stove also showed much higher emissions
factors for PAHs for over fuelled operation.
Fuel quality has also been shown to influence PM emissions. Emissions of PM and to a lesser extent
NOx from a fixed bed 5.7 kW stove with a single combustion chamber were found to be lower from
the combustion of torrefied fuels compared to raw biomass from spruce, willow and olive stone
sources by Maxwell et al48. However, even for the torrefied fuels, NOx emissions ranged from 45-90
g/GJ. Moisture content has been shown to particularly influence PM emissions with freshly cut, high
moisture woods giving higher PM emissions than kiln dried woods (small logs) for the soft wood
spruce within a study of a 5.7 kW domestic stove.139 Kiln dried fuels were shown to lead to higher
combustion temperatures and rates of combustion, particularly during the early phases of
combustion. However, oxygen supply was also shown to be critical, and insufficient air supply led to
high PM emissions for logs of the kiln dried hard wood beech. Calvo et al. also found higher PM
emissions from hard woods (eucalyptus), compared to pine soft woods.140
There have been relatively few emissions studies of manufactured fuel logs and briquettes in
domestic scale devices. However, Mitchell et al. recently compared emissions from briquetted
bagasse, miscanthus, wheat and barley straw with low moisture wood logs and wood briquettes. PM
emissions factors for the bagasse and miscanthus briquettes were found to be lower than those for
the straws and comparable to the wood briquettes. The study of Trubetskaya et al. compared PM
emissions of torrefied olive stone briquettes, smokeless coal, sod peat, wood logs and bituminous
coal.108 The torrified olive stones and smokeless coal gave lower emissions factors than the other
fuels but the differences were dependent on the test method used (hot filter sampling of the raw
flue gas vs. sampling of cooled, diluted flue gas). A large scale study of spatially resolved emissions
factors for a region in rural China suggested slightly lower PM2.5 emissions per Kg of fuel burned from
processed biomass pellets than from uncompressed straws, although overall the emissions factors
for biomass were slightly higher than those for coal.141
The UK implemented a “Ready to Burn” certification scheme in May 2021 through the Air Quality
(Domestic Solid Fuels Standards) (England) Regulations 2020, which bans the sale of wood with a
moisture content of > 20% in amounts of two cubic metres or more. France is also planning a similar
labelling scheme certifying wood of low moisture wood and sustainable forestry origins. 142 The US
EPA manages the Burn Wise program with the aim of providing information to the public on how to
store and dry wood, as well as lists of EPA certified appliances and appropriate installation and
maintenance regimes.143 Schemes such as these, and the penetration of stoves built to EDD/EPA
specifications may partially offset the increase in stove sales that has occurred over the last decade.
However, as appliances are slowly replaced by newer EDD compliant versions, it will be important to
conduct real world emissions studies to establish whether they meet regulatory limits during every
day usage and for typical ranges of fuel types used.
Strategies for the mitigation of emissions from small scale devices are focussed in two main areas: i)
improvement of the combustion process and/or fuel quality improvement, or ii) flue gas cleaning
technologies. Primary measures address issues around air supply through, for example, staged
combustion, or by pre-heating the primary combustion air.68 Fuel improvement by processing as
discussed above, or through the use of additives can also be used to reduce alkali compounds,
decreasing the formation of inorganic particles.144 Several options exist for flue gas removal such as
electrostatic precipitators (ESPs), condensing scrubbers145 and catalytic converters.
The difficulties of controlling PM from domestic scale biomass combustors were highlighted in a
report prepared as part of the IEA Bioenergy project.146 The report compares efficiencies of particle
collection devices for residential biomass systems (<50 kW) including ESPs, catalytic converters,
ceramic filters and condensing heat exchangers. The scarcity of available PM control systems at the
domestic scale that can be retro-fitted is of particular concern. The challenge is technically more
difficult than, for example, the use of diesel particle filters, due to the fact that small scale domestic
appliances tend to rely on natural draughts for air supply. Pressure drops can be induced by flue
inserts such as catalyst carriers which can affect primary combustion conditions and thus influence
emissions.147 IEA Bioenergy Task32 found no promising results from catalytic converters for wood
boilers and stoves. The high flue gas temperatures required for catalytic oxidation, mean that these
devices are typically not operating effectively during start-up where the highest emissions often
occur. They can also suffer from fouling and clogging due to tar and soot emissions and hence
cleaning regimes would be vitally important.68 The pressure drop of ceramic filters or catalytic
converters was found to negatively influence the combustion behaviour of natural draught systems.
