SUPPLEMENTARY INFORMATION: UNDERSTANDING THE UK GREENHOUSE GAS INVENTORY

NPL REPORT CCM 2 SUPPLEMENTARY INFORMATION: UNDERSTANDING THE UK GREENHOUSE GAS INVENTORY An assessment of how the UK inventory is calculated and the implications of uncertainty DAVID BUTTERFIELD COMMISSIONED BY THE COMMITTEE ON CLIMATE CHANGE MAY 2017 i

Supplementary information: Understanding the UK Greenhouse Gas Inventory An assessment of how the UK inventory is calculated and the implications of uncertainty David Butterfield National Physical Laboratory Commissioned by the Committee on Climate Change ii

NPL Management Limited, 2017 ISSN 2056-9572 National Physical Laboratory Hampton Road, Teddington, Middlesex, TW11 0LW Extracts from this report may be reproduced provided the source is acknowledged and the extract is not taken out of context. The National Physical Laboratory is operated by NPL Management Ltd, a wholly-owned company of the Department for Business, Energy and Industrial Strategy (BEIS) Approved on behalf of NPLML by Heather Powell Group Leader, Emissions and Atmospheric Metrology iii

I. CONTENTS I. CONTENTS... IV 1 EMISSION TOTALS CALCULATION METHOD & DATA SOURCES... 1 1.1 OVERVIEW BY IPCC SECTOR... 1 1.2 DATA SOURCES... 4 1.3 CALCULATION METHOD FOR POWER STATIONS IPCC CATEGORY: 1A1a... 5 1.4 CALCULATION METHOD FOR GAS LEAKAGE IPCC CATEGORY: 1B2b4 & 1B2b5... 6 1.5 CALCULATION METHOD FOR ENTERIC FERMENTATION- IPCC CATEGORY: 3A... 8 1.6 CALCULATION METHOD FOR MANURE MANAGEMENT IPCC CATEGORY: 3B... 10 1.7 CALCULATION METHOD FOR AGRICULTURAL SOILS - IPCC CATEGORY: 3D... 11 1.8 CALCULATION METHOD FOR CROPLAND IPCC CATEGORY: 4B... 14 1.9 CALCULATION METHOD FOR SETTLEMENTS IPCC CATEGORY: 4E... 16 1.10 CALCULATION METHOD FOR LANDFILL IPCC CATEGORY: 5A... 17 2 CCC SUB-SECTOR BREAKDOWN OF EMISSION AND UNCERTAINTIES FOR CO 2, CH 4 AND N 2O... 21 2.1 CO 2... 21 2.2 CH 4... 37 2.3 N 2O... 53 3 OVERVIEW OF CHANGES TO CALCULATION METHODOLOGY FOR 1990 EMISSION TOTALS... 69 3.1 2004 INVENTORY (2006 SUBMISSION)... 69 3.2 2005 INVENTORY (2007 SUBMISSION)... 69 3.3 2009 INVENTORY (2011 SUBMISSION)... 69 3.4 2010 INVENTORY (2012 SUBMISSION)... 69 3.5 2013 INVENTORY (2015 SUBMISSION)... 69 3.6 2014 INVENTORY (2016 SUBMISSION)... 70 iv

IPCC Sector 1A 1 EMISSION TOTALS CALCULATION METHOD & DATA SOURCES 1.1 OVERVIEW BY IPCC SECTOR The following table lists the methods used to calculate emission totals for each IPCC category along with the approach used for the activity data and emission factors. Common Method statements are used across different categories for consistency. Comments on calculation method Tier Activity Data Emission Factor Basic combustion module (fuel use * emission factor) Transport models (see MS 7, MS 8 and MS 6) Carbon balance approach (See MS 4) 1B Carbon Balance approach (See MS 4) 2A DECC EEMS inventory (See, MS 18) Gas leakage data from network operators (See MS 20) Cement production: IPCC Tier 2 approach Lime production: Approach is comparable to IPCC Tier 2, although the Tier 1 default factor is used in the reporting of emissions. Glass:, Brickmaking:, Other carbonates 2 3 3 3 3 3 2 1 3 3 3 Bottom up sectorial compared against top down fuel use Aviation, road transport & off road machinery Fuels used in coke ovens, sinter plant, and blast furnaces Fuels used in coke ovens, sinter plant, and blast furnaces UK study of closed coal mines Natural gas leakage data in energy and mass units DUKES DUKES Site specific agglomerated to give estimates for each sub-sector Site specific Site specific Specific EFs + carbon content of fuels Carbon content of aviation fuel, specific road vehicle type EFs + carbon content of fuel EU ETS data or operators of integrated steelworks EU ETS data or operators of integrated steelworks Modelled Natural gas compositional data Operator-reported EFs from EU ETS Operator-reported EFs from EU ETS UK-specific factors from EU ETS UK-specific factors from EU ETS UK-specific factors from EU ETS 1

