This project has now been completed. Please find the final project report through the following link - The contribution of indoor domestic solid fuel burning to Scotland’s air pollution. Final report to Scottish Government - NERC Open Research Archive

Questions 1 and 2 
1. How much domestic biomass burning takes place in Scotland? 
2. How much pollution does this cause? 
 

Questions 1 and 2 were addressed in Work Package (WP) 1, which was sub-contracted by SRUC to UKCEH (UK Centre for Ecology and Hydrology). The methodology and results are comprehensively described and discussed in the draft report provided to RESAS by UKCEH on 1st June 2025 (hereafter cited as Di Marco et al., 2025; Appendix 1). We consider the objectives of WP1 to have been achieved in full. 
 

Task 1.1 Biomass burning inventorying 
Task 1.2a Measurement quantification of biomass burning contribution to PM2.51 in village/ small town Task 
1.2b Measurement quantification of biomass burning contribution to PM2.5 for understudied smoke-controlled area Task 
1.2c Analysis of Scottish Black Carbon network sites for biomass burning aerosol 
 

Methods and Results from Tasks 1.2a, b and c are described in Chapter 2 of Di Marco et al. (2025). For practical reasons, measurements were first made in an area of central Edinburgh (Task 1.2b, winter of 2022/2023) and then in a rural location in Fife (Task 1.2a, winter of 2023/2024). Methods of choice were aerosol mass spectrometry and multiwavelength aethalometry, as described in report sections 2.1.2 and 2.1.3. The latter section also describes the derivation of estimates of black carbon (soot) produced during domestic biomass burning. 

Results in Task 1.2b (section 2.3) confirmed some intuitive expectations, such as the diurnal distribution of biomass burning (peaking in the evening) and its relatively even spatial distribution across the city (compared with emissions from traffic, concentrated in the city centre). A novel finding was the relatively large contribution from cooking, likely arising from restaurants and hospitality venues. Aethalometer data showed that domestic biomass burning accounted for 16% of black carbon emissions, with the remainder generated by traffic. 

Results in Fife (Task 1.2a, section 2.4) showed that, compared with Edinburgh, there was less of an evening peak in emissions from traffic (less of a rush-hour) and an absence of emissions from cooking (lack of nearby restaurants). Emissions from domestic coal and wood burning could be distinguished: both showed peaks in the evening, as expected. Of black carbon (soot), 27% was due to wood burning (a higher proportion than in Edinburgh) and the rest to traffic. The timings of peaks in emissions from biomass burning suggest more regular use throughout the week, as a primary source of domestic heat, compared with the urban pattern suggesting greater use of wood and coal fires at the weekend. 

In Task 1.2c (section 2.5), results from the two sites used in this project (Edinburgh and Fife) were compared with data from two other sites in Scotland (rural Auchencorth Moss and urban Glasgow Townhead) that are part of a long-term DEFRA study. Black carbon emissions from biomass burning accounted for between 4 and 6% of the concentration of PM2.5, across all four sites. Concentrations of black carbon (‘cwood’) were higher from Nov-Mar, coinciding with the season of biomass burning.

Task 1.3 Model quantification of biomass burning contribution to PM2.5 

After generating new data for the contribution of biomass burning to PM2.5 at two specific locations in Task 1.2, the objective of modelling this contribution, at high resolution (1km2), across Scotland was pursued in Task 1.3. This work is described in Chapter 3 of Di Marco et al. (2025). 

Compared with the original project plan, this task was delayed2, in agreement with RESAS, in order to take advantage of a DEFRA fuel survey conducted in 2022/23 and published in 2024, which repeated a similar survey conducted in 2018/19. Compared with the 18/19 survey, the 22/23 survey suggests an increase of 260% in the estimate of domestically burned wood, and an increase in Scottish wood burning from 7% to 12% of the UK total. Use of the 22/23 survey in this project will therefore inevitably mean an increase in the estimate of PM2.5 from domestic burning. 

Emission Factors were derived for specific combinations of fuel type (separating, for example, seasoned and unseasoned wood) and burning device (e.g., different ages of wood burning stove). Several surprising observations were made based on this work, for example low emissions for seasoned wood (and unseasoned wood) and burning device (e.g., different ages of wood burning stove). Several surprising observations were made based on this work, for example low emissions for seasoned wood (<20% moisture) compared with dried wood, and lack of lower emissions from modern stoves compared with their immediate predecessors.