A recent study by Klauser et al. did however, see an effect of flue gas catalytic removal in reducing
emissions of high molecular weight PAHs from burning beech wood in a domestic room heater,
albeit at the expense of slightly increased total PM emissions. 148 It was postulated that the catalysts
might have led to oxidative formation of semi-volatile species, thus contributing to overall PM
emissions. Wöhler et al. compared the use of a foam ceramic element, a catalytic active coated foam
ceramic element, and a honeycomb catalyst for the potential reduction of emissions of CO, PM and
OC from a domestic wood stove.149 The honeycomb catalyst showed the highest potential of systems
tested, leading to up to a 33 % reduction in PM emissions, although the effectiveness of all devices
was found to be lower in the start-up phase of the stove. Reichert et al. recently demonstrated that
the use of honeycomb catalysts could be effective at achieving emissions rates stipulated for wood
stoves in the EDD 147 under real-life related testing conditions. Hence more recent studies are
starting to show some promising results for the use of catalysts, but the literature remains
inconsistent and further studies are required.
IEA Bioenergy Task32 found ESPs to be the most promising technologies for boilers, with collection
efficiencies for PM of up to 85%. Costs can be reduced by utilising a metal chimney tube as the
collecting electrode, but particle removal would need to be done manually by the operator or a
chimney sweeper for most of the examples considered, and hence efficiencies will depend on
maintenance regimes. Most of the devices considered, seemed to be effective for the removal of
PM1 which is promising in terms of reducing respiratory effects of PM. However, where there are
high emissions of condensables, the aerosol load can be higher at the filter outlet compared to the
inlet, and the use of ESPs may therefore alter the PM composition68 potentially changing the toxicity
of the emitted particles.150 At an estimated capital cost of 1000 to 3000 €, ESPs would approximately
double the price of a domestic stove installation which may impact on voluntary uptake. More
recent examples of ESPs (e.g. Exodraft151) offer chimney integrated ESP particle filters with a
chimney fan, and are starting to reach near market status. These have the aim of ensuring a constant
up-draught in the chimney, thus improving air supplies and preventing the backflow of smoke into
the room. Automatic particle removal is also claimed which would potentially reduce the influence
of poor maintenance regimes.
Without such mitigation options, domestic stoves are likely to continue to make a substantial
contribution to particulate emissions in both urban and rural areas and research into cost effective
retro-fit PM control systems at this scale is greatly needed. Stricter emissions limits tend to
accelerate technological development and would encourage the market introduction of, for
example, particle precipitation or catalytic devices. A common international approach regarding PM
emission measurements is also urgently needed to assess the comparative efficiency of control
technologies, particularly in real world settings.146
As discussed in section 4, PM emissions from biomass burning are compositionally complex including
a range of organic and inorganic species such as heavy metals, PAHs, dioxins and furans. There is a
substantial literature which shows that the combustion of biomass, and wood in particular, is a rich
source of PAHs, both in the gas phase and bound to PM, which may arise via pyrosynthesis. PAHs are
of particular concern because of the known carcinogenicity of the mixture. McDonald et al. 54 derived
an emissions estimate for total PAH emission rates from wood stoves as 74.7 mg/kg fuel.