IPCC Sector 2B Comments on calculation method Tier Activity Data Emission Factor Emissions calculated based on emissions data from industry, EU ETS and the environmental regulators inventories, except for: o Use of IPCC default factors for CH 4 from ethylene oxide, acrylonitrile, carbon black in years where no environmental regulators inventories data available o Use of IPCC default factor for CO 2 from ethylene dichloride across full time series 3 1 1 Site specific Installation specific review Installation specific review UK-specific factors from EU ETS IPCC default factor IPCC default factor 2C Iron and Steel - 2 stage carbon balance and EU ETS/operator carbon factors for carbonate use and arc furnaces (see MS 4) Spreadsheet model and operator reported emissions for aluminium and magnesium production. Tier 1 approach for non-ferrous metal production 2D Emissions calculated based on IPCC defaults for non-energy use of fuels IPCC method based as a proportion of the amount of fuel consumed for urea consumption in road transport 2E, 2F Spreadsheet model to estimate emissions of F- gases 2G Spreadsheet model to estimate emissions of F- gases NHS research into anaesthetic use Statistics on cream consumption and Danish inventory assumptions for N 2O as a propellant for whipped cream Pollution inventory data for other uses of N 2O; 2 2 1 1 3 DECC energy statistics & ISSB BGS & operators Estimates are based on emissions data reported by the process operator in the Pollution Inventory DUKES Proportion of fuel consumed and proportion of Euro IV & V HGVs and buses UK-specific factor from carbon balance, UKspecific from EU ETS UK-specific factors plus operator reported data, based on IPCC T2 method Estimates are based on emissions data reported by the process operator in the Pollution Inventory IPCC default factor EMEP/EEA Guidebook 2 & 3 Literature review + Expert Judgement Carbon Emission Factor + Country specific data 1,2 & 3 1 3 2 Royal Air Force Industry, Universities NHS UK cream consumption data are available from (DEFRA) Process operators IPCC default emission factor Carbon Emission Factor + Country specific data EU GHG inventory derived emission factor Danish Inventory Method Process operators, emissions from chemical sources, not combustion. 2

IPCC Sector Comments on calculation method Tier Activity Data Emission Factor 3A Emissions calculated based on animal 1 & 2 Agricultural Census IPCC Tier 1 defaults except for population data and appropriate EFs Dairy, beef cow and lambs Country specific 3B Emissions calculated based on animal 1 & 2 Agricultural Census Tier 2 except for deer which is Tier 1 population data and appropriate EFs 3D Emissions calculated based on animal 2 IPCC Tier 2 Methodology Country specific emission factor for direct N 2O population data, fertilizer data and appropriate and IPCC for indirect N 2O. EFs 3F Emissions calculated based on IPCC methodologies and USEPA EFs 4 Mathematical models used to estimate emissions and removals from Land-Use and Land Use Change CARBINE model used to estimate emissions and removals from Forestry, provided by Forest Research 5A The Methane Emissions from Landfill model (MELmod) 5B UK waste activity data and IPCC default emission factors 5C Country specific emission factors, partially based on Pollution Inventory data 5D IPCC default method using country specific activikty data for all N 2O and CH 4 from private waste-water management systems and industrial waste-water treatment Data from operator returns to the regulator for water company waste-water management 1 Crop production data No estimates since 1993 due to the cessation of field burning 1 & 3 Agricultural Censuses, Fire & Rescue Service, Satellite data IPCC default emission factors Changes in carbon stock, IPCC default for non- CO 2 3 Forestry Research survey Changes in carbon stock, IPCC default for non- CO 2 2 Local Authority data on waste sent to Landfill Country specific 1 Annual organics recycling reports IPCC default emission factors 1 & 2 Pollution Inventory & Expert Judgement Literature studies Country specific 1 & 2 Municipal waste-water treatment volumes UK-specific factors UK-specific per capita Biochemical Oxygen Tier 1 N 20 Demand Tier 2 CH 4 UK population data UK waste-water estimates UK-specific factors Table 1: Methods used to calculate emission totals for each IPCC category and the approach used for the activity data and emission factors. Common Method statements are used across different categories for consistency. 3

1.2 DATA SOURCES Source (and publisher) Relevant activity data contained in the source Digest of UK Energy Statistics (UK Department of Business, Energy and Industrial Strategy [BEIS]) DUKES Emissions Trading System (EU ETS regulatory agencies in the UK; data supplied via BEIS) EU ETS Transport Statistics GB (UK Department for Transport) TSGB Northern Ireland Statistics: Inventory of Statutory Releases, transport data (NI Department of the Environment, NI Department for Regional Development) ISR Civil Aviation Authority CAA Pollution Inventory (Environment Agency and Natural Resources Wales) PI Scottish Pollutant Release Inventory (Scottish Environment Protection Agency) SPRI United Kingdom Petroleum Industry Association UKPIA Environmental Emissions Monitoring System (EEMS) (BEIS Offshore Inspectorate) EEMS UK Iron and Steel Industry Annual Statistics (International Steel Statistics Bureau) ISSB United Kingdom Minerals Yearbook (British Geological Society) UKMY Annual Abstract of Statistics (Office for National Statistics) ONS Department for Transport ANPR Table 2: Sources and publishers of relevant data Energy statistics for the UK (imports, exports, production, consumption, demand) of liquid, solid and gaseous fuels Calorific values of fuels and conversion factors Emissions from installations and characteristics of fuels consumed Energy data are aggregated by sector and used to inform inventory estimates Fuel quality data are used to derive up to date carbon emission factors for major fuels in energy intensive sectors. Vehicle km according to vehicle type and road type Vehicle licensing statistics (split in vehicle km by fuel type) Selected domestic and international civil aviation aircraft km flown Traffic count and vehicle km data for Northern Ireland Information on regulated processes in NI Detailed domestic and international civil aviation aircraft km flown Information on emissions from regulated processes in England and Wales Information on regulated processes in Scotland Refinery emissions; Lead and sulphur contents of fuels, benzene content of petrol, RVP of petrol. Detailed inventory of oil and gas emissions Energy production and consumption in the Iron and Steel industry Other statistics regarding the Iron and Steel industry. Statistical data on minerals production, consumption and trade Population data Automatic Number Plate Recognition (ANPR) data used to help define fleet composition on different road types in the UK. 4