The distribution of domestic biomass burning (all solid fuels burnt indoors) is mapped at 1km2 scale in Figure 3.3 of Di Marco et al., (2025). Hot spots not surprisingly coincide with locations with high population density.

Results from this mapping exercise were compared with two other published sources. Several differences are noted (e.g., Table 3.9). Compared with the National Inventory (2021), total PM2.5 emissions in Scotland from the domestic burning of wood, coal and MSF, as estimated in this project, are 22% higher, while emissions from wood are 30% lower.

Finally, bringing all the preceding work in Task 1 together, the impact of these new estimates of emissions on air pollution was modelled using the EMEP4UK model3, as described in Chapter 4 of Di Marco et al., (2025). Perhaps the most salient excerpt is: ‘…the magnitude of PM2.5 contributions from Scottish domestic solid fuel burning is generally small ((<0.5 μg/m3, compared to total concentrations 4-8 μg/m3) except in hotspots around small towns such as Dumfries, Fort William and Ayr.’

The validity of the EMEP4UK model, as used in this project, was checked by comparing its predictions with the measurements made at the sites used in this project (Edinburgh and Fife) and in the DEFRA (2024) report (Auchencorth moss and Glasgow). As described in section 4.3, the model performed well in 3 of these 4 sites, the exception being Edinburgh, where the model may overpredict emissions from coal and MSF. Model uncertainties are discussed in Chapter 5of Di Marco et al. (2025).

Question 3. Does domestic biomass burning affect human health? 

This question was due to be addressed by SRUC in Work Package 2. 

Much of the work anticipated in WP2 was dependent on WP1 and was therefore affected by the same delays (e.g., availability of relevant DEFRA data for Task 1.1 and Task 1.2c). Loss of staff resource then coincided with the decision of RESAS to terminate the project. This objective was not achieved, and the question remains unanswered. 

Task 2.1 Identification of the spatial distribution of life expectancy, and respiratory or coronary diseases in Scotland 

A background report (Degiovanni, 2025; Appendix 2) on methodologies that could be used to model the spatial distribution of life expectancy and respiratory and coronary diseases was prepared as part of Task 2.1. However, these models were not applied to available health data, and Task 2.1 was not completed. 

Task 2.2 Identification of the spatial distribution of sources of air pollution (other than those associated with biomass burning) across Scotland 

UKCEH ran the EMEP4UK model six times to separate and spatially map different sources of air pollution (see Chapter 4 of Di Marco et al., 2025): 
1. Reference run with all emissions included. 
2. Rest-of-the-UK run with domestic solid fuel emissions for non-Scottish UK sources removed (i.e. England, Wales & N. Ireland). 
3. No-wood run with all wood and wood product emissions removed for Scotland 
4. No-coal run with all house coal/lignite/peat emissions removed for Scotland 
5. No-MSF run with all MSF emissions removed for Scotland 
6. No solid fuel run with all solid fuel emissions removed for Scotland 
 

Results (annual and winter averages for the pollutants PM2.5, NOx, NO2 and SO2) are shown in Figures 4.3 to 4.10 of the UKCEH report. These model runs provide the data required for completion of Task 2.2. Run 1 (panel (a) in Figures 4.3 to 4.10) minus Run 6 (panel (f) in Figures 4.3 to 4.10) represents the spatial distribution of sources of air pollution other than those associated with domestic biomass burning. 

Task 2.3 Measurement of the effect of PM2.5 concentration levels due to biomass burning on health outcomes in Scotland 

This task would have used existing information on the spatial distribution of life expectancy and the incidence of respiratory and coronary disease (gathered and modelled in Task 2.1) and information on the magnitude and spatial distribution of pollutants from biomass burning (WP1 and Task 2.2) to assess the possible impact of biomass burning on health. 

Little progress had been made before the project was closed at the end of March 2025. 

Task 2.4 Evaluation of the effect of the Low Emissions Zones (LEZ) on health outcomes in Scotland 

The intention of this task was to assess the impact of LEZ on health outcomes and to use the spatial distribution of emissions and health outcomes (from Tasks 1.3, 2.1 and 2.2) to identify opportunities for new LEZ. In the final project proposal, it was anticipated that this task would be performed in the fourth year of the project (2025-2026), by which time LEZ were expected to have been fully geographically identified and their effects on emissions computed. 