Predominant species were found to be acenapthylene, naphthalene, anthracene, phenanthrene,
benzo(a)pyrene and benzo(e)pyrene. Rogge et al. found that the smoke from the burning of
synthetic logs had the highest total PAH emission rate and the most diverse range of PAHs present
when compared to burning pine and oak.56 The review of Vicente and Alves68 and the detailed wood
stove study of Pettersson et al. using birch, spruce and pine logs, provide a detailed discussion and
emissions estimates for PAHs respectively.152 One striking conclusion of Pettersson’s study is the
high variability of emissions factors for PAH, with values ranging from 1.3 to 220 mgMJ-1fuel. In most
cases the mass of PAH was dominated by phenantrene, fluoranthene, and pyrene, and mostly found
in the particle phase.152 Factors influencing the level of PAH emissions included wood type, and the
same types of operation conditions that also affected overall PM emissions such as oxygen
availability and burn rates.
The review of Lavric et al. provides a summary of emissions of dioxins (PCDDs) and furans (PCDFs),58
concluding that as complete as possible combustion is critical for a low emission level of dioxins,
mainly depending on combustion chamber temperature146, turbulent burning rates, residence time
and excess oxygen. Zhang et al. showed that combustion mode: either flaming or smouldering had a
considerable influence on the amounts of dioxins emitted from the system. 59 An analysis of soot
scrapings from the chimneys of wood-burning stoves operating in various regions of the United
States showed that average total PCDD/F levels in the chimney deposits were 8.3 ng/kg in the
eastern region, 42.1 ng/kg in the central region, and 10 ng/kg in the west153 with differences
attributed to the wide variability in device design affecting sampling points and/or possible
contamination of the fuel wood. The burning of waste wood found in urban areas has been
highlighted as leading to significantly higher levels of dioxins in the fly ash from a range of
combustors, likely due to the use of treated, varnished or PVC-coated wood.58 Tame et al. reviewed
the role of preservatives in the formation of dioxins from the domestic combustion of wood. They
found that preservatives such as copper could influence dioxin formation via acting as catalysts, as
well as promoting the smouldering of wood char following the cessation of flaming, proving the low
temperature environment needed for dioxin formation. Chlorinated organics, added as secondary
preservative components, were also found to yield dioxin precursors on thermal decomposition.154
The disposal of preservative impregnated timber via combustion, through for example domestic
bonfires, is clearly hazardous, even when arsenic is not present.
The ratio of Black Carbon (BC) to Organic Carbon (OC) within total carbon (TC) emissions is also
important in relation to the relative impacts on climate forcing, since BC is a major warmer, but OC
aerosol is thought to have a cooling effect on the climate as discussed in section 2.1.16 Mitchell et al.
showed that the BC/TC ratio varied during the combustion cycle for a 5.7 kW domestic stove burning
a variety of biomass fuels. A higher ratio of ~0.4 was found during the flaming mode for typical wood
logs, with a lower value during smouldering combustion.130,155 They also point out the importance of
measurement technique for determining this ratio, since measurement temperature clearly affects
the collection of semi-volatile organics. Fine et al. found that the BC content of fine particle
emissions for woods burned in the NE US, ranged between 3-7% although, white pine and paper
birch produced much higher ratios due to high sap and bark content respectively. 51 Fernandes et al.
found that the BC content of PM from an energy-efficient "chimney type" logwood stove was
substantially higher than from a conventional cast iron stove and fireplace, with the opposite
observed for the OC fraction.156 Sun et al. present BC and OC emissions factors for a range of
biomass fuels for PM emissions from mini-boiler stoves under close to real world operation, finding a
dominance of OC, but variability between fuel types.157 Clearly there is a large variability in BC/TC
ratios of PM from different biomass sources and appliances which would affect their climate
impacts.