1.3 CALCULATION METHOD FOR POWER STATIONS IPCC CATEGORY: 1A1a Relevant Gases: CO2, CH4, N2O Background The main fossil fuels used by the UK electricity supply industry are bituminous coal and natural gas. The number of coal stations has decreased markedly across the time series 1990-2014, and the number of gas fired stations has increased. The share of electricity from coal and gas in 2014 was 30% and 30%, respectively. Bio-fuels are burnt at an increasing number of power generation sites to help electricity generators meet Government targets for renewable energy production. Electricity is also generated in a large number of engines running on biogas at landfill sites and sewage treatment works. CO 2 emissions associated with biofuel combustion are estimated and reported as memo items, but not included in this category; these emissions are reflected in LULUCF carbon stocks. Electricity is also generated at an increasing number of Energy from Waste (EfW) installations in the UK. All such installations are now required to be fitted with boilers to raise power and heat, and their emissions are therefore reported under CRF source category 1A1 (electricity generation), rather than 5C (Waste Incineration). The following table lists the UK power stations by fuel type. Year Coal Fuel oil Gas oil Gas Waste Biomass Biogas Nuclear Fission 1990 44 9 11 0 2 0 Unknown a 19 1995 23 8 11 5 4 0 Unknown a 16 2000 22 5 11 35 15 4 267 15 2005 17 5 12 47 20 5 461 12 2010 17 4 12 53 24 6 554 10 2012 17 2 12 59 26 8 565 10 2013 15 2 13 53 28 10 621 10 2014 14 2 15 53 33 10 628 10 Table 3: UK power stations by fuel type 1990-2014 Note: a Number of power stations for early years is unknown although emissions are reported, biogas consumption is obtained from DUKES. Key data sources Activity data: Emission Factors: DUKES, EU ETS, UK PIA Carbon factors are predominantly derived from EU ETS data (2005 onwards) and from the 2004 Carbon Factors Review [1], with some solid fuel factors derived from UK research [2]; non-co2 EFs are predominantly IPCC defaults (IPCC, 2006 1 ). Method approach The calculation of direct greenhouse gases from the power stations is calculated according to: Emission estimate = Emission Factor Activity Data 1 References to IPCC, 2006 corresponds to [3] throughout 5

Uncertainties Uncertainties for both activity and emission factors are based on expert judgement. Uncertainties in fuel use statistics are typically low. The carbon emission factors are based on UK specific data. Since there is a direct link between the carbon emitted and the carbon content of the fuel, it is possible to estimate CO 2 emissions accurately. The uncertainty in the CO 2 emission quantity is ~2%. Non-CO 2 emissions are dependent on a greater number of parameters, and are largely based on IPCC defaults and hence have larger values. The uncertainty in the CH 4 emission quantity is ~38% and the uncertainty in the N 2O emission quantity is ~58%. However, emission totals for these 2 gases are very small compared to CO 2, so the total CO 2 equivalent emissions for public electricity and heat production (1A1a) is ~2%. Assumptions & observations For the 2014 inventory, default emission factors for CH 4 and N 2O from the 2006 IPCC Guidelines have replaced old and uncertain UK-specific factors for some sector-fuel combinations. In the case of methane, the IPCC default factors are mostly higher so their use yields generally more conservative emission estimates. In the case of N 2O, the IPCC factors are mostly lower than the previous factors, so emission estimates are now lower. The new N 2O factors are approximately 70% of the previous figures for energy sector coal/coke use and this leads to a significant reduction of N 2O emissions from 1A1a, equal to a reduction in total UK emissions of N 2O in 2013 of about 1%. DUKES reports less fuel burnt by power producers than is reported by operators either directly to the Inventory Agency or via the EU Emissions Trading System (EU ETS). Therefore fuel oil, gas oil, and burning oil are reallocated from industry (1A2) to power stations to ensure consistency with operator data, while maintaining consistency with the overall fuel consumption data in DUKES. Activity data and emission factors for this sector are well understood, which results in low uncertainties in emission totals. 1.4 CALCULATION METHOD FOR GAS LEAKAGE IPCC CATEGORY: 1B2b4 & 1B2b5 Relevant Gases: CO2, CH4 Background The UK GHG inventory includes estimates of methane and carbon dioxide emissions from natural gas leakage from the downstream gas supply network, including releases from: high pressure transmission network; distribution network & gas leaks at point of use. Key data sources Activity data: Emission Factors: Natural gas leakage data in energy and mass units, from the UK downstream natural gas network operators: National Grid, Scotia Gas, Northern Gas Networks, Wales & West, and Airtricity (NI). Natural gas compositional data supplied by the gas network operators. UK estimates of natural gas consumption within each local distribution zone are used to generate a weighted-average UK compositional analysis of natural gas consumed annually. From 2007 these data are available from Long Term Development Plans published by each of the gas network operators. Earlier are based on Local Authority-level consumption estimates aggregated into LDZs. 6