Scotland’s LEZs were introduced on 31 May 2022 with Glasgow beginning enforcement on 1 June 2023, Dundee on 30 May 2024, and Aberdeen and Edinburgh on 1 June 2024 (https://lowemissionzones.scot/about). 

No work was undertaken on this task before the project was closed at the end of March 2025. Given their relatively recent dates of introduction, more time may be needed before impacts of existing LEZ on emissions are sufficiently well quantified. However, the opportunity to use the work done in this project on the spatial distribution of air pollutants as a baseline when evaluating effects of LEZ and opportunities for future LEZ, is clear. 
 

References: 

Degiovanni, H (2025) Task 2.1.1 Review of Methodological Approaches 
Di Marco CF (2025) The contribution of indoor domestic solid fuel burning to Scotland’s air pollution. Draft final report to Scottish Government
 

Outcomes

1. The estimate of Scottish PM2.5 emissions from solid fuel burning made in this project is higher than previous estimates (e.g., National Inventory 2021), due partly to higher estimates of the quantity of solid fuels used, and partly to higher emission factors describing the emissions per unit of fuel 
a. An exception is the burning of wood, where a reduction in the emission factor outweighs a higher estimate of the amount burned, resulting in a reduction in national emissions from this fuel type

2. A number of observations challenge assumptions about PM2.5 emissions from domestic biomass burning, such as: 
a. New emission factors show limited benefit of switching from coal to MSF 
b. Modern stoves may not have lower emissions than older stoves, although more data are needed
c. Seasoned wood may have lower emissions than pre-dried wood

3. The project provides insights into the spatial and temporal (diurnal and seasonal) distribution of emissions from domestic biomass burning. These insights arise from direct measurement in two specific locations and from application of an atmospheric chemistry and transport model (with a high degree of temporal and spatial granularity) to the whole of Scotland: 
a. From direct measurements, the proportional contribution of biomass burning to PM2.5 emissions was greater in rural Fife (17%) than in urban Edinburgh (8%). Restaurant cooking was identified as an underappreciated source of emissions in Edinburgh. 
b. From application of the model, the largest local values (for PM2.5 concentration from domestic biomass burning) are in the central belt between Bathgate and Livingston (with a major contribution from MSF). The contribution from wood burning was proportionally greater in small towns in rural areas, such as Fort William.

4. There was generally good agreement between modelled and measured estimates of emissions except for: 
a. Modelling accounted for secondary PM2.5 formed from gases generated by solid fuel burning which are not included in the measurements 
b. The model may overestimate the amount of coal and MSF burnt in Edinburgh 

5. The contribution to PM2.5 in Scotland from biomass outside Scotland is relatively small and confined to the Scottish Borders (most likely originating from Carlisle and vicinity). 

6. The contribution of biomass burning to NO2 (of concern to human health) is very small (around 1%) 

7. There is evidence of long-term decline in PM2.5 concentration (Glasgow, from DEFRA statistics)

Project Insights
UK-wide surveys of domestic fuel usage and domestic biomass burning practices conducted by DEFRA added great value to the project (e.g., making project outputs of greater value to the next National Inventory) but also caused delay. The rationale for this change was identified early by UKCEH and accepted by RESAS. 
Closer involvement of SRUC scientists responsible for WP2 in the later stages of work by UKCEH scientists on WP1 (in late 2024) would have allowed a faster start to work to explore relationships between PM2.5 emissions and human health. 
The decision to terminate the project (therefore not answering the question of whether emissions of air pollutants from domestic biomass burning affect human health) should not detract from the high value of work conducted by UKCEH in WP1 and WP2 Task 2.2. This work represents a major advance in our understanding of the spatial and temporal distribution of air pollutants, particularly PM2.5, in Scotland, and the contribution of domestic biomass burning to those pollutants. This work adds value to future National Inventory calculations and asks questions of practical relevance to the significant proportion of the Scottish population who use wood, coal and MSF in domestic settings. 
The opportunity to use the outputs of WP1 to pursue the questions not answered by the aborted WP2 (associations with human health and evaluation of LEZ) remains.