Variability in emissions of the GHG methane was also noted in the simulated real world study of
several residential heating biomass appliances (6-11 kW) by Ozgen and Casarini.158 Using a GWP
based method, they compared the CO2 equivalent emissions of methane with those of N2O, CO,
NOx, VOC, and BC emitted by the same appliances, and showed that the warming impact of
methane was lower than that of BC and CO, but was still an important fraction of the CO2 avoided by
the substitution of fossil fuels with biomass. Importantly, they concluded that within a 20-year time
frame, the CO2 eq for all the non-CO2 forcers offset the CO2 benefits of biomass use.158
As discussed in section 4, VOC emissions can also affect the formation of secondary smog
components such as O3 and NO2 as well as leading to the formation of SOA. Sun et al., based on a
study of traditional and clean domestic heating appliances in northwest China, suggest that the
compounds with the greatest contribution to O3 formation tend to be alkenes, and to SOA,
aromatics.159 They found significant differences in emission profiles between coal and biomass, with
wood giving the highest mass based emissions of non-methane VOCs (NMVOCs) and anthracite the
lowest. Volatile matter content and combustion efficiency of the appliances were the two biggest
factors affecting NMVOC emissions factors. Poor air supply and poor mixing were prevalent in
traditional stoves and significant improvements were seen for new devices. NMVOCs yielded by
traditional stoves were five times higher than that from semi-gasifier stoves. Pettersson et al. found
that emissions for NMVOCs varied from 17 to 2300 mgMJ-1fuel in their domestic stove study, with the
highest concentrations being of ethene and acetylene generally followed by benzene, propene,
ethane, and toluene.152 Emissions were highest during intensive flaming combustion. Emissions for
1,3-butadiene, of relevance to AQ target compliance due to being an identified carcinogen, varied in
the range of 0.1−61 mg/MJ and were again highest in the intensive combustion phase. McDonald et
al. also found elevated emissions of 1,3-butadiene and 1,3-cyclopentadiene, for wood combustion
relative to other sources and suggested that they probably arise from the decomposition of
terpenoids.54 McDonald et al. carried out a highly detailed study of both particle and gaseous phase
emissions from the combustion of US softwoods, hardwoods, and synthetic logs in fireplaces and
hardwoods burned in a wood stove, providing emissions data for over 350 elements, inorganic
compounds, and organic compounds.54 They found high emission rates for alcohols, particularly
methanol, alkenes, C2-C9 carbonyls and aromatics. Evtyugina et al., in their study of VOC emissions
from the combustion of European woods in fireplaces and wood stoves, found aromatics to be the
most abundant group, followed by oxygenated compounds and aliphatic hydrocarbons. The relative
emissions factors for individual VOCs were found to be strongly affected by the wood composition,
the type of burning device and operating conditions.160 Many previous studies of solid fuel
combustion have focussed on VOC emissions from coal as a fuel source and more information is
needed on VOC emissions from biomass, particularly since emissions profiles seem to be very
variable depending on fuel type and combustor design and operation.
From an AQ perspective, the impact of a new biomass burning device installation will very much
depend on what it replaces. Figure 10 shows emissions factors for the domestic combustion of
various fuels used within the 2018 version of the UK National Atmospheric Emissions Inventory
(NAEI). The PM emissions for wood and peat are particularly high, and are representative for the
types of stoves prevalent in the UK for this year i.e. in advance of the adoption of limits set in the
EDD. The values are estimates, but are used here to illustrate the potential influence of fuel or
device switching. If a modern EDD compliant device is replacing an open fireplace or an old stove for
the delivery of the same heat, then improvements in emissions, particularly for PM are likely to
result. Comparison of the emissions factors for wood in Figure 10 with those from the EDD in Figure
5 suggest at least a halving of PM emissions. Similarly, switching from peat to wood as a fuel would
have large benefits for PM emissions. However, in many cases wood stoves and/or domestic boilers
could be replacing gas fires or gas/oil boilers as a source of secondary heat, and in this case there are
likely to be dis-benefits for both NOx and PM. A comparison of NOx emissions factors within the
literature for various biomass fuels compared to fossil fuel alternatives shown in Figure 11 also
reinforces this point, particularly for non-woody biomass.131 This recent review of NOx emissions
also highlights that not all of the tested appliances in the literature meet the EDD regulation limit,
i.e. that there will continue to be a problem with legacy devices, unless suitable retro-fit mitigation
options are developed.