EFs for the gas leakage at point of use are derived from UK data on gas fitting performance and assumptions regarding unit operational cycles, ignition times. Method approach The leakage estimates are calculated using separate methodologies to cover: 1. Natural gas leaks from the high-pressure transmission mains (1B2b4 Transmission); 2. Natural gas leaks from the low pressure distribution network, medium pressure gas mains, Above Ground Installations (AGIs), AGI working losses and interference (1B2b5 Distribution); 3. Other losses of natural gas at the point of use (1B2b5 Distribution). For methods 1 & 2 from 2004 onwards the gas network operators provide annual gas leakage estimates on a mass basis, providing a breakdown of emissions across the LDZs. In addition, each of the gas network operators provides annual natural gas compositional analysis for their networks. Prior to 2004, the data on gas leakage (activity data and compositional analysis) was all provided by British Gas, which operated all of the UK networks. Methane losses from the high pressure transmission system (1B2b4) are estimated by National Grid based on periodic fugitive emission surveys for the NTS, compressor stations and LNG terminals, and National Grid records of intentional venting actions on the network. Methane loses for the low pressure distribution network (1B2b5) are based on the aggregate mass of gas leaked across all networks (low pressure mains and other losses), with the methane content of based on compositional analysis from all of the gas network operators. UK Gas Network Leakage Model The UK gas network operators use a common industry leakage model to derive their annual estimates of gas leakage from the low and medium pressure distribution systems. The UK gas network leakage model was developed by British Gas and uses factors and assumptions on leakage rates for different types of gas mains and installations, based on measurements and surveys conducted in 1992 and 2002, with annual updates to maintain the representation of the UK gas network infrastructure (such as length and type of pipelines and other units) and reflect the rolling programme of network replacement. Historical data for the leakage from the low-pressure distribution network and other losses is based on studies from British Gas in the early 1990s. Point of use The third inventory estimation methodology is used to determine estimates of natural gas leakage at the point of use, and these estimates are also reported in 1B2b5. Leakages are estimated for a range of different appliances that use gas, combined with national statistics on natural gas consumption in the domestic and commercial sectors. The number of boilers from 1990 to 2014 is thought to have increased (ca. 22 million in 2008) due to the increasing use of gas central heating for space heating, and the increase in the number of houses. However, it is assumed that pre-ignition gas loss in boilers installed in houses in 1990 were greater than in the current boilers installed, as technology has improved. Therefore, it is assumed that the proportion of gas leaked (i.e. % of the total gas use) from domestic heating and water heating appliances per annum is steady across the time series. Uncertainties Uncertainties in the emission estimates from leakage from the gas transmission and distribution network stem predominantly from the assumptions within the industry model. For above ground installations the methane content of the gas released is known to a high degree of accuracy, but the mass emitted is based on industry calculations. 7

The uncertainties for the estimates of gas leakage at point of use are high due to the lack of source data, an IPCC method and the need to use a series of assumptions and expert judgement to estimate the leakage from different gas appliance types. Methane is the predominant emission from this IPCC category. The uncertainty in the CH 4 activity data or natural gas transmission is 3%, while the uncertainty in the related emission factor is 20%. This gives a combined uncertainty of 20.2% in the emission total of CH 4 from natural gas transmission. Observations Measurements of methane emissions from high pressure compressor and distribution facilities would help to characterise the nature and magnitude of these losses, thus reducing the uncertainties associated with relevant emission factor data. Improvements in the UK Gas Network Leakage Model would help to reduce uncertainties in the calculation of methane losses in the medium and low pressure distribution network and improve the quality of emission factors. The calculation method in 2016 was updated resulting in the following recalculations: 1) 1B2b4 Natural Gas (transmission leakage) National Grid provided updated data for transmission network leakage in 2012, and a correction to previous data for 2011 (data transcription error by data provider) was also identified through Inventory Agency quality checks. These changes affected the 2005 to 2010 time series as these were previously interpolated between2004 and 2011. 2) 1B2b5 Natural gas (distribution leakage) Activity data revised due to changes in gas distribution category. Previous submissions had combined category data. 3) Natural Gas (leakage at point of use) Estimates of natural gas use in domestic and commercial appliances in recent years have been significantly revised by DECC (see 2015 DUKES report [3]) leading to revisions in the estimates for gas leakage at point of use from 2009 onwards. 1.5 CALCULATION METHOD FOR ENTERIC FERMENTATION- IPCC CATEGORY: 3A Relevant Gases: CH4 Background The agriculture sector has the second largest contribution to total GHG emissions in the UK, after the energy sector. It contributes approximately 8.7% to the total emissions. The emissions from this sector have shown an overall decrease of 18% since 1990, reflecting trends in livestock numbers and emissions from fertiliser application. Enteric fermentation makes up 53.6% of the emissions from the agriculture sector. Methane is produced in herbivores as a by-product of enteric fermentation. Enteric fermentation is a digestive process whereby carbohydrates are broken down by microorganisms into simple molecules. Both ruminant animals (e.g. cattle and sheep), and non-ruminant animals (e.g. pigs and horses) produce CH 4, although ruminants are the largest source per unit of feed intake. 8