Next Steps/Future Plans
The logic of co-mapping air pollutants and health outcomes remains and could lead to targeted, impactful actions to reduce emissions to deliver improvements in human health. It is recommended that new projects are initiated to complete the work not delivered by this project.
 

Objective 1. Improve the emission inventory for biomass burning for Scotland
Recent UK projects provide revised activity figures for the NAEI 2021 inventory (National Atmospheric Emissions Inventory, 2023 release) and revised emission factors. We accessed these data and processed emission fields for various solid fuel types for input to our Atmospheric Chemistry and Transport Model (Objective 3 below). We are awaiting updated data from DEFRA's current survey of fuel use across the UK, to update information for Scotland. 

Objective 2. Quantify the contribution of biomass-burning aerosol to PM2.5 (PM - Particulate Matter) in understudied settings (database to inform emission inventory work and for model assessment) 
We completed the analysis of the winter PM measurement campaign 2022/23 targeting a smoke controlled area (Edinburgh University Drummond St Campus) and are nearing completion of measurements in a rural village setting (winter 2023/24; Charlestown, Fife). We compared two quantifications of biomass burning aerosol (black/brown carbon using a three-wavelength aethalometer and positive matrix factorisation of aerosol mass spectrometer data), finding large inconsistencies of the former, more commonly applied approach. Results show clear presence of biomass burning emissions with a diurnal cycle, lasting later in the day than other sources such as traffic and cooking. Analysis suggests average wintertime concentrations of burning-related PM of 0.2 µg/m3, peaking at 0.4 µg/m3 around 8-9pm (exceeding the organic aerosol contribution from tailpipe emissions). Measurements highlight cooking-derived PM (assumed dominated by deep-frying) at this city location: a poorly controlled source offering opportunities for further reduction. Because cooking occurs throughout the year, it likely makes a larger annual contribution than domestic biomass burning. Comparison of modelled and measured contributions are reasonably consistent. 

Objective 3. Assess the contribution of biomass burning to PM2.5 concentrations through high resolution modelling. Coupling the 2020 NAEl with 2019 meteorology at area of 3 km x 3 km, we have explored the contribution of domestic solid fuel burning. We used a state-of-the-art atmospheric chemistry and transport model (EMEP4UK) to simulate the PM concentration before and after removal of emissions from domestic burning of wood and smokeless fuels from the inventory. The difference quantifies the PM due to these sources. At an annual average, this modelling suggested maximum contributions of about 0.2 to 0.3 µg/m3, peaking at about 5% of the total PM. For January (representative winter month) the average contribution increases to about 0.7 µg/m3, peaking east of Glasgow and locally exceeding 10% of the PM. We have now calculated meteorology at area of 1 km x 1 km and are preparing the NAEI 2021 (2023 release) for the final calculations, now also including the impact of house coal. Following comments from NatureScot, we have started to investigate the contribution of wild fires (in Scotland mainly representative of muirburn) through an improved satellite-derived dataset on fire counts to quantify its contribution.

Results are continuing to emerge and solidify and have not yet been communicated in written form to ScotGov. With the village campaign still ongoing and the final model runs awaiting input from a DEFRA-funded study, first indications are that domestic burning of solid fuels makes a significant contribution to the PM concentrations both in smoke controlled cities like Edinburgh and rural settings. We still need to put these results into context of total PM2.5 concentrations encountered during this period. Measurements are broadly consistent with the emissions represented in the recently revised NAEI. The measurements highlight cooking (presumably dominated by deep frying in restaurants and takeaways) as a contributor that, in city centre locations, is potentially larger (on the basis of the annual average) than domestic fuel burning. This source is currently also poorly controlled and policed. In the urban setting both sources appear to be larger than primary tailpipe emissions during winter.

Results are still emerging and solidifying, as the measurements still require full quality control assessment. First indications are however that domestic burning of solid fuels makes a significant contribution to the PM concentrations in smoke-controlled cities like Edinburgh. These results still need to be put into the context of total PM_2.5 concentrations encountered during this period. First indications are that the measurements are broadly consistent with the emissions represented in the recently revised NAEI. The measurements highlight cooking (presumably dominated by deep frying in restaurants and take-aways) as a contributor that is potentially larger at the annual average for city centre locations than domestic fuel burning. This source is currently also poorly controlled and policed. Both sources appear to be larger than primary tailpipe emissions during winter.