Figure 10. Emissions factors expressed in g/GJ (net) for different fuel types as included in the 2018
National Atmospheric Emissions Inventory (NAEI) for the UK. Note, although the majority of NOx
mass emissions are likely to be NO, since NO is subsequently converted to NO2 in the atmosphere,
they are often reported as NO2 as is the case in the NAEI.
Figure 11. Comparison of the literature residential heating experimental emission factors for
biomass combustion and common fossil fuels.131 Reprinted from Renewable and Sustainable Energy
Reviews, 135, Ozgen, S.; Cernuschi, S.; Caserini, S., An overview of nitrogen oxides emissions from
biomass combustion for domestic heat production, 110113, Copyright (2021), with permission from
Elsevier.
It is perhaps not surprising that most technologies and strategies proposed for the reduction of GHG
emissions are aimed at the industrialised or rapidly industrialising world. However, tackling problems
related to the domestic use of solid fuels in developing countries could provide significant benefits
to both health-related air pollution problems, as well as climate change. It was estimated in 2000
that approximately 50% of the global population, mostly in developing countries, relied on coal and
biomass in the form of wood, dung and crop residues for domestic energy. 14 The IEA suggests some
progress since 2000, with the number of people in developing countries with access to clean cooking
(LPG, natural gas and electricity), growing by 60% compared to 2000. 161 However, they state that a
third of the world’s population – 2.5 billion people in mainly sub-Saharan Africa and Asia – still rely
on the use of solid biomass for cooking. Data from the 2017 IIASA Gains model suggests that
anthropogenic BC emissions within the African continent are dominated by biomass usage for
cooking (>50 % of the total). Within Asia, a mixture of biomass cook stoves and coal stoves used for
cooking and heating dominate the BC emissions profile. Much of this usage is likely to involve fuels
being burnt in simple, often open stoves, although they can vary greatly in terms of design and
construction material.162 Traditional cooking systems are based on burning coal or firewood on
stoves with cooking pots placed on top, resulting in low thermal efficiencies and therefore excessive
fuel consumption and high emissions.163 They are usually naturally ventilated, without scope for
controlling air supply, leading to incomplete combustion in unvented indoor environments. As a
consequence, women and young children are exposed to high levels of indoor air pollution on a daily
basis. Although uncertain, estimates suggest that exposure to indoor air pollution may be
responsible for nearly 2 million excess deaths in developing countries. Causes of death include
COPD, tuberculosis, stroke, heart disease, respiratory infections, pneumonia and lung cancer.14,164–166
The same studies estimate that over 3% of the global burden of disease can be attributed to indoor
solid fuel burning, a level comparable with tobacco smoke and higher than the 1% attributed to
urban outdoor pollution.57 The WHO estimate that close to half of deaths due to pneumonia among
children under 5 years of age, are caused by PM inhaled from household air pollution.165
Indoor exposures from cook stove emissions can far exceed even heavily trafficked outdoor
environments. A study of the impacts of traditional stove usage on lung functions in adult women of
a rural Indian village, found average indoor levels of PM2.5 as high as 728.90 ± 50.20 μgm-3 from the
use of biomass compared with 99.76 ± 41.80 when using LPG.167 This led to respiratory symptoms
such as wheezing and chronic cough being significantly more common in the biomass group
compared to the LPG control group. Concentrations of up to 2000 μgm-3 were found during cooking
in biomass-using households in studies of rural households of southern India168 and in Mexico.169
Measurements of indoor AQ in 23 houses in Costa Rica170 using wood burning stoves indicated that
average daily PM10 levels reached up to 132 μgm-3 (c.f. European standard of 50 μgm-3). Peak PM10
levels during cooking reached 18,900 μgm-3. The effects of such extreme levels of short term
exposure are not well understood. However, some studies suggest that high numbers of particles
may overload the lung macrophage system and thus short terms peaks may be as important as
lower exposures over longer time periods.44 Ventilation conditions were found to be the most
significant factor determining exposure by Nayek and Padhy171 implying that huge benefits to global
health could be achieved by improvements in stove technology and building ventilation. Jaakkola
and Jaakkola suggest that sustainable development calls for interventions that replace traditional
biomass fuels with more processed, cleaner fuels, and that improve stoves, ventilation, and housing
conditions in an affordable way that is easy to implement.172 In the context of the Paris Agreement
however, cleaner fuels should not be fossil based going forward.