Method approach Emissions from enteric fermentation are calculated from detailed animal livestock population data collected in the June Agricultural Census, by the devolved administrations, and the appropriate emission factors. Apart from dairy and beef cows and lambs, the methane emission factors are IPCC Tier 1 defaults (IPCC, 2006) and do not change from year to year. Tier 2 methods are used for dairy and beef cows, while the T1 emission factor for lambs has been modified to reflect UK conditions. Dairy cows The dairy cattle emission factors (for dairy cows only) are estimated following the IPCC Tier 2 procedure (IPCC, 2006), using country-specific data for dairy cow live weight, milk yield, milk fat content, feed digestibility and activity (proportion of the year spent grazing) and vary from year to year. A country-specific value for the digestibility of feed (DE), expressed as a percentage of the gross energy, for dairy cows is used of 74.5%. This value is on the high side of the IPCC (2006) default value for Western Europe of 55-75% for pasture fed animals, but is based on typical diets for cows over the lactating and non-lactating period, combining forage and concentrates, with energy values for the various feeds according to MAFF (1990). Beef cows A Tier 2 methodology is used for the calculation of the enteric emissions from beef cows. Mature weights for the different beef size categories were obtained from an analysis of abattoir data (net carcase weight) from four abattoir companies across Great Britain for the years 2008-2013 combined with British Cattle Movement Society (BCMS) data. The digestibility value for beef cows used by the UK is 65% for annual average feed composition, based on UK expert opinion, reflecting the poorer quality diet that beef cows will generally receive in comparison with dairy cows. Other cattle A Tier 1 methodology is used for the calculation of the emissions from other cattle with default emission factors (IPCC 2006 guidelines) and include: dairy cows, beef cows, dairy heifers, beef heifers, dairy replacements > 1 year, beef all other > 1 year, dairy calves < 1 year, beef calves < 1 year. Sheep The UK is currently undertaking a programme of work to improve methodology for calculating emissions from this sector due to the complex structure of sheep farming in the UK, which has many different breeds of sheep and a range of hill, upland and lowland rearing and finishing systems. This will allow the implementation of Tier 2 data and emission factors. The current approach is to assume the IPCC Tier 1 default emission factor for enteric fermentation for all mature sheep (> 1 year old). Other Tier 1 Emissions The UK emission factor for pigs, goats, horses and deer are IPCC default values (2006 guidelines). Uncertainty Emissions are calculated from animal population data and appropriate emission factors. The estimates of uncertainties in emissions were calculated using Approach 2 (Monte Carlo simulation) described by the IPCC for 1990 and 2014. Activity data (animal population) uncertainties are provided by the devolved administrations. Tier 2 methods were used to estimate the emission factors for dairy / beef cows and lambs. For all other animal categories, IPCC Tier 1 emission factors are used. 9

The calculation of uncertainty in activity data and emission factor has been revised for the 2016 submission and has concluded that uncertainty for emission factors and activity cannot be separated, only a combined uncertainty is quoted. The combined uncertainty for 2014 is 13.7%. In the previous method the uncertainty in the activity data was nearly always <=1%, while the uncertainty in the emission factor was in the order of 20% between 1990 and 2013. Observations & Improvements The UK is currently undertaking research to improve activity data on typical forage diets for a range of livestock production systems and aims to provide preliminary data feeding into 2016 submission. The digestibility values for the different forage components are taken from 1990 MAFF data and could be updated by new work. Both these measures should reduce the uncertainty in the emission in the inventory by the move to Tier 2 methods for all animals. 1.6 CALCULATION METHOD FOR MANURE MANAGEMENT IPCC CATEGORY: 3B Relevant Gases: CH4 & N2O Background This category reports emissions of CH 4 from animal manures as well as N 2O emissions from their manures arising during its storage. Method approach Methane emissions from animal manures Methane is produced from the decomposition of manure under anaerobic conditions. When manure is stored or treated as a liquid in a lagoon, pond or tank it tends to decompose anaerobically and produce a significant quantity of methane. When manure is handled as a solid or when it is deposited on pastures, it tends to decompose aerobically and little or no methane is produced. Hence the system of manure management used affects emission rates. Activity data for animal manures are calculated from livestock population data provided by the devolved administrations. The emission factors for manure management are calculated following IPCC Tier 2 methodology for all animals except deer, where default IPCC data (Tier 1) is used. IPCC default values are also used for volatile solids. An IPCC methane producing potential is also used to take into account the different animal waste management systems, i.e. liquid ponding, daily spreading. Nitrous Oxide emissions from Animal Waste Management Systems Animals are assumed not to give rise to nitrous oxide emissions directly, but emissions will arise from N excreted by livestock. Emissions from manures during storage are calculated for a number of animal waste management systems (AWMS) defined by IPCC. The conversion of excreted N into N 2O emissions is determined by the type of manure management system used. Country specific activity data for nitrogen excretion by the different livestock types and the proportion of manure managed according to the different AWMS is based on a 2006 Defra report with additional Expert Judgement. Emission factors are IPCC default factors for the different AWMS. Emissions from the following AWMS are reported under the Manure Management IPCC category: Uncovered anaerobic lagoons. These are assumed not to be in use in the UK; Liquid/slurry; 10

Deep bedding (previously deep bedding); Poultry manure with/out bedding or destined for incineration. According to IPCC (2006) guidelines, the following AWMS are reported in the Agricultural Soils category: All animal manures and slurries applied to soils; Pasture range and paddock. Emissions from the combustion of poultry bedding for electricity generation are reported under power stations. Emissions occurring during storage of poultry bedding that will later be used for energy generation are included in the agricultural inventory. Indirect N 2O emissions from manure management comprise N volatilisation from manure management systems calculated using IPCC 2006 guidelines. Along with country specific fractions derived directly from the UK agriculture ammonia emission inventory, for N loss due to volatilisation of NH 3 and NO x, disaggregated by manure management system. Emissions of N 2O from the leaching/runoff associated with the storage of deep bedding as field heaps have been estimated using IPCC 2006 guidelines and country specific data. Uncertainty Emissions are calculated from livestock population data and appropriate emission factors. The estimates of uncertainties in emissions were calculated using Approach 2 (Monte Carlo simulation) described by the IPCC. The calculation of uncertainty in activity data and emission factor has been revised for the 2016 submission and has concluded that uncertainty for emission factors and activity cannot be separated, only a combined uncertainty is quoted. The combined uncertainties for CH 4 and N 2O for 2014 are 4.8% and 68% respectively. In the previous method the uncertainty in the activity data was nearly always <=1%, while the uncertainty in the emission factor was in the order of 30% for CH 4 and 250% for N 2O. Observations & Improvements The recalculated uncertainty for N 2O in 2014 is a very significant decrease compared to the uncertainty used in previous years, 68% compared to 250%. This is due to the implementation of country specific values derived directly from the UK agriculture ammonia emission inventory (N loss from manure management due to volatilisation of NH 3-N and NO X-N). The evidence for this change is from a literature review and a field measurement programme. 1.7 CALCULATION METHOD FOR AGRICULTURAL SOILS - IPCC CATEGORY: 3D Relevant Gases: N2O Background Agricultural soils is the largest emitter of N 2O in the inventory, making up 70% of the emissions. Direct emissions of nitrous oxide are estimated using the IPCC recommended methodology (IPCC, 2006) but incorporating country specific emission factors and UK specific parameters. The IPCC method involves estimating contributions from: 1. The use of inorganic fertilizer; 2. Application of livestock manures to land; 3. Application of sewage sludge to land; 11