Within developing countries, particularly in rural environments, domestic fuel burning for cooking
and heating is likely to remain a significant environmental challenge that must be addressed in order
to improve public health. Reductions in air pollution-related disease burden (both for household and
outdoor) is being used to monitor the progress towards attaining the Sustainable Development Goal
on Health (SDG 3)165 and mitigation strategies are vitally needed. In addition, ensuring universal
access to clean fuel and technologies is a target of the Sustainable Development Goal on energy
(SDG 7). Whilst electric based cooking might be the cleanest alternative, it is rarely used in the
developing world due to the high costs of electricity and limited access in rural areas. Hence, where
biomass fuel sources are to be used, research priorities should focus on vastly improving both
efficiency and AQ impacts. Large scale penetration of efficient small scale technologies for the
effective burning of biomass fuels is therefore required with potential benefits to health and GHG
equivalent emissions such as BC. Bhattacharya and Salam for example, suggest that the use of wood
stoves with improved efficiency could reduce GHG emissions by at least a factor of two over
traditional stove technologies162 as well as relieving pressure on vital biomass fuel resources.
The Clean Cooking Alliance was established in 2010 and works across a global network of partners
with the aim of making clean cooking accessible to millions of households.173 Simple mitigation
measures such as encouraging a more open ventilation environment and using forced draught
stoves or fan assisted stoves174 with mixed biomass solid fuel sources can lead to improvements in
indoor levels of PM2.5 compared to traditional stoves in more enclosed spaces, with 20–80%
reductions in concentrations measured in a rural Indian study.175 The use of pellet stoves, as oppose
to burning logs and unprocessed biomass, has also been shown to lead to improvements in indoor
AQ, although a study in Chile suggested that social barriers to such interventions exist, even when
financial burdens are removed.176 Switching to gaseous fuels would also lead to substantial
reductions in PM as indicated in the study of Pathak et al. mentioned above.167 However, LPG is a
petroleum derived fuel and thus a poor choice from a climate perspective. The use of biogas rather
than solid fuels has been suggested as a possible route for rural communities. This could be based on
micro-gasification technologies at the individual stove level,163,177 on anaerobic digestion or on
community scale biogas production which has become relatively well established in China178 and
India.179 Stoves with up to 55% efficiency are available for biogas combustion offering lower GHG
emissions than traditional designs. Gasification also offers a potential route to the reduction in open
agricultural waste burning and would potentially reduce unnecessary sources of PM and CO2
emissions if such fuels could be burned to provide useful energy.163 Biogas can also be produced
from the anaerobic fermentation of organic matter from waste products such as agricultural
manure, food waste as well as landfill gas.180
The potential success of using biogas as a fuel depends on the availability of appropriate gasification
or anaerobic digestion technologies. However, barriers to wide scale implementation need to be
addressed which are not necessarily technical in nature, and can include issues related to
infrastructure, financial and perceptual barriers.181 Economic viability is also a key issue which is in
part related to gasification and combustion efficiencies. Several studies identify solving problems
related to secondary environmental pollution as key to the wide spread implantation of biomass
gasification in developing countries.178 Methods for improved tar removal could remove the need for
waste water treatment methods, improving efficiency and reducing costs. There have been a
number of studies on the use of catalysts for the removal of tar during gasification with scope for
further research at the practical scale.182,183
There are a growing number of studies investigating the development of gasifying stoves, although it
is claimed that such designs may have already existed 300 years ago.184 Such stoves are based on
two stage combustion with a fraction of air introduced into the lower part of the reactor in order to
gasify solid biomass into syngas. Secondary air is introduced at the top of the reactor, to ensure
complete combustion. Both natural draft or forced draft designs have been developed. Perez et al.