4. Urine and dung deposited by grazing animals in the field; 5. Crop residues returned to soils; 6. Mineralisation; 7. Cultivation of histosols (organic soils). In addition to these, the following indirect emission sources are estimated: 8. Emission of N 2O from atmospheric deposition of agricultural NO x and NH 3; 9. Emission of N 2O from leaching and run-off of agricultural nitrate. Method approach New emission factors were calculated based on IPCC compliant historical experimental data. The experiments covered a range of N sources (mineral fertiliser, livestock manure and dung and urine as grazing returns) to across representative UK sites in terms of soil type and annual cumulative rainfall. Any treatments which included a nitrification inhibitor were omitted from the analysis. Statistical analysis showed there to be no significant differences between N 2O emissions from mineral fertilisers with increasing numbers of application timings compared with the standard fertiliser treatment, so data from these treatments were included in the analysis. For each treatment-site combination, the means were calculated from a minimum of three replicates, and a maximum of six replicates together with standard deviations and standard errors for each of the N sources. In the case of the fertiliser treatments, mean emission factors were derived for those applied to grasslands and those applied to arable land. These were then further subdivided by fertiliser type. There were insufficient data to derive separate emission factors for farm yard manure or slurry applications according to land use type (i.e. grassland and arable), and therefore individual means for N source (farm yard manure and slurry) were calculated across both land types. The grazing returns were based on experiments where urine and dung from cattle had been applied to grassland (Defra project AC0116 [4]), with mean emission factors being derived separately for the dung and the urine components. A weighted emission factor was derived based on the assumption that 60% of total N excretion is as urine and 40% as dung. 1) Inorganic Fertiliser Emissions from the application of inorganic fertilizer are calculated using the IPCC (2006) Tier2 methodology with country specific emission factors applied to different fertiliser types and land use. The annual consumption of synthetic fertilizer is based on crop areas provided by the Devolved Administrations and the British Survey of Fertiliser Practice. 2) Application of livestock manures to land Emissions from animal manures are calculated using the IPCC (2006) Tier 2 methodology with country specific data for the amount of manure nitrogen applied to the land. Country specific emission factors are used for different manure types: a) liquid; b) deep bedding; c) poultry manure without bedding and poultry manure with bedding and destined for incineration. 3) Application of sewage sludge to land Tier 1 Methodology following the IPCC 2006 guidelines. The calculation involves estimating the amount of nitrogen contained per dry matter unit of sludge that is applied to land and applying IPCC emission factors. 4) Urine and dung deposited by grazing animals in the field The method of calculation is the same as that for N 2O emissions from animal waste management systems using country specific emission factors derived from experimental studies (AC0116). 12

5) Crop Residues returned to soils Emissions of nitrous oxide from the ploughing in of crop residues are calculated using the 2006 IPCC guidelines and IPCC default emission factors. Activity data comes from production data of crops, supplied by Defra. 6) Mineralisation N 2O emissions from mineralisation of soil organic matter on land converted to Cropland more than 20 years ago are included in the Agricultural Inventory (emissions from more recent land use change are included in the LULUCF inventory). The emissions are estimated using the areas of Forest land and Grassland converted to Cropland from the land use change matrices. 7) Cultivation of histosols (organic soils) Emissions from histosols are estimated using the IPCC (2006) default factor of 8 kg N 2O-N/ha/yr. The area of cultivated histosols is estimated at 285,700 ha. 8) Atmospheric deposition of NO X and NH 3 Indirect emissions of N 2O from the atmospheric deposition of ammonia and NO x are estimated according to the 2006 IPCC guidelines using default emission factors for fertiliser N application and manure application to soils, and country specific value derived directly from the UK agriculture ammonia emission inventory, for the fraction of N that is volatilised. Another source of NH 3 and NO x is sewage sludge applied to soils for which the default emission factor is used. The method used corrects for the N content of manures used as fuel (poultry bedding incineration). 9) Leaching and runoff Indirect emissions of N 2O from leaching and runoff are estimated according the 2006 IPCC guidelines with the default emission factor. The sources of nitrogen considered, are synthetic fertiliser application and animal manures applied as fertiliser, sewage sludge applied to soils and crop residues. The method used corrects for the N content of manures used as fuel (poultry bedding incineration). Uncertainty The estimates of uncertainties in emissions were calculated using Approach 2 (Monte Carlo simulation) due to the non-normal probability density function (log-log). The calculation of uncertainty in activity data and emission factor has been revised for the 2016 submission and has concluded that uncertainty for emission factors and activity cannot be separated, only a combined uncertainty is quoted. The combined uncertainty for N 2O is 68%. In the previous method the uncertainty in the activity data was nearly always <=1%, while the uncertainty in the emission factor was in the order of 260%. Observations & Improvements Emission factors were reviewed during 2015 based on the evidence from three areas: 1. Emission factor for direct soil emissions; from a literature review and a field measurement programme; 2. Emission factor for manure management systems; from a literature review and a field measurement programme; 3. Emission factor for nitrogen leaching/runoff factor; from a field measurement programme. 13