investigated a 3 kW forced air multi-fuel gasification stove using a range of solid fuels including
briquettes made from the densification of agricultural solid wastes.163 They found significant
reductions (84-93% for high and low power operation) in fine particulates compared to traditional
stove designs. They also noted a decrease in fuel usage of around 60% due to the improved
efficiency of the stove design. Gitau et al., found reductions in indoor concentrations of CO, CO2, and
PM2.5 by 57%, 41%, and 79% respectively when using a gasifying stove compared to using a three-
stone open fire.185 Mitchell et al. compared emissions from a raised grate charcoal stove, a rocket
stove using dry wood, wet wood and straw, and a gasifier ‘top lit’ down-draft stove using wood
pellets.186 The gasifier-type stove was found to emit the lowest levels of all mass based pollutant
emissions, although high particle numbers were found which may require further investigation. A
more recent study of this gasifying stove found a low signal from Laser induced incandescence (LII)
measurements, indicating that the ultrafine particles seen using a fast particle analyser, were
composed of inorganic or tar particles that would vaporise at temperatures too low to generate an
LII signal, rather than of soot.187 KCl is suggested as the most likely component, but clearly further
work on ultrafine PM from gasifier stoves for a range of fuel types would be useful, since
composition would clearly affect potential toxicity. A number of studies have addressed design
principles of gasifying cook stoves with Kirch et al. for example188 indicating that low primary air
supply rates with utilisation of the resulting char for subsequent applications, may be beneficial from
an air pollution point of view. 188
The AQ, epidemiological and toxicological evidence raises questions as to whether the domestic
burning of biomass is environmentally sustainable, since without suitable mitigation technologies,
unacceptable exposure to harmful pollutants can result both within indoor and outdoor settings. In
terms of combustion efficiency and the ability to control harmful emissions, especially of PM, large
scale combustion of biomass in power stations or CHP units makes more sense. However, such large
scale utilisation puts pressures on biomass supply chains, which must utilise carbon neutral sources
involving sustainable replanting or the use of genuinely waste materials. Even then, the timescales
involved in carbon sequestration of large trees pose problems for carbon accounting and for
achieving shorter term carbon reduction goals.
In terms of domestic use, even in industrialised countries, it is difficult to envisage legislation that
severely limits an individual’s choice to burn solid fuels. Indeed, attempts made to ban wood burning
in Paris in 2014 met with fierce public opposition and plans were scrapped only days before the ban
was due to come into force.189 Other mitigation measures such as the prohibition of wood burning
on open fires, tightened emissions limits for new stoves, and regulations on fuel quality may have
some impact, but their success will depend on the degree of enforcement, as well as turnover rates
for old appliances and the energy system that they are replacing. Turnover rates could potentially
be encouraged through grants and scrappage schemes such as those introduced in Denmark and
Norway190, but even an ultra-modern domestic stove meeting the EDD emissions limits, is still likely
to emit as much PM per hour as 6 Euro VI heavy goods vehicles.119
If we are not as a society willing to ban the use of solid fuels in domestic settings, then the
development and mandatory use of both primary and secondary control technologies will be vital to
limiting AQ impacts from biomass burning. These could include primary measures such as
restrictions on fuel quality, education programmes on appropriate device usage, and designs that
limit poor operation, as well as secondary measures such as domestic scale ESPs and catalytic
removal methods. Fuel treatment methods provide further scope for reducing emissions by
removing moisture, semi-volatiles and inorganic components from the fuel, but emissions from
treatment plants themselves need to be properly regulated. Further research is necessary to
establish the effectiveness of all these mitigation options for a range of biomass fuel types, in terms
of speciated emissions both within gas and particulate phases, and within real world as well as in
laboratory based settings. Again, we can look to the automotive sector to realise the importance of
real world testing, where based on a programme of real world tests, the US EPA revealed that
Volkswagen had intentionally programmed diesel engines to activate NOx emissions controls only
during regulatory laboratory emissions testing.191 As a result, NOx emissions in the real world were
substantially higher and did not meet regulatory limits. Amongst other things, the diesel scandal
highlights the challenges in designing pollution mitigation systems that simultaneously reduce
emissions of relevance to both climate change and AQ.