This has enabled the move from a Tier 1 methodology to a Tier 2 methodology, resulting in the uncertainty in N 2O emissions from soils contributing to 52% of the total uncertainty in N 2O emissions in 2014. In 2013, the contribution from soils was 82%. 1.8 CALCULATION METHOD FOR CROPLAND IPCC CATEGORY: 4B Relevant Gases: CO2, CH4, N2O Background The category is disaggregated into 4.B.1 Cropland remaining Cropland and 4.B.2 Land converted to Cropland. Ongoing carbon stock changes in soils arising from historical land use change to Cropland more than 20 years before the inventory reporting year are reported under 4.B.1 Cropland remaining Cropland, along with emissions from organic soils as a result of drainage. Carbon stock changes from drainage of Cropland on organic soils arise from areas which were drained many decades ago for agriculture, allowing oxygen into previously water logged soils. As a result, soil carbon in these areas continues to oxidise and is released as CO 2, resulting in an ongoing change in soil carbon stock. Changes in soil carbon stock resulting from changes in Cropland Management are also reported under 4.B.1. Carbon stock changes and biomass burning emissions due to conversion of other land categories to Cropland in the previous 20 years before the reporting year are reported under category 4.B.2 Land converted to Cropland (biomass burning emissions occur in the same year as the land use conversion, while loss of soil carbon occurs over a longer period). All forms of land use change, including deforestation, are considered and both mineral and organic soils are included. N 2O emissions from soil disturbance associated with land-use conversion to Cropland are reported, these arise from Forest Land and Grassland being converted to Cropland. Method approach Representation of land use area The UK uses Approach 2 (IPCC 2006) for the representation of land use areas in the inventory. Data sources are available at the individual country level and results are combined to give UK totals. Data sources that contain area information for reporting carbon stock changes and/or emissions from Cropland are habitat/landscape surveys; an assessment of Cropland drainage, and data on wildfires on agricultural land from Fire and Rescue service and satellite data. Areas of Cropland that are a losing carbon due to historical drainage (4.B.1) were been reassessed in 2013. This reassessment gives a more complete picture of the area of Cropland on drained organic soils than previous work in 1997. The vast majority of Cropland on drained organic soils is in England, but small areas in the other UK administrations are also identified. The areas of the main crop types used assess changes in soil carbon stocks resulting from Cropland Management are obtained from the June Agricultural Censuses for each UK administration. The areas of Cropland receiving inputs of manure, fertiliser and crop residues are obtained from the British Survey of Fertiliser Practice. Both of these are collected on an annual basis. From 2010 areas of wildfire on Cropland are taken from Fire and Rescue service data. Between 2001 and 2009 the area of wildfire on Cropland is calculated by using satellite data on the total area of wildfires in the UK which are apportioned to land use using the same ratios as found in the Fire and Rescue service data. Cropland wildfire areas prior to 2001 are extrapolated. Land-use definitions and classification system 14

Cropland is defined in accordance with the Agriculture, Forestry and Other Land Use Guidance (IPCC 2006). Post-1980, Cropland is the area of cropland reported in the annual June Agricultural censuses, and broken down by type as recorded by the census. For pre-1980 land use matrices cropland is the sum of the Crops and Market Garden land cover types in the Monitoring Landscape Change project (MLC 1986). Methodological Issues A dynamic model of carbon stock change is used with the land use change matrices to estimate soil carbon stock changes due to all land use change, including change to and from Cropland. In the model soil carbon stock changes follow an exponential path between initial and final land uses with the most rapid change in the early years after land use change. The carbon stocks for each land use category are calculated as averages for Scotland, England, Northern Ireland and Wales using a database of soil carbon density for the UK (1997, 1998, 2005). N 2O emissions associated with the conversion of land to Cropland are reported using the areas of Forest land and Grassland converted to Cropland from the land use change matrices and the IPCC Tier 1 emission factors. Carbon stock change in soils as a result of Cropland Management is estimated using Tier 1 stock change factors for most activities, although a Tier 2 stock change factor is used for tillage reduction. Activity data is from agricultural surveys. Carbon stock change in biomass as a result of Cropland Management is estimated using literature derived Tier 2 stock change factors and activity data is again obtained from agricultural surveys. Emissions from Cropland on drained organic soils are reported using Tier 1 emission factors which assume constant rates of carbon loss and activity data from expert judgement. Emissions from wildfires on Cropland are reporting using a Tier 1 emission factors and activity from the Fire and Rescue Service s Incident Reporting system from 2010 onwards, remotely sensed FIRMS thermal anomaly data from 2001 2009 and extrapolation prior to this. Uncertainty There are very large differences in the uncertainties for activity data and emission factors for CO 2, CH 4 and N 2O. In all cases the uncertainty in the activity data is 1% due to the annual Agricultural Survey, while the uncertainty in emission factor is 45%, 55% and 55% for CO 2, CH 4 and N 2O respectively. Therefore the combined uncertainties for these 3 GHGs are the same as the uncertainties in emission factor. The total uncertainty in CO 2 equivalent emissions for this category is 43.8%. The areas undergoing land use change are the biggest source of uncertainty in the inventory. Other significant sources of uncertainty are parameterisation of the forest model. Emissions from cropland on drained organic soils has the largest uncertainty of the minor emissions sources (i.e. not land use change) as the effects of drainage are highly uncertain. Work in implementing the Wetlands Supplement may decrease this uncertainty. Observations & Improvements The main change between the 1990-2013 inventory and the 1990-2014 inventory is the inclusion of biomass carbon stock changes arising from Cropland Management activities. Corrected deforestation data has also been used to improve the inventory. The uncertainty in the land use change areas is currently being addressed by the development of a new vectorbased approach, combining multiple sources of land use data and will be used for future submissions. 15