For the biomass sector, further challenges remain for the mitigation of semi-volatile emissions and
ultrafine particles, as well as for gaseous emissions that lead to the formation of atmospheric ozone
and SOA. Previous research has demonstrated an extremely high level of variability in emissions
factors for manually operated small scale devices, although there are clearly some trends related for
example to fuel quality/moisture content, ash, trace species and volatile content, wood
contamination, air supply, combustion mode, fuel supply rates and firing intensity. A better
understanding of speciated emissions factors under different mitigation options used over a range of
operational conditions and fuel qualities, could assist the development of guidelines on fuel quality,
as well as localised emissions limits or “no burn days” that could potentially be used in areas with
existing poor air quality, or during periods of unfavourable meteorological conditions such as low
wind speeds or limited vertical mixing due to temperature inversions.190
Enforcement of any future policies is a key issue and requires sufficient resources in order to be
carried out effectively. Most urban areas in the UK for example, have been designated as smokeless
zones as part of the Clean Air Act. This means that indoor burning of wood on open fires is already
prohibited. However, the recent survey carried out on behalf of DEFRA, suggests that this practice is
still widespread, implying that insufficient resources are available within local government to
enforce the ban.76
Long term ambient monitoring strategies that can identify the contribution of biomass burning,
particularly to PM, are needed within residential areas, and would help to develop a better
understanding of concentration trends, as well as the impact of any technological and policy
interventions. Improved monitoring is also needed within indoor environments, as well as at the
level of personal exposure, in order to support epidemiological studies on health outcomes that are
able to couple exposure measurements with symptoms tracking to establish causal links. An
improved understanding of exposure to pollution from biomass burning and the resulting health
impacts may also help to promote behavioural change. A substantial part of biomass burning in
industrialised countries is currently providing secondary heat, and therefore alternative options exist
in many households. Over the next few decades however, large changes to the provision of domestic
heat will be required, which may involve switching to electricity based systems, expanded use of
heat networks and/or the use of non-carbonaceous fuels such as hydrogen. How biomass fits into
future low carbon strategies for heat, requires careful consideration of resulting potential changes to
AQ and pressures on supply chains. System wide studies including LCAs with a scope broad enough
to include both climate and AQ relevant emissions could help to support overall strategy
development for the domestic heat sector.
Within developing countries, particularly in rural environments, domestic fuel burning for cooking
and heating is likely to remain a significant environmental challenge that must be addressed in order
to improve both public health and climate impacts. Here though, there are opportunities to improve
sustainability goals in both these areas, since currently used inefficient stoves lead to over usage of
precious fuel resources, as well as to high emissions of CO2, NOx, PM and VOCs. Gasification
technologies do exist for domestic cook stoves, and interventions are required to promote their
large scale uptake, potentially providing positive benefits to both human health and GHG emissions,
although further work may be needed to investigate ultra-fine particle emissions. In all areas of
biomass usage for energy, we should be aiming for such win-win strategies that combine
improvements for climate and air quality outcomes and thus human health.
Alison Tomlin is Professor of Environmental Modelling at the School of Chemical and Process
Engineering, University of Leeds. She obtained a BSc in Mathematics and Philosophy of Science, and
a PhD in Nonlinear Chemical Kinetics, both from the University of Leeds. She then held PDRA
positions at Leeds and Princeton Universities, joining Faculty at Leeds in 1994. Research interests
include the combustion of alternative fuels and air quality impacts of combustion. She is a Fellow of
the Combustion Institute and was co-Editor of PROCI (2012-2018).
TABLE OF NOMENCLATURE
REFERENCES