The methodology and emissions factors for calculating emissions from controlled burning following deforestation were updated to follow the IPCC 2006 guidance. Previous inventories had used the methodology and emissions factors from the IPCC 2003 guidance. To assess the uncertainty due to drainage work, implementing the IPCC Wetlands Supplement [5] is underway and this may improve the emission factor for cropland on drained organic soils. Parameterisation of the forest model is the second largest source of uncertainty and the move to the CARBINE model (19 tree species) from the CFlow model (2 tree species) has helped to reduce the uncertainties. Results from the latest National Forest Inventory has also provided additional information on carbon stocks in trees [6]. 1.9 CALCULATION METHOD FOR SETTLEMENTS IPCC CATEGORY: 4E Relevant Gases: CO2, CH4, N2O Background This category is disaggregated into 4.E.1 Settlements remaining Settlements and 4.E.2 Land converted to Settlements. Ongoing carbon stock changes in soils and direct N 2O emissions from N mineralization arising from historical land use change to Settlements more than 20 years before the inventory reporting year are reported under 4.E.1 Settlement remaining Settlement. Carbon stock changes, N 2O emissions from N mineralization and biomass burning emissions in the previous 20 years before the reporting year are reported under category 4.E.2 (biomass burning emissions occur in the same year as the land use conversion). All forms of land use change, including deforestation, are considered and both mineral and organic soils are included. Method approach Representation of land use area The activity data on areas of Forest Land converted to Settlement (deforestation) from 2000 onwards has been updated with data collated from multiple sources. This has substantially reduced the estimated area of forest land converted to settlement by 0.4-0.7 kha per year from 2000 onwards. Before 2000, data on forest-urban land conversion in England was obtained from the Ordnance Survey. Land conversion ratios from Countryside Survey were then used to extrapolate from England to the other countries in the UK. Land-use definitions and classification system Settlement is defined in accordance with the Agriculture, Forestry and Other Land Use Guidance (IPCC 2006). For pre-1980 land use matrices Settlement land is the sum of the Builtup, Urban open, Transport, Mineral workings and Derelict land cover types in the 1986 Monitoring Landscape Change project Since 1980, settlement land corresponds to the broad habitat types: Builtup and Gardens and Boundary and linear features as reported in the 2000 Countryside Survey. Some components of the Boundary and linear features type could fall under the definition of Cropland or Grassland. It is not possible to disaggregate this broad habitat further and the assignment to a single land use category avoids double-counting. In the latest 2007 Countryside Survey Boundary and linear features covered 2% of the UK land area. Methodological Issues A land use matrix approach and dynamic model of soil carbon stock change is used to estimate changes in biomass and soil carbon due to land use change. 16

Land use conversions or land management changes that result in a loss of soil carbon, leading to N mineralization and N 2O emissions were reported for the first time in the 2013 submission, reflecting updated guidance in the 2006 AFOLU guidance. The Tier 1 methodology described in the IPCC 2006 Guidelines is used. IPCC default C:N ratio of 15 is used for estimating mineralised N. UK expert opinion considers the values reported in the Countryside Survey are not applicable as different depths of soil are evaluated by the Survey and the Inventory. Uncertainty There are very large differences in the uncertainties for activity data and emission factors for CO 2, CH 4 and N 2O. In all cases the uncertainty in the activity data is 1% due to the data collected in the Countryside surveys, while the uncertainty in emission factor is 50%, 20% and 20% for CO 2, CH 4 and N 2O respectively. Therefore the combined uncertainties for these 3 GHGs are the same as the uncertainties in emission factor. The total uncertainty in CO 2 equivalent emissions for this category is 47.4%. The areas undergoing land use change are the biggest source of uncertainty in the LULUCF Inventory, but model choice and soil carbon parameters are also significant. Observations & Improvements The activity data on areas of Forest Land converted to Settlement (deforestation) from 2000 onwards have been updated with data collated from multiple sources. This has substantially reduced the estimated area of forest land converted to settlement from 2000 onwards. The methodology and emissions factors for calculating emissions from controlled burning following deforestation were updated to follow the IPCC 2006 guidance. Previous inventories had used the methodology and emissions factors from the IPCC 2003 guidance. The uncertainty in the land use change areas is being addressed by the development of a new vector-based approach, combining multiple sources of land use data and will be used for future submissions. The collation of multiple deforestation datasets should reduce the uncertainty in this area, a full assessment has started but is yet to be completed. Work is being undertaken on carbon stock changes in perennial biomass in cropland and grassland: this will allow hedgerow areas (permanent vegetative boundaries between agricultural fields) to be separated out from the Boundary and Linear features habitat type and moved from the Settlement category to the Grassland category. 1.10 CALCULATION METHOD FOR LANDFILL IPCC CATEGORY: 5A Relevant Gases: CH4 Background The NAEI category Landfill maps directly on to IPCC category 5A Solid Waste Disposal for methane emissions. Emissions are reported from landfills that started receiving waste in 1980, when legislative changes took effect to improve management of landfill sites, and old unmanaged waste disposal sites that closed prior to 1980. Estimated emissions from this sector in 2014 were 13.5 Mt CO 2e. Emissions have been on a downward trend since 1996. Emissions from this category contribute 71% of the emissions from the waste sector. In addition to CH 4, anaerobic decomposition also produces an approximately equivalent amount of carbon dioxide and further CO 2 is also produced by aerobic decomposition processes. However, as the decaying organic matter originates from biomass sources derived from contemporary crops and forests, the emission of this carbon dioxide is not included in the. Waste also contains fossil-derived organic matter, predominantly in the